Friday, 1 May 2015

T-72: Part 1

The T-72 is known as one of the most numerous tanks in the world today, and the tank is likely to remain in some form of service for the rest of the 21st Century, so before we examine the history of the T-72, let us see how many of these tanks were actually built. A few years ago, Uralvagonzavod (UVZ) made efforts to declassify much of the history of the T-72 tank, including the number of tanks built by the factory. The table below comes from the factory archives of UVZ, published in the book "T-72/T-90: The Experience of Creating Domestic Main Battle Tanks" authored by S. Ustmantsev and D. Komalkov (head designer of the UVZ transport engineering bureau).

From 1973 to 1990, a total of 18,373 T-72 tanks and T-72 derivatives were manufactured at the UVZ factory floor and another 1,600 tanks were manufactured from 1991 to 1996. The Chelyabinsk Tractor Factory also took part in the manufacture of the T-72 tank, producing 1,894 units themselves between 1978 and 1990. In total, 20,267 T-72 tanks were produced in Soviet Russia, making it the second most numerous tank ever produced in both the USSR and the world, outstripping the T-62 for that title by a slim margin. But how did it come about? The 2010 book "T-72 Ural armor versus NATO" by noted military historian Mikhail Baryatinsky details the history of the development of the tank, and is the source for many of the diagrams and pictures shared below.

Firstly, it should be clear that the T-72 is indeed a "mobilization model" with slightly inferior performance compared to the T-64. Some Internet sleuths found this chart of prices showing that the T-72A (1979) was significantly more expensive than the T-64A, despite the fact that the technology of the T-72 Ural was also very conspicuously inferior to the T-64A in several major ways. The more expensive pricing of the T-72 does not change the fact that it is a less sophisticated product compared to the T-64, although it was not originally intended to be such by the chief designer of the UKBTM design bureau Leonid Kartsev. In fact, the original T-72 outpaced the T-64A in some technical areas due to the implementation of certain technologies even though the T-64 series could still be rightfully considered more effective in terms of overall combat capability.

Let us take a look at the Object 167M. It is a precursor to the T-72, but is better described as a T-62 taken to the extreme.

The Object 167M featured the now-famous AZ autoloader, composite armour for the upper glacis and turret, a 125mm D-81T cannon, a V-26 engine which developed 700 HP, a reinforced transmission to deal with the increased power, hydraulically powered gear shifting systems, the new "Liveni" two-plane stabilizer system, and a new suspension composed of six roadwheels with three return rollers. The tank did not enter mass production because the Object 432 had already been ordered to enter production under the designation of "T-64" by a resolution from the Council of Ministers of the USSR and the Object 167M had numerous drawbacks of its own. In February 26, 1964, the scientific-technical council GKOT examined the Object 167M project and rejected it. This was the end of the road for the Object 167M, but it was destined to leave its mark on Soviet tank history, as we shall see later on.

The Minister of Defence Industry S.A Zverev came to the experimental workshop of the Uralvagonzavod factory on October 26, 1967, the 50th anniversary of the October Revolution. After an invitation to enter and examine a modified T-62 with a 125mm gun (a special development in response to Kharkov's failure to begin mass production of the T-64 on schedule) by UKBTM design bureau chief Leonid Kartsev, the minister requested a demonstration. The designer E.E Krivosheya and researcher L.F Terlikov who were loitering around the tank quickly joined the two inside, started it up and activated the autoloader. The minister was impressed by the autoloader and was enthusiastic on the idea of putting it in the T-64, but rejected Kartsev's proposal to also replace the 5TDF engine on the T-64A with a supercharged derivative of the V-2 engine (from the T-34) developed in Chelyabinsk. He only agreed on changing the engine on the next day after a private meeting with the head of the Military Industrial Commission I.V Okunev. It was agreed that six T-64A tanks were to be sent to UVZ to be given the modifications specified by Kartsev, but the minister also dictated that the suspension and chassis were to be untouched. On 5 January, 1968, Zverev officially gave the order to begin the "modernization" of the T-64A by the Uralvagonzavod factory. Thus the life of the T-72 began, but for the next few years, all prototypes of Kartsev's new tanks would either be modifications of these six T-64A tanks or modified copies thereof.

The first of these modifications, dubbed "Object 172", was completed in the summer of 1968, and the second was completed in September of that same year. The Object 172 differed from the T-64A obr.1969 only in the fighting compartment, which had to be rearranged to fit the new autoloader, and in the engine compartment, which was completely reworked for the V-45K engine and T-54-style cooling system. 

Because the modifications were focused only on internal components and the engine compartment, the Object 172 was practically indistinguishable from a typical T-64A obr.1989 from the front, but from the rear, the turret was clearly modified for the autoloader's ramming and ejection system, while the engine deck was completely distinct from that of the T-64 series. The engine compartment itself was also lengthened to accommodate the new powertrain and cooling system, so the tank is not as compact as the T-64 in length. The left side of the hull gained an exhaust port just above the second rearmost roadwheel - a location reminiscent of the earlier T-54 and T-62.

Baryatinsky's book gives us the details of the late stages of gestation of the T-72. Here are a few translated paragraphs:

"Then in the Design Bureau of the UVZ, which since August 1969 was headed by V.N. Venediktov, it was decided to use the chassis of the Object 167 with rubberized roadwheels of increased diameter and more durable tracks with open metal track pins similar to those of the T-62 tank. The development of this tank was carried out under the designation "Object 172M". The engine, boosted to 780 hp, received the index of V-46. A two-stage air-cleaning cassette system was introduced, similar to that used on the T-62 tank. The weight of the Object 172M increased to 41 tons, but the mobility characteristics remained at the same level (author's note: same level as the T-64A) due to the increase in engine power by 80 hp, the capacity of fuel tanks by 100 liters and the width of the track by 40 mm.

From November 1970 to April 1971, Object 172M passed a full cycle of factory tests and then on May 6, 1971, was presented to the defense ministers A.A. Grechko and the defense industry SA. Zverev. By the beginning of the summer, an installation lot was set up from 15 vehicles, which, together with the T-64A and T-80 tanks, passed many months of unprecedented scale. At the suggestion of Major-General Yuri M. Potapov, a battalion composed of platoons of three companies was formed. At the same time, each company was manned by tanks of the same type. The route of the traffic was chosen from Dnepropetrovsk through Ukraine to Belorussia to Slutsk and then back to Dnepropetrovsk, and then through the Donbass and the North Caucasus to Baku, across the sea by ferry to Krasnovodsk, through the Karakum desert and the Kopetdag mountain range. The tests were due to be completed at a range of 60 km from Ashgabat. During the march, live firing tests were conducted at various firing range, and platoon and company level exercises with live firing and driving were carried out at various tankodroms (training grounds).

After the end of the tests, a report with the title "Report on the results of military trials of 15 tanks 172M, manufactured by Uralvagonzavod in 1972." was submitted. The final part of the report contained these remarks:

  1. Tanks have passed the tests, but the lifespan of the tracks of 4,500 - 5,000 km is insufficient and does not fulfill the requirement for tank travelling capability of a distance of 6500 - 7000 km without replacement of tracks.
  2. Tank 172M (warranty period - 3000 km) and engine V-46 (350 m / h (?)) worked reliably. In the course of further testing up to 10,000 - 11,000 km, most of the units and assemblies, including the V-46 engine, operated reliably, but a number of major units and assemblies showed insufficient lifespan and reliability.
  3. The tank is recommended for adoption into the armed services and serial production, provided that the identified shortcomings are eliminated and the effectiveness of their elimination is checked before serial production. The scope and time frames for improvements and inspections should be agreed between the Ministry of Defense and the Ministry of Defense Industry.

In accordance with the decision of the Central Committee of the CPSU and the Council of Ministers of the USSR of August 7, 1973, Object 172M was adopted by the Soviet Army under the name of T-72 "Ural". The official order of the Minister of Defense of the USSR was published on August 13, 1973. In the same year, an pilot batch of 30 tanks was produced at Uralvagonzavod."

And thus, the T-72 was born. An amalgamation of the T-64A and the Object 167M, the T-72 would go on to become the second most widely produced tank in the world, behind only the T-54.

In the end, the T-72 turned out to be so similar to the T-64 that you could not easily tell them apart, yet different enough that there was minimal parts commonality between them. This was one of the many headaches caused by the rivalries in the Soviet tank building industry, but this does not change the fact that the T-72 was an extremely capable tank. When viewed in terms of cost efficiency, the T-72 was second to none during its heyday and still remains a relatively competitive product to this day if modernized.

Although many T-72 models could be considered technically inferior to a T-64 or T-80 counterpart from the same period, the difference in combat capabilities was generally quite small. According to captured Soviet documentation from 1977 presented in this CIA report, the combat potential coefficient of the T-64A is 1.50 and the T-72 is rated at 1.50. Thus, the T-72 was considered to be the equal of the T-64A during the earlier years of its life. However, the gap in capabilities appears to widen as the years go by: the combat potential of the T-64B is rated at 2.10 and the T-72 "with TPD-K1" is rated at only 1.70. However, the T-72A (1979) did not yet exist by the time the document was published (1977) whereas the T-64B (1976) was already in mass production, so it is not so easy to directly compare the T-64B with the closest T-72 counterpart. Generally speaking, however, the determining factor appears to be in the fire control system and the corresponding abilities of the tank. Indeed, the inclusion of the TPD-K1 laser rangefinding sight apparently gave the T-72 a 0.2 point advantage over the basic model. The additional advantages enjoyed by the T-64B includes its ability to fire guided missiles which the T-72 family lacked until 1985 when the T-72B began mass production. The T-72 series also retained the same TPD-K1 sight in one form or another throughout its career in the 80's and into modern times whereas the T-64 series served until 1987 with the substantially more advanced 1A33 fire control system beginning with the T-64B obr. 1976. However, the cost efficiency of the T-72 was a different matter.

In the book "Combat Vehicles of Uralvagonzavod: T-72 Tank" published by the Information and PR department of Uralvagonzavod, it is claimed that the money spent on purchasing a single T-80U could otherwise be invested in three T-72Bs (with an additional 16,000 Rubles), so in other words, the funds required to purchase two T-80U tanks could cover the cost of a platoon of T-72B tanks as well as enough spares to last several years. One may be reminded of the stereotype of the Red Army favouring hordes of inferior tanks instead of a numerically small collection of technically superior ones, but this is simply not the case. Having a large number of tanks is necessary to ensure that tank support is regularly available to the infantry and there must be enough tanks to fulfill certain strategic objectives. Furthermore, being able to deploy platoons instead of individual tanks or to deploy companies instead of platoons leads to a exponential increase in the chances of success in any given military task.

Referring to the theory of cost efficiency detailed in the book "Tanks: Tactics, Technology, Armament" by Yu.P Kostenko, the Uralvagonzavod editors claim that the calculated cost efficiency coefficient of the T-72B is 3.38 whereas the calculated cost efficiency coefficient of the T-80U is only 1.25. Theoretically, the cost efficiency of the T-72B is 2.7 times higher. The method by which combat effectiveness is quantified is a complex affair so it is difficult to independently verify these numbers using third party (i.e non-UVZ) primary sources, but the hard numbers are merely academic. The high cost effectiveness of the T-72 undoubtedly had a large influence on the monumental export success of the tank, especially considering that the core customer base for Soviet and Russian arms are Second World or Third World nations that typically do not have a large procurement budget.

On another note, it is interesting to observe that although the turret of the T-72 lacks handrails for tank riders like preceding Soviet tanks, the practice of hitching a ride was still occasionally taught and exercised.

But before we take a look at the T-72 in earnest, we must first remember that the original Ural variant underwent several major upgrades throughout its lifetime, creating significant discrepancies between each successive model. To complicate matters, each model in itself may have subtle improvements implemented during overhauls. For example, the transition from the T-72 Ural to the T-72A was gradual, with many of the changes embodied by the so-called "Ural-1" model that contained a mixture of features from both the T-72 Ural and the T-72A. Since 1977, the T-72 Ural-1 already began to be produced with a new turret with a "Kvartz" non-metallic filler. This turret was standard for the T-72A (Object 172M-1) and became associated with it. The transition of the T-72A to the T-72B was similarly difficult to track. According to Alexey Khlopotov, the 172.10.077SB turret commonly associated with the T-72B entered production in September 1982 and the T-72A began to receive these turrets in 1983 together with a new chassis and a new upper glacis armour design. These late model T-72A tanks have the external appearance of a T-72B, but are not actual T-72B tanks (Object 184). Without going into very much detail, we can condense the evolution of the T-72 tank into a few main models. Some of the information below comes from Russian military historian A.V Karpenko.

Object. 172M (T-72 Ural) 1973-1974

The original T-72 model with a monolithic cast steel turret and optical coincidence rangefinder-based sighting system. The IR spotlight was originally located on the left side of the cannon like the T-64A, but it was relocated to the right side in 1974 in order to improve driver safety.

Object. 172M1 (T-72 Ural-1) 1975-1979

In this model, the "Gill" armour panels on the side of the hull from the T-72 Ural (originally from the T-64) were replaced with conventional side skirts sometime in the middle of the production run. The optical coincidence rangefinder-based sighting system was replaced by a laser rangefinder-based version sometime during the production run, at an unknown point. Modified turrets lacking the protrusion for the second optic port for the coincidence rangefinder were devised for these variants. The tank also began to receive a thermal shroud on the cannon barrel in 1975.

Object. 172M-1 (T-72A) 1979-1985

First total upgrade of the tank. Almost everything was changed; the tank had a revised hull armour and a new turret with a composite filler was implemented, the D-81TM cannon was installed, the 902A "Tucha" smoke grenade system was added, a new convoy light with a digital numerical display was installed, and more.

Object. 184 (T-72B) 1985

Second serious upgrade of the T-72. The new tank featured completely revised hull and turret armour, a new autoloader, a guided missile firing capability, a new cannon, a new engine, new sighting systems, and more.

Object. 184-1 (T-72B1) 1985

Simplified T-72B variant without the missile firing capability and with the original Ural autoloader. This aspect of the T-72B1 is examined later on in the article, in the section on the autoloader.

Again, it must be stressed that this is only a very basic list of variants. It is unwise to generalize with regards to the T-72, as the model designation sometimes does not reveal the full story.

Table of Contents

  1. Commander's Station
  2. TKN-3M
  3. Commander's Fire Controls
  4. Communications

  5. Gunner's Station
  6. Sighting Complexes
  7. TPD-2-49
  8. 1A40
  9. 1A40-1
  10. Auxiliary Sights
  11. TPN-1-49-23
  12. TPN3-49
  13. 1K13-49
  14. 1A40-4 Sosna-U

  15. Stabilizers
  16. 2E28M "Sireneviy"
  17. 2E42-2 "Zhasmin"
  18. 2E42-4 
  19. Meteorological Mast
  20. D-81T Cannon
  21. 2A26M-2
  22. 2A46
  23. 2A46M
  24. 2A46M-5

  25. Ammunition Stowage
  26. Autoloader
  27. Loose Stowage

  28. Ammunition
  29. HE-Frag
  30. HEAT
  31. APFSDS

  32. PKT Coaxial Machine Gun
  33. NSVT Anti-Aircraft Machine Gun

Due to length restrictions, this article has been divided into two parts. Part two is available here.


From Stefan Kotsch's fantastic website

The commander's station is somewhat cramped compared to an M60A1 or a Leopard 2, which can be exacerbated by bulky winter clothing, but it is still noticeably less cramped than the gunner's station, which is suitable for the commander since his duties involve more movement. If we refer to this diagram from "Human Factors and Scientific Progress in Tank Building" by M.N. Tikhonov and I.D. Kudrin as provided by Peter Samsonov, we can see that the commander of a T-72 apparently has much less space (0.615 cubic meters) compared to a T-55 commander (0.828 cubic meters), but this is obviously not possible. For one, the commander in a T-55 has to wrap his legs around the gunner seated in front of him because there is simply not enough legroom and the breech guard squeezes him against the turret wall where the radio is located. It is the exact opposite for the T-72. As the commander's station in the T-72 is completely separated from the gunner's station, there is nothing in front of him below chest level, and as a result, he has an abundance of legroom and there is an abundance of vertical height. It is perfectly possible for extremely tall people to command a T-72 without any ergonomic issues, and the commander can stretch his legs out as far as he desires even when the turret rotates. By all accounts, the T-72 is actually quite user-friendly for a tank of its time.

According to Sergey Suvorov, the use of an carousel-type autoloader in the T-72 as opposed to a basket-type autoloader as in the T-64 reduced the vertical space in the tank by around 25 cm. However, the height of the crew compartment in the turret was still more than sufficient. The commander in a T-72 does not have the abundance of space that an M60A1 commander may be accustomed to, but it is a sizable improvement over the T-54 and T-62 as much of the equipment attached to the wall of the commander's station (like the bulky radio) has been moved forward so as to free up more space for his shoulders, as you will see in the many photos in this article. In the T-72, the commander's seat is only a few inches above the tallest point of the false floor of the turret so the commander sits with his legs outstretched. Overall, the T-72 definitely offers more space for the commander than a T-55, though that may not be very high bar to pass in the first place. It is likely that the volume figures provided in "Human Factors and Scientific Progress in Tank Building" are calculated by modeling each crew station as a cylinder where the diameter represents the available width and the height represents the height as measured from the tip of the crewman's foot to the top of his head. This would not accurately represent the actual space available in this case.

According to a media presentation from the TASS news agency, the combat compartment (turret) has a volume of 5.9 cubic meters, the driver's compartment has a volume of 2.0 cubic meters and the engine compartment has 3.1 cubic meters of volume, for a grand total internal volume of 11.0 cubic meters. This is corroborated by "Боевые Машины Уралвагонзавода: Танк Т-72" published by the Uralvagonzavod Production Association which states that the volume of the combat compartment is 5.9 cu.m, the volume of the driver's compartment is 2.0 cu.m, and the volume of the engine compartment is 3.1 cu.m, for a total of 11.0 cubic meters. However, Vasily Chobitok writes in "Основы теории и история развития компоновки танка" ("Fundamentals of the theory and history of the development of the tank layout") that the total internal volume of the T-72 is 11.8 cubic meters.

Internally, the width of the turret is equal to the diameter of the turret ring, which is 2,162mm wide, and the width of the lower part of the crew compartment is equal to the internal width of the hull, which is around 1,810mm wide (calculated by subtracting side hull armour and anti-radiation lining from external width of hull). The turret ring diameter of the T-72 is considerably larger than the turret ring diameter of the Leopard 2 (1,980mm) and is equal to the M1 Abrams (2,160mm). The main culprit of the smaller internal volume of the turret of the T-72 is the teardrop geometry which maximizes the frontal protection of the turret with minimal armour mass.

The commander's cupola follows the same pattern set by the cupola of the T-54, but with some significant differences. The T-72 cupola is slightly taller, has a more thickly armoured hatch, and the hatch has a clam shell shape rather than a simple semicircular shape. In terms of width, the two cupolas are very similar in diameter. The race ring of the T-72 cupola extends below the turret roof, whereas the one in the T-54 cupola does not. This is because the T-72 cupola has an additional toothed ring that engages with the counter-rotating motor - we will explore this further later on.

T-72 (left), T-55 (right)

The cupola housing is secured to the cast steel turret roof by a ring of bolts around its circumference, but unlike the T-54 cupola, the bolts are sheltered against gunfire and the weather. The T-72 cupola is also more complex as its race ring is not directly connected to the fixed base bolted to the fixed cupola housing but to an intermediate metal band, and that connects to the cupola housing via a larger race ring. The intermediate metal band is between the inner cupola (which carries the optics and hatch) and the fixed base, and the anti-aircraft machine gun mount is installed on this band. By releasing a locking mechanism, the intermediate band can be freed from the fixed base, thus allowing it and the machine gun installed on it to be rotated degrees independently of the rest of the cupola, as you can see in the photo below (photo from Russian Ministry of Defence).

This is also apparent in the photo below showing that the cupola is completely independent from the machine gun mount.

The independence of the machine gun mount from the cupola is demonstrated in this video, and in this video. This video from TV Zvezda shows a fully assembled turret with the machine gun cradle on its mount, traversed to a forward position. By separating the machine gun mount from the cupola, the commander does not have to bear the weight of the heavy NSVT machine gun (25 kilograms), the machine gun ammunition and the machine gun mount itself when he rotates the cupola which is especially important because the machine gun is mounted eccentrically to the axis of rotation of the cupola, so the center of gravity would be shifted and the cupola would become imbalanced. Overall, this design feature reduces the amount of physical effort required to rotate the cupola, especially if the tank is on a slope. This issue was solved in the T-64A by using an electric cupola traverse system (part of the remotely controlled anti-aircraft machine gun complex) so that the commander is not required to rotate it manually, but the T-80 and T-80B had its machine gun mounted directly on the cupola and the T-80U skirted the problem by having its anti-aircraft machine gun installed on fixed posts welded to the turret roof at various points. This aspect of the commander's cupola is further discussed in the section on the anti-aircraft machine gun later on.

The diameter of the cupola is 700mm based on drawings of the T-64A turret and other information, but the dimensions of the hatch are more difficult to find. By scaling the TNPA-65A periscope housing to the rest of the cupola in the diagram, the width of the hatch was found to be 665mm, and the maximum length to be 413mm which seems to fit the diameter of the cupola. For comparison, the loader's hatch in the M1A1 Abrams has a diameter of 584mm.

Like in the gunner's station, the commander is ventilated by a single adjustable DV-3 fan, a simple 5.2W fan running on the tank's 27V electrical system. Although it may seem silly simplicity, it is worth noting that the Leopard 1, M60A1 and Chieftain all lacked such a basic amenity. The DV-3 is shown in the photo below.

The DV-3 is closely related to the DV-302T, which is a very similar fan used in aircraft like the Mi-8 helicopter, Il-76 and many more. In other words, the DV-3 was essentially an off-the-shelf product at the time the T-72 began mass production.

Because the commander has his own hatch, he may opt to simply stick himself out of the hatch and ride on the turret roof if he feels uncomfortable inside the tank. For additional ventilation in the crew compartment, the air source for the engine can be switched from the external intake to the crew compartment intake. The crew compartment intake sucks air from the crew compartment (unsurprisingly), which greatly increases the airflow inside the tank if the hatches are open.

Interestingly enough, the thick anti-radiation lining inside the turret and hull of the tank may provide heat insulation for the crew and internal equipment in cold weather. Having a layer of insulation allows the interior of the tank to retain heat better during winter and prevent the personnel heater from heating up the metal of the turret and hull. The insulation may also be useful during summer if shade is available as solar heating of a tank parked in the open can increase the internal temperature of the tank to a dangerous level, sometimes causing heatstrokes and hallucinations. By parking the tank in the shade and leaving the hatches open, the insulated interior of a T-72 could theoretically remain relatively cool after leaving cover.

It may seem trivial, but temperature regulation inside a tank is seriously important. Having a reliable internal heater and a layer of insulation on the walls of the tank is a basic necessity for cold winters - something which the T-72 has and some tanks do not. For instance, the personnel heater in the M60A1 and M60A3 was astonishingly unreliable. According to a 1983 TACOM report titled "M60 Tank Personnel Heater Comparison Test", the original Model "B" heater in the M60A1 was atrociously unreliable and frequently caused the automatic fire extinguisher system in the tank to discharge accidentally. The problems were recognized and the newer Model "C" heater was installed in M60 tanks since May 1980, but these were still extremely unreliable and the issue of frequent accidental fire extinguisher discharges was still not solved. The Model "C" heater had a mean time before failure (MTBF) of just 70 hours and the mean starts before failure (MSBF) was only 25 starts. Without reliable heating, the efficiency of the crew in wintertime would have been severely reduced. Even worse, some tanks like the Chieftain did not have a heater at all, forcing the crew to depend entirely on multiple layers of clothing and thick mittens.

Each crew member was provided with a two-liter aluminium bottle which was stowed in a special holder near their respective stations. Officially, the bottle is meant for drinking water, but it can be filled with anything that the crew desires.

The commander's main means of battlefield surveillance is a forward-facing TKN-3M pseudo-binocular periscope, augmented by two rectangular TNPO-160 periscopes on either side of it and two narrow TNPA-65A viewing prisms aimed to his rear quarters. There is no periscope that allows the commander to see directly behind the turret. For that, he must spin the cupola just a little to one side, and look out of either one of his TNPA-65A viewing prisms. Due to the conformal slant of the gunner's hatch, the commander's view to the left of the turret is largely unimpeded. Even the gunner's night vision sight does not completely block the commander's view as the height of the sight does not exceed the height of the turret roof. In general, it can be said that the commander has an uninterrupted field of view in all 360 degrees. However, the commander's view to the left of the turret in the 10 o'clock to 11 o'clock sector was obstructed when Kontakt-1 reactive armour blocks were installed on the roof of the turret (and on the tank as a whole) beginning with the T-72AV modification. This problem persisted when Kontakt-5 blocks replaced the Kontakt-1 blocks in the T-72B obr. 1989 model and continues to plague the T-90A. All viewing devices are electrically heated using the RTS-27-4A system to prevent fogging in cold weather. RTS stands for "Регулятор Tемпературы Стекла", which means "Glass Temperature Regulator".

The two photos below show the TKN-3MK, a slightly updated version of the TKN-3M. It is impossible to visually distinguish them from each other - the only way to know is to see what model of T-72 you are looking at.

As for equipment, the commander's station is packed chock full of various knick-knacks essential for commanding the tank. There is also an assortment of accessories that are not directly related to his job, but are placed near him because it was the only available space in the squeezed turret.

In the photo above, we can see the R-123 radio transceiver (BLUE) at the very bottom. The silver-gray box above it is a switch box (RED) for the communications system to switch between radio and intercom communication, and the white box beside it is a master control panel (GREEN) for most of the functions in the tank. This control panel (pictured below) gives the commander dominion over things like the lights and the ventilator, and behind the silver and milk-white metal flaps at the corners of the panel are the emergency engine stop button and the emergency fire extinguishing system engagement (activates all the fire extinguishers connected to the automatic firefighting system in the fighting compartment) button, respectively.

The commander is responsible for setting the fuse on HE-Frag shells, and this control panel enables him to do so. Pressing one button partially activates the autoloader so that it stops before the ramming cycle commences. The commander will then use his special fuse setting tool to set the fuse to either the High Explosive or Frag mode. Then, another press of a button finishes the loading procedure.

The silver box (YELLOW) to the right of the intercom switch enables the commander to control the autoloader for the purpose of unloading it.

The box flips open to reveal control toggles for operating the individual elements of the autoloader system, like raising and lowering the shell casing catcher, opening and closing the ejection port, activating the rammer, and so on. If the autoloader is only partially malfunctioning, the commander can use this control box to operate some parts of the loading procedure automatically, and operate other parts manually. If the autoloader carousel malfunctions, it is possible to rotate the carousel manually, and crank the autoloader elevator by hand to extract ammunition and use the electric chain rammer to ram the ammunition into the breech.

Above that is a TN-28-10 dome light and the already-mentioned DV-3 fan. The dome light is part of the PMB-71 lighting system.  At the upper left corner is a wooden dowel with a rubber head. This is a ramming stick for the commander to use when manually loading the cannon. Beside the dome light is the gyroscopic tachometer for the stabilizer system.

Besides the dome light in front of the commander, there is another dome light light directly above the gun breech, making it quite easy for him to perform his duties, including loading and unloading the autoloader and loading the coaxial machine gun.

Here is another view of the station, this time from below. Photo from KyivPost.

Toggle switches for turning on the external and internal lights and the periscope heating system are located around the cupola ring. Two such switches are shown below. The switch on the right is to turn on all of the forward facing lights on the tank, and the switch on the left is to turn on all of the rearward facing lights.


The TKN-3M is a combined pseudo-binocular periscope with night vision capability in two modes; passive and active. The periscope itself has a fairly average angular FOV of 10 degrees in the day channel, or 8 degrees in the night channel. The periscope has two eyepieces, but only one aperture, making it a pseudo-binocular periscope. Since it is not a truly binocular periscope, the TKN-3M offers practically no depth perception. This does not make much difference at long distances, but the viewing experience may take some getting used to. The single aperture of the periscope is seen below.

It has a fixed 5x magnification in the day channel and 3x magnification in the night channel. This is quite limited, making long-distance observation problematic, especially if the weather is unfavourable. It can be manipulated to elevate and depress to a reasonable degree, offering some limited aerial view for the commander. Due to the fact that the periscope is unstabilized, identifying another tank at a distance is very difficult while the tank is on the move over very rough terrain. The commander is meant to bear down and brace against the handles of the periscope to control his line of sight, and that is adequate for keeping the target within view for the smoother parts of off-road driving, but the degree of accuracy is not enough for range finding or precise target designation. Overall, it is not a great system, and it was outstripped as early as 1973 by the new TRP 2A sight installed on the Leopard 1A3, and by the highly advanced PERI-R12 panoramic surveillance and sighting system installed in the Leopard 1A4 in 1974. Both of these devices were capable of very high magnification and had powered traverse with stabilization, and the PERI-R12 had the additional function as the sighting complex for the commander when he used the gunnery override mode. Having this ability in an independent surveillance device was a breakthrough for the early 70's, and many tanks would not have a similar feature until the late 80's or 90's.

In the passive mode of operation, the TKN-3M employs a 1st Generation light intensifier tube, which is usable in lighting conditions as dark as a typical moonless, starlit night (0.005 lux). As the amount of light increases, the effective viewing distance increases. An enemy tank can be identified at up to 400 meters at 0.005 lux ambient light, but identifying the same tank is entirely possible at slightly further distances in moonlit nights, but excess brightness cannot be tolerated as the image intensifier tube may be damaged from the power surge. The two most significant advantages of the passive imaging system is that no infrared light is emitted, unlike an active infrared imaging system, and the image intensifier system enables the commander to detect the minute amounts of visible light emitted from enemy IR spotlights and headlights from long distances.

To switch between the day and night channels, the user simply rotates a dial on the right side of the periscope housing by 90 degrees. This flips an internal mirror by 90 degrees, thus changing the optical path between the night vision unit and the regular daytime optic. The diagram below shows the two choices. Diagram (a) on the left shows the path of the light from the aperture through the night vision system and into the eyepiece, while diagram (b) on the right shows the mirror flipped 90 degrees and the light from the aperture passing through the normal optical channel for daytime use.

The TKN-3MK is a slightly updated variant with a 2nd Generation image intensifier system, producing a brighter and clearer image, but the viewing distance is still hamstrung by the low 3x magnification of the optics. The nominal maximum tank identification range is 500 meters under the same lighting conditions stated before (moonless, starlit nights with ambient light levels of 0.005 lux). 2nd generation image intensifiers differ from their 1st generation counterparts by the implementation of an MCP, a so-called "electron multiplier". The addition of an MCP substantially increases the amplification factor of the device compared to a 1st generation image intensifier, but the price of a 2nd generation image intensifier is also much higher. All T-72B tanks are equipped with the TKN-3MK. Sadly, even the latest modifications of the T-72B3 is also equipped with the TKN-3MK, which is entirely inappropriate for its time.

Besides passive image intensification, the TKN-3M/K features an active infrared imaging system using an electron-optical converter that allows the device to pick up infrared light, convert it into visible light, and amplify it to a sufficient brightness. The active mode requires the use of the OU-3GK IR spotlight which is mounted on the rotating cupola. Inside the spotlight is a rather low-powered tungsten filament lamp designed to run on 110W when connected to the tank's electrical system.

The maximum distance at which a tank-sized target can be identified in the active mode is stated to be around 400 meters, although the spotlight can illuminate objects much further away than that. The main issue is the low resolution of the image and the low magnification factor.

Without the infrared filter, the spotlight emits white light at 240 candelas. The infrared filter prevents all but approximately 0.001% of the light in the visible spectrum (360-760 nm) to pass through, and the light that passes through is at the higher end of the visible spectrum. As such, the spotlight emits a very faint red light with an intensity of around 0.24 candelas that can be perceived by the naked eye at close range when the OU-3 spotlight is activated in low light conditions. The intensity of near infrared light is much higher, of course. This light can be detected by the photosensor of a digital camera without an infrared blocking filter. The photosensor displays this infrared light - which is otherwise invisible - as pink light. This can be seen in the photos below, showing the OU-3 of a BRDM-2 (pictures from kmshik from the Gaz 69 forum).

When the OU-3 spotlight is used without the infrared filter, the TKN-3M/MK can be used at night in the daylight mode possible and the viewing distance can be increased at the cost of exposing the exact location of the tank to everyone. This is permissible in certain circumstances, but it is not common.

Overall, the TKN-3M offers very poor night viewing capabilities compared to modern thermal imaging sights, but it was at least equally advanced as other devices of its type built in the 60's (the TKN-3 first appeared in 1964 on the T-62), and the use of image intensification technology was a novel feature up until the 70's, when the TKN-3M/MK was outstripped by more advanced Western passive image intensifying optics. The TKN-3M/MK was not replaced or sufficiently upgraded as the T-72 entered the 1980's.

The periscope aperture has a small wiper, as you can see in the photo below.

Rotation of the cupola can be done by either using the handgrips on the TKN-3M/MK to slide the cupola around the race ring, or by using the cupola-mounted anti-aircraft machine gun cradle's handles, if the commander is outside the hatch. By rotating the cupola, the commander can attain a full 360 degrees of vision.

At the end of the left hand grip of the TKN-3M is a button to designate a target for the gunner, in the same way as the hunter-killer system in the T-54B using the TPK-1 optic. Unlike the T-54, though, the T-72 features an additional electric motor that automatically counter-rotates his cupola so that his original orientation is preserved while the turret is spinning. The photo below shows the direction sensor, painted red, in contact with the three metal bands on the cupola ring above the golden toothed band. These metal bands interface with a roller inside the direction sensor, and the sensor detects which direction the cupola is facing relative to the turret by detecting the direction in which the roller is deflected. The counter-rotating motor is installed at the turret ring and transmits rotational energy to the cupola via a cardan shaft which can be seen in the photo below. The cardan shaft ends in a drive gear in contact with the toothed band around the circumference of the cupola ring.

Once the turret is slewed towards the target, the gunner will see the target, lay the gun more precisely, and then engage. The commander has duplicated controls for ammunition selection, so can select the most appropriate shell type for the type of target upon spotting it, allowing the gunner to open fire as soon as he has laid the sights on target. This sort of cooperation between the gunner and commander helps the T-72 to attain a higher rate of fire.

It is worth noting that while the target designation system is activated with a single click of the left hand grip button on the TKN-3M periscope, the button can be held down to slave the turret to the commander. Wherever he aims the reticle, the turret will follow. Turret rotation is always done at maximum speed, so small corrections in turret orientation may be a little jerky.

The TKN-3M sight has a stadia reticle intended for approximate manual range estimation of tank-sized targets with a height of 2.7 meters from a distance of 800 m to 3,000 m or 800 m to 3,200 m depending on the TKN-3 variant. It is entirely possible for the commander to see tank-sized targets from several kilometers if weather conditions and the geography of the battlefield allows for it, but in reality, conditions are almost guaranteed to be degraded in some way. The most difficult scenario would involve camouflaged targets hiding in foliage with minimal exposure and without moving.

Stadiametric rangefinding is not an accurate way to determine target distance. At long distances, the errors in estimation may amount to hundreds of meters. This mattered for earlier tanks like the T-54, but the optical or laser rangefinder provided to the T-72 gunner makes the imprecision of stadia rangefinding irrelevant to any real combat scenario.

A horizontal stadia rangefinder is objectively superior to a "choke" type stadia rangefinder, like the type found on M551 Sheridan light tanks. A "choke" rangefinder indicates target distance based on width whereas a horizontal rangefinder depends on height. This means that a "choke" rangefinder would not be useful for measuring distances if the target tank was not oriented directly towards the observer or if the target does not show a full side profile silhouette. For example, a moving tank may be weaving side to side. This keeps the width of its silhouette indeterminate, but the height of the tank would not change. A horizontal-type stadia rangefinder, on the other hand, can measure distance no matter which direction the target is travelling in, and if a tank was in a hull-down position, the height of a tank would generally be halved given that only the turret is exposed, giving the observer a reasonable chance of guessing the approximate target distance with some degree of accuracy. 

As mentioned before, the TKN-3M sight depends on an OU-3GA xenon arc IR spotlight for illumination when operating in the active infrared imaging mode. An inherent shortcoming to the usage of IR spotlights is that the opposing forces using a similar night vision system can also see the beam of light along with its source. The SVD sniper rifle, for example, was fitted with the PSO-1 scope with a solar-powered rechargeable IR filter that gave the designated marksman or sniper the ability to see sources of active infrared illumination without the need for a large image intensification tube and a power source. Devices like this make it easy for the T-72 to be caught in an ambush at night by other tanks of the era like M48s, M60s, Leopard 1s, Chieftains, etc, although it must be said that the inverse also applies. However, the OU-3 design is particularly flawed in this respect because it lacks an occluder. The lack of an occluder means that around half of the light from the spotlight is projected directly forward instead of into the parabolic reflector. As such, an enemy observer will not only see a circular patch of light. When observing a Soviet tank with its IR spotlight on, a large portion of the tank will be brightly illuminated. The additional illumination brings the minor benefit of lighting up the ground for the driver to see more clearly, so the common issue of speed control due to short visibility distance with the complementary IR periscope for the driver is slightly alleviated in battle conditions.


None of the Soviet era T-72 models featured a set of firing controls for the commander. This feature only came on the recent T-72B3 modernization. In the T-72B3, the commander is now equipped to override the gunner entirely. He has a flatscreen display linked to the Sosna-U sight and the necessary controls for firing the main gun and the coaxial machine gun at his disposal in the form of a control unit similar to type used for the PNK-4 sighting complex. This arrangement is no different from what most Western tanks already had for decades. These fire controls are part of the 1A40-4 fire control system. The resolution of the flatscreen display is not known directly, but it is reported that the Catherine-FC thermal imager has an image resolution of 754x576 so it should be safe to assume that the display is configured for this.

Control of turret traverse and gun elevation is accomplished using the thumbstick. The decision to use a thumbstick was because a full joystick could not be easily manipulated with precision if the operator's body and arm was rocking around if the tank were going over rough terrain. However, the thumb would be completely stationary if the hand was securely gripping a handle. The index finger rests on the trigger. The control unit also allows the commander to activate the autoloader for load the cannon.


The T-72 was originally supplied with an R-123 radio. The R-123 radio had a frequency range of between 20 MHZ to 51.5 MHZ. It could be tuned to any frequency within those limits via a knob, or the commander could instantly switch between four preset frequencies for communications within a platoon. It had a range of between 16km to 50km. The R-123 had a novel glass prism window at the top of the apparatus that displayed the operating frequency. An internal bulb illuminated a dial, imposing it onto the prism where it is displayed. The R-123 had an advanced modular design that enabled it to be repaired quickly by simply swapping out individual modules.

Beginning in 1984, the R-123 was replaced by the R-173 radio. The R-173 had a frequency range of between 30 MHz to 75.999 MHz and 10 preset frequencies. It had an electronic keypad for entering the desired frequency, and a digital display. Both the radio and intercom system are directly routed to the throat mike and headset, which are integral parts of the iconic Soviet tanker's helmet.

The throat mike facilitates good voice clarity as it doesn't pick up any ambient noise, making a throat mike system inherently superior to open mikes. However, this is partly counteracted by the low sound quality from the headphones installed in the helmet. Nevertheless, the voice output is loud and clear enough for the task.

Communications through the R-173 are rather easy to intercept and jam or listen in to if the proper instruments are available. For instance, Chechen fighters during the Chechnya campaign were able to listen in to radio chatter and even interject bogus commands over Russian airwaves. For this reason, the new, the development of the frequency-hopping R-168 series was accelerated and the new R-168-25UE-2 radio system was rapidly launched into service in the 2000's to replace it.


The R-168-25UE-2 frequency-hopping encrypted radio system replaces both the R-173M and R-123 radio stations in the T-72B3 and in some T-72BA tanks.


The R-168 family of radios is now the standard throughout the Russian ground forces, from infantry platoons to tank companies. It can produce frequency hops 100 times a second, and the data is encrypted as well. The radio enables the T-72B3 to operate in the new network-centric data sharing system currently implemented in some Russian units.

Besides the updated communications hardware, the tank's intercom and radio control panel was also replaced with an all-new digital one shown below:

Command variants of the Soviet T-72 were equipped with an additional R-123 radio. As of today, the R-123 radio is completely antiquated. It is an analogue design first used in the T-62 back in the early 60's to replace the R-113. Command variants were identifiable via their distinctively elongated second antenna. The modern day Russian army no longer fields command variants of the T-72 due to a drastic shift in combat doctrine. Instead, all modern T-72B3 tanks have only a single R-168-25UE-2 radio allowing all the members of a tank platoon to communicate with each other instead of relying on top-down commands from the platoon commander, who in turn relies on the company commander. Command variants of Soviet era T-72s have reportedly been reverted to their base variants.

Unlike some NATO tanks like the M60A1, the commander's means of surveying the battlefield is conducted with periscopes and not with vision blocks. The commander's head is located below the cupola ring as well. The implications of this design decision is that the commander has rather unremarkable all-round visibility compared to an American tank with their large cupolas and large vision blocks, but like all design decisions, this one has a few advantages of its own. The commander is completely withdrawn from large-caliber sniper fire (12.7mm-type) and concentrated machine gun fire directed at the cupola. There is absolutely zero chance that his eyes may be injured by broken glass due to the nature of the periscope and because the periscope eyepiece is protected by ballistic glass, as shown in the photo below, to the left.

For forward observation, two TNPO-160 forward-facing periscopes are provided. Each periscope has a total horizontal range of vision of 78 degrees, and a vertical field of view of 28 degrees; 12 degrees above the horizontal axis and 16 degrees below.

Two TNPA-65A periscopes bring up the 4 o'clock and 8 o'clock positions. They are mounted directly in the hatch, and give the commander a view of the rear two quadrants of the turret. Unfortunately, there is a blind spot directly behind the cupola, since this is where the hatch's locking latch handle is located.

TNPA-65A provides only 14 degrees of binocular vision horizontally and 6 degrees of vertical vision, meaning that its width is within the normal and acceptable range, but it is very narrow.

Together with the TKN-3M/MK, the two forward-facing TNPO-160 periscopes comprise the forward vision assembly of the commander. Despite the limited all-round visibility (compared to NATO tanks) offered by the commander's five periscopes, he can still compensate by simply rotating his cupola. This essentially negates the smaller number of observation devices, but it does not compensate for the periscopes being more constricted than the type found in typical NATO tanks. Nevertheless, while the commander may not have perfect immediate all-round awareness, he has a very reasonable degree of coverage, definitely enough to fight with.

The commander's hatch is of a forward-opening half-moon type, mounted on the rotating cupola. The hatch is quite small, and exiting through it in a hurry may be problematic if the commander is wearing winter clothing.

It is spring-loaded to assist the commander in opening the heavy hatch. A simple rotating handle locks the hatch when closed, preventing it from bouncing up and down when the tank is in motion, and a smaller handle at the bottom of the hatch serves to lock the hatch in place when it is opened, which is useful when the commander wishes to view the battlefield from outside the hatch, or when he needs to use the complementary cupola-mounted machine gun..

Because it opens forward, the thick hatch gives the commander full-body protection from machine gun fire whenever he wants to pop out for a tactical assessment with binoculars. To look over the hatch, all he needs to do is to stand on his seat.

The commander is shielded from machine gun and sniper fire by his hatch

In some modifications beginning in the mid-70's, the commander's cupola may also have peculiar shield installed forward of the hatch. All T-72s operated by the Russian ground forces today feature this shield.

The lower part is a simple hanging canvas sheet, which isn't intended to be part of the protection scheme. The upper part is just a face shield for the commander for if he were to sit outside on the turret while on road marches, probably to protect his face from bugs.

The shield made of very thin sheet steel with an equally thin polycarbonate or perspex window and is thus not bulletproof, splinter-proof or fragmentation-proof (though the commander's hatch is). Therefore, the protection afforded to the commander does not change. The only ballistic protection the commander gets still comes from his hatch, only now he has dust and bug protection. See the photos above and below.


Screenshot from video

The gunner is seated on the left side of the turret. He is provided with a single forward-opening hatch. Its most distinctive feature is the smaller circular port hole at its center, intended for snorkel installation. The hatch is spring loaded to hold it in place when open, and to give a little leeway for the gunner when opening it. It is locked with a simple rotating latch. There is a single TNPA-65 periscope embedded in the hatch, facing to the left. It is rather small and slit-like, but it provides the gunner with some precious limited sideways visibility. It provides only 14 degrees of binocular vision horizontally and 6 degrees of vertical vision.

The gunner is responsible for all of the weapons-related equipment, including the autoloader, stabilizer, cannon, coaxial machine gun, the sighting devices and their associated instruments. The gunner's station is dominated by the massive GPS (Gunner's Primary Sight) which completely fills the space between the gunner and the wall of the turret and tips the scales at 80kg. The gunner's station is the most cramped position in the T-72, and even more so if he is wearing winter clothing. However, it would be a mistake to consider the cramped nature of the gunner's station as a unique and defining feature of the T-72. As a whole, the T-72's turret does indeed have a much smaller volume than most tanks, but the space delegated to the gunner is very much on par with its contemporaries.

Looking again at this table from "Human Factors and Scientific Progress in Tank Building" by M.N. Tikhonov and I.D. Kudrin as provided by Peter Samsonov, we can see that the space afforded to the gunner is seriously tight, only 0.495 cubic meters. However, this is still an improvement over the T-55, which gave its gunner only 0.395 cubic meters of space.

In any case, internal space in this tank seems to be more psychological than physiological. Volume and comfort-wise, the gunner's station in a T-72 is quite adequate for a legacy tank, though still undoubtedly cramped. However, that is not to say that crampedness of the gunner's station is entirely negative. A snug fit ensures that the gunner will not be knocked around too much while the tank is in motion, which is undoubtedly a small benefit to targeting precision while driving on uneven ground. It isn't so much an issue while on long marches, because both turret occupants may simply sit on the turret roof instead. In this respect, the T-72 has a slight ergonomic advantage over many tanks in that the gunner has his own hatch and he can exit whenever he likes to sit on the roof, or to stand upright. In the event of an internal fire, the entire crew can bail out through their own hatches with no fuss. This is quite unlike tanks like the T-55, Leopard 1, Abrams, or indeed, any other manually-loaded tank except for a few oddball designs like the M60A2 as the gunner is usually not provided with his own hatch. On long marches, he might be forced to stay put in his decidedly cramped station for hours at a time. This is not the case for the gunner of a T-72.

Case in point: in Part 2 of his "Inside The Chieftain's Hatch" video review of the Centurion tank, Mr. Nicholas Moran from Wargaming noted that after just 20 minutes, it was beginning to get uncomfortable in the gunner's seat. If it began to get uncomfortable in his seat, the gunner of a T-72 can open his hatch and sit on the roof, or just stand on his seat and stretch. Additionally, in a typical manually loaded tank, if the commander were incapacitated or killed, the gunner would have to squeeze through the commander's body or shift it aside in order to bail out. Again, this is not a problem for the T-72.

Mr. Moran also noted that the gunner's station in the T-55 was very well laid out, but mentioned that legroom was somewhat limited unless the turret was pointing straight forward, in which case he could stretch his legs all the way into the driver's station. The T-72 fully preserves the reportedly excellent layout of the T-55, but is more spacious by 0.1 cubic meters and offers a more generous allotment of legroom no matter where the turret is pointed. This is not due to the lack of a turret basket, but to a combination of the larger turret ring diameter and separated seating of the commander and gunner.

Ventilation is provided by a DV-3 fan, like in the commander's station. It is more than enough in European climates where temperatures are usually around 20° C (68° F) or less, as it is a relatively powerful 5.2W fan, but in hot, desert regions averaging 30° C to 40° C is only useful for increasing air circulation to stave off stuffiness, and little else. Still, it is unquestionably better than tanks that do not provide any personal ventilation at all. Many tanks made during the 60's and 70's lacked this minor amenity.

For general visibility, the gunner is provided with a single forward-facing TNPO-165 periscope and another TNPA-65A periscope on his hatch, facing to the left. The TNPO-165 periscope has a large field of view. It is placed there for the gunner to check the orientation of the gun barrel, and to make sure that the gun barrel is elevated safely when the tank is entering a ditch. In daytime, the periscopes are also sources of light. Most importantly, having the forward-facing TNPO-165 periscope allows the gunner to effectively use the tank from a turret-down status. The TPD-2-49 or TPD-K1 primary sight is periscopic, of course, so the gunner will have no trouble looking over the crest of a hill or a berm when the tank is hiding in a turret-down position. When the commander spots a target, he can designate it using the TKN-3M/MK optic and the turret will slew over to point at it. The gunner will be able to see the target through his primary sight, lase it, and prepare to fire. Once the commander gives the order for the driver to move forward, the gunner can immediately open fire once the muzzle of the cannon clears the crest of the hill or berm. However, the primary sight has an 8x magnification, so the gunner cannot see if the muzzle has gone over the crest of an obstacle from that alone. This is where the TNPO-165 is particularly useful as it allows the gunner to independently verify the position of the cannon barrel without the commander's assistance.

The TNPA-65A periscope installed in the hatch seems less useful, especially since it only provides a narrow view of the outside world. One conceivable purpose of this periscope is to permit the gunner to see if there are any obstacles on the left side of the tank, allowing him to see if it is safe to traverse the turret to the left. This may be important when the tank is fighting in a forest or in urban areas. The commander would be responsible for checking the right side of the turret, of course.

In a more general sense, the two periscopes are also useful in that the gunner gains a sense of spatial awareness. This gives the gunner a wide view of the terrain ahead of the turret which is important as the primary sight has a fixed 8x magnification with no reduced magnification setting like the earlier telescopic TSh2-22 of the T-54 (3.5x or 7.0x) and TSh2B-41 of the T-62 (3.5x or 7.0x). Having some spatial awareness may also help to prevent motion sickness.

1A40-1 sighting complex and 1K13-49 night vision/auxiliary sight

As you can see in the photo above, the gunner is supplied with a duplicate of the commander's master control panel. Besides being able to initiate the fire control system, control the ventilation, turn on the lighting system, and much more, having the master control panel gives the gunner complete control over most of the electrical equipment in the tank, and also enables the gunner to set the fuse on a HE-Frag shell in lieu of the commander if necessary. This means that the T-72 can theoretically be operated with only a 2-man crew with a minimal loss in combat capabilities. This may be useful when a tank company or battalion is understaffed and there are no sufficiently qualified substitutes for the tank commander's position, however unlikely that may be. Of course, this capability is also largely thanks to the omission of a human loader.

Beginning in 1979, the T-72A began to receive the 902A "Tucha" smoke grenade discharge system. The gunner is in charge of using the system, but usually follows the instructions of the commander when doing so. Referring to the photo above, the control box for the system is located to his immediate left, at the gunner's elbow. Using this control box, he is able to deploy smoke grenades from one of three groups (the twelve smoke grenades are divided into three groups) individually, volley fire the grenades from each group, or fire the grenades from all three groups at once, either individually or by volleys.

Referring again to the photo above, the circular box seen between the TPD-K1 eyepiece and the gunner's handgrips is the AZ-172 autoloader control box. The autoloader is turned on from this box, and the ammunition type can be selected by the gunner.

The T-72B uses a different autoloader, sometimes referred to as the AZ-184, and the control box is also different as a result. With the introduction of a new ammunition type, the new control box has an additional option on the ammunition selector dial: missiles. The T-72B3 features a modification of the AZ-184 autoloader, and it can be seen in the updated control box. The new control box retains the same layout as the old one, but is digitized.

The autoloader control box can be used to set the autoloader to automatic operation or manual operation. Setting it to manual operation mode enables the crew to manually load the gun with partial assistance from individual components of the autoloader, such as the chain rammer, or to manually load the gun entirely by hand. If the crew intends to load the cannon manually, setting the autoloader to the manual mode is mandatory as it allows the sighting complex recognize the readiness of the cannon once a shell is loaded.

The gunner is also provided with an autoloader ammunition indicator. The indicator is rather crude, even for its time, as the indicator display is based on simple milliammeter technology. Due to the small size of the indicator pin, it may be difficult to easily see the indicated number in a high intensity situation.

The indicator does not have any selectors or dials on its own housing. Rather, it works in conjunction with the autoloader control box. When the gunner selects an ammunition type on the dial on the autoloader control box, the ammunition indicator automatically displays the ammunition reserve for that ammunition type currently stowed in the autoloader carousel. The number of empty slots in the autoloader carousel is determined by setting the ammunition selector dial to the "Load" position. The ammunition indicator only goes up to eleven, so if the number of rounds for any ammunition type exceeds eleven, the exact number of rounds can only be determined by finding out the number of rounds of the other ammunition types and the number of empty slots in the carousel. Needless to say, it was not a very good system.

Besides the autoloader controls, there is also a turret azimuth indicator, installed just next to the manual turret traverse flywheel. It is a part of the horizontal turret drive mechanism.

The indicator is akin to a clock, with an hour hand and a minute hand. The hour hand is mainly a tool of convenience as it shows the direction the turret is pointing to, but it is also an important tool for laying the gun for indirect fire. The minute hand is read with the hour hand to obtain a precise reading of the orientation of the turret for indirect fire purposes.


Because of the T-72's status as a "mobilization model", the more expensive parts were usually kept as affordable as possible. It was to be manned by conscripts with minimal training (though I emphasize that it was still much better and more thorough training than what many 3rd world country tank crews received), and T-72 crews received fewer opportunities to conduct firing exercises during peacetime than T-64 and T-80 crews. The sighting systems suffered the most from this practice. The T-72 never had a true ballistic computer and the fire control system required far more manual input than the best analogues of the time. Furthermore, T-72 units usually received new ammunition later than units equipped with the T-64 or T-80. This fact exacerbated the lack of sophisticated sighting devices, and this shortage of technology in an increasingly technological stage of the Cold War was not comforting.

T-72 Ural


The T-72 first entered service in 1973 equipped with the TPD-2-49 sighting complex with an integral optical coincidence rangefinder. This was the most modern sighting complex available in the USSR at the time having just entered service one year prior as one of the new elements of the T-64A obr. 1972. The sight is independently stabilized in the vertical plane. The internal gyroscope installed at the far end of the sight housing, in a protruding block underneath the sight aperture. The vertical stabilizer of the cannon is slaved to the sight. This improves the accuracy of the cannon in the vertical plane.

The sight assembly is mounted to the roof of the turret by a single large bolt. The load-bearing suspension elements of the sight are protected by shock-dampening bushings. The armoured nut holding the sight assembly to the turret roof is behind the armoured housing for the sight aperture, as you can see in the two photos below (from Facebook group). The armoured nut has a diameter of 110mm.

Further protection is provided by a 50mm layer of anti-radiation lining known as "Podboi" attached to the turret ceiling, between the sight assembly and the turret roof. This thick lining acts as a spall liner.

There are two eyepieces for this system. The viewfinder is split into two halves, top and bottom. The two measuring optics see the same target, but half of it is blocked out, and the gunner must use the adjustment dial near his hand grips to line up both halves and obtain a seamless picture. This process was relatively cumbersome and somewhat inaccurate - the error margin was 3% to 5%, which meant that the range could be off by up to ±200 meters at a distance of 4 kilometers, or a much less serious ±30 m at 1 kilometer range. Nevertheless, this is not a serious drawback when considering that the average tank engagement distance expected in Europe was estimated to be around 1500 m, not to mention that the use of hypersonic APFSDS ammunition meant that the error margin could usually at closer ranges be ignored since the ballistic trajectory was so flat. The problem was much more pronounced with HEAT and HE-Frag ammunition, which were heavier, had more drag and came out of the barrel at much lower velocities. With the advent of long range ATGM systems mounted on jeeps, scout cars, IFVs and even light tanks, accurate long-distance fire with HEAT and HE-Frag shells was imperative.

The sight has a fixed 8x magnification with a field of view of 9 degrees. The second measuring optic also has a fixed 8x magnification, but has a much smaller field of view of only 2 degrees. The left eyepiece on the sight shows the view through the primary optic and the right eyepiece is for the secondary optic. This is shown in the diagram below.

Because TPD-2-49 is independently stabilized in the vertical plane, it is possible to conduct rangefinding while the tank is in motion. Like the advanced TPS1 sighting complex of the T-10A and T-10B heavy tanks and unlike the simpler "Meteor" stabilizer for the T-62, the vertical stabilizer of the cannon is slaved to the vertical stabilizer of the TPD-2-49, meaning that the vertical motion of the cannon is directly slaved to the stabilizer for the sights. This allows the system to achieve a higher accuracy as the system will only fire when the elevation angle of the cannon is aligned perfectly with the sights, which has a more precise stabilizer than the cannon. The much larger range of elevation of the independently stabilized sight also allows the gunner to maintain constant visual contact with the target as the tank and even engage it while the tank moves across undulating terrain, as the system will automatically fire the cannon when it is in the proper elevation angle if the gunner holds down the trigger button. This feature also allows the gunner to see a target from a turret-down position behind the crest of a hill or a specially prepared ditch, which can be useful in certain situations. The gunner can search for a target, acquire a ballistic solution, lay the reticle on the target and proceed to press and hold the trigger button. When the commander gives the order for the tank to move forward, the gun will automatically fire when the stabilizer system detects that the cannon is aligned properly with the sights. This reduces the time spent in a non-turret-down position, thus shortening the period of vulnerability.

The gunner turns a range adjustment wheel located just above his hand grips to line up the two halves, as shown in the GIF below and in this short video (link).

The gunner must use both eyepieces, but the rangefinder is not stereoscopic in the same context as stereoscopic rangefinders. The gunner sees one half of the target from each of the eyepieces, but since the field of view from the second measuring optic is very narrow (2 degrees) compared to the view from the main sight (9 degrees), the gunner must find the target using his main sight and then place target near or at the center of the viewfinder of the main sight, or the target will not be in view of the second measuring optic. See the diagram below.

In case of low visibility from poor weather conditions or from enemy countermeasures, the rangefinder can be set to a secondary mode, where instead of splitting the target into two halves, two full images of the target are displayed on top of one another. There is a fixed vertical line at the left side of the viewfinder, and the gunner must lay the line on the edge of the target tank in the bottom image by using his handgrips, then turn the range adjustment wheel until the same edge of the target tank in the top image touches the line. In other words, if the same part of the tanks in both images touch the vertical line, then the two images are aligned. Refer to the diagrams below.

This method tends to be less precise, but may be easier to use if the outline of the tank is not clear or if the target is not a tank but something with an irregular shape. Both techniques can be used to determine the range to terrain features like trees or man-made structures like telephone poles and buildings, thus enabling the tank to accurately engage infantry-type targets hiding in wooded areas and urban areas alike. Most importantly, this allows the tank to open fire at anti-tank missile teams at long range with reasonable accuracy.

A major flaw with optical coincidence rangefinders in general is that they don't work very well on camouflaged targets or in inclement weather, especially without a very high magnification sight. Even tanks simply painted the same shade as the environment can be difficult to accurately range because the outlines of the tank may not be very clear to the gunner and this is made much worse during poor weather. As mentioned before, small ranging errors were more forgivable to a T-72 gunner because the gun fired APFSDS ammunition with an extremely high muzzle velocity, but firing HEAT on targets would be very difficult at longer ranges, not to mention moving ones.

While the TPD-2-49 would have qualified as among the world's best sighting systems when it was introduced with the T-64A in 1967, it was not quite as fresh by the time the T-72 Ural came onto the scene in 1974. As time went on, it became increasingly clear that optical rangefinders were no longer satisfactory, largely because it took a great deal of concentration from the gunner to operate, and in the case of the TPD-2-49, it was very expensive to manufacture an advanced independently stabilized sight with an integrated optical rangefinder. They were also fragile, despite extensive shockproofing and the use of anti-vibration bushings in the mounting points. Any misalignment as a result of shocks from tank shell impacts could cause some lens to be misaligned even slightly and that would be enough to put it out of commission for the duration of a battle, and this was a big problem with the T-72 (and indeed, every other tank with such a rangefinder) because an optical tube connecting the first aperture to the main sighting unit ran across the turret ceiling above the cannon breech block. A shell impacting the turret roof might bounce off and not penetrate the steeply sloped armour, but the strong shock of the impact and the shifting of the relatively soft and relatively thin cast steel roof could cause enough damage to the optical tube that it might no longer be usable. The thick rubber bushings on the optical tube can be seen in the photo below.

The location of the optical tube connecting the two apertures can be seen in the photos below (credit to

This, in addition to the issues mentioned above, meant that production of TPD-2-49 sighting complexes was summarily discontinued soon after the improved TPD-K1 sight was available and modernization programmes to refit T-72 Urals with TPD-K1 laser rangefinding sights began after that. The TPD-2-49 continued to be installed on mass produced tanks until around 1977 and Ural turrets cast with the armoured protrusion for the aperture port of the coincidence rangefinder optic continued to be produced along side it. T-72 Ural tanks modernized with the TPD-K1 did not have the armoured protrusion removed, but the aperture port was blocked off and permanently welded shut. Some modernized T-72 Ural tanks had the armoured protrusion cut off the turret roof, such as the example in the two photos below. The weld seams for the armour blocks to seal off the gap in the roof from the missing armoured protrusion are visible.

The housing for the primary optic of the TPD-2-49 sight is heavily armoured. The external thickness of the side of the armoured housing is 20mm as measured from surface to edge of top cover plate, shown in the photo on the right (from the Facebook group). The external thickness of the front of the armoured housing is 80mm as obtained from subtracting the length of the cover plate (190mm) from the length of the housing (270mm) as shown in the photo on the left.

However, these external measurements on the armoured housing do not represent the actual thickness of the housing because there are recesses where the armoured cover plates are bolted, as you can see in the two photos below (from T-64A). The thickness of the front of the armoured housing is the thickest part by far, appearing to have the full thickness of 80mm judging by the photo on the right.

Based on actual measurements and other known facts, the minimum thickness of armour on all sides is one inch and the front of the housing is around three inches thick. The extensive protection offered to the sight aperture indicates that it would be extremely difficult to damage with airbursting artillery shells or any other means. A similar armoured housing was used on the T-72A and T-72B.

The glass aperture is protected by a layer of SET-5L ballistic glass (19mm thick) to protect the internal mirrors from bullets and shell splinters. The ballistic glass panel comes with an integral heating system to prevent fogging, and it is provided with a small external wiper to remove any debris or mud that might obstruct the gunner's vision. There is also a sheet steel hood over the sight aperture that shelters it rain and snow, or even mud thrown up during rough cross-country driving. The hood is shown below. Tank crews carry an extra sight aperture in internal stowage for quick field replacement. To replace a damaged sight aperture, the bolts on top of the armoured housing must be removed.

TPD-2-49 placed the T-72 Ural on at least equal footing with the best NATO tanks at the time, including the Leopard 1. As the optical coincidence rangefinder was integrated into the sight and the whole package was independently stabilized (which no other system could boast of), the TPD-2-49 could be considered a rather advanced sighting complex of the time, on par with the fire control system of the Leopard 1 and superior to the setup on the M60A1, which had a separate primary sight and M17 rangefinder unit. The gunner of an M60A1 would have to conduct the rangefinding process before switching over to the primary sight to engage the target whereas the TPD-2-49 sight was adjusted concurrently with the rangefinder, and target acquisition time was slashed accordingly. The only flaw is that the commander of the T-72 is not able to take over the rangefinding procedure via a sight extension, like on the aforementioned Western tanks.

T-72 Ural-1, T-72A (Early)


The TPD-K1 is a sighting system which consisted of the sight itself in addition to the internal electronic ballistic calculator and the sight-stabilizer interface, including the gunner's controls which are integral to the sight as you can see in the photo above. It was first installed as standard tank equipment in 1978 on the T-72 obr. 1978 (one of the models colloquially known under the Ural-1 umbrella term) and was retrofitted on a large number of older Ural tanks as part of a modernization effort. The TPD-K1 was later carried over to the T-72A in 1979 and to the T-72B in 1985 in a modernized form. The TPD-K1 first appeared in a T-72 in early 1975 and ten tanks with the new sight rolled off the production line at the tail end of the same year.

Around the same time, the T-64B had already entered mass production in 1976 and featured a new and more advanced 1A33 fire control system along with a guided missile launching capability. The 1A33 paired with the "Kobra" missile system was tested in a single T-72 prototype in 1976-1977 but did not enter mass production. It is interesting to note that older models of the T-64A were later upgraded with the TPD-K1 only in 1981. The Object 172M-1-E3 export variant of the T-72 is equivalent to the Ural-1 model. This model is known simply as the T-72M and features a TPD-K1 paired with the cast monolithic steel turret of the original Ural. It was used in various Warsaw Pact nations and in East Germany, as seen in the photo below.

The TPD-K1 sight is a modification of the TPD-2-49 and shares many components. The sight has a fixed 8x magnification and a 9° field of view, like the TPD-2-49. TPD-K1 gave the T-72 a 3-year head start over its American nemesis the M60A1, which received its own AN/VVG-2 laser rangefinder unit in 1978 as part of the M60A3 upgrade. German Leopard 1 tanks did not receive their own laser rangefinders until the 80's rolled around, and British Chieftains had to wait until 1988 to get theirs. The ballistic computer in the sight is able to automatically account for ambient temperature, ammunition charge temperature, atmospheric pressure, and barrel wear.

The TPD-K1 is independently stabilized in the vertical plane and it has an internal gyroscope installed in the same location as the one in the TPD-2-49. The range of independent vertical elevation is from -15 to +25 degrees, and the vertical cannon stabilizer of the T-72 is slaved to the stabilizer of the TPD-K1 sight in the same way as discussed previously with the TPD-2-49. However, the sight is not stabilized in the horizontal plane. This has implications that we will explore later. The stabilizer system for the sight is connected to the cannon for referencing purposes - the mechanical rods that connect the sight (left side) to the cannon (right side) can be seen below. It is a parallelogram-type system. This allows the sighting complex to determine the orientation of the cannon relative to itself so that the fire control system does not fire unless the gun is perfectly aligned with the sighting complex. This greatly reduces the effect of the larger error margin of the gun stabilizer compared to the independent stabilization system in the sighting complex.

The armoured sight aperture housing on the turret roof is very similar to the one used on the TPD-2-49, but instead of a simple sheet steel hood, additional ballistic protected is afforded to the sight aperture by the presence of two armoured "ears". The armoured "ears" extend quite a long distance from the aperture itself and fulfill the same purpose as the armoured doors of modern sight housing designs. The TPD-2-49 lacks such doors, so the aperture is only protected by a layer of ballistic glass without the option to seal it with an armoured shield like the night vision sights on the tank. A sheet steel hood is bolted to the armoured "ears" to provide protection from the elements in the same way that the hood for the TPD-2-49 did. Beginning in 1984, the T-72A began to receive the "Nadboi" anti-radiation cladding on many of the external surfaces of the tank. The armoured housing of the sight also received a layer of "Nadboi" as did the sheet steel hood, as you can see in the photo below.

The thickness of the two armoured "ears" is 11mm and the thickness of the sheet steel hood is around 4mm, as shown in the photo below kindly supplied by Jarosław Wolski (measurement was done on a T-72M1). The same shroud from the TPD-2-49 housing to shelter the sight aperture from the elements. As you can see in the photo above (showing a T-72B), the armoured "ears" may help limit the damage done to the sight aperture if one of the Kontakt-1 explosive reactive armour boxes next to the sight housing are detonated. It is unclear if the hood is made from armour-grade steel or just simple mild steel.

The TPD-K1 alone is shown in the two photos below. Note the two distinct polarized halves of the sight aperture. The laser rangefinder emitter and receiver is installed in the left half and the gunner's optical viewing channel is on the right half. Note that the TPD-K1 in the two photos below is being suspended from its mounting point. Like the TPD-2-49, the sight is mounted to the roof and does not contact the frontal armour of the turret. As such, it is almost completely isolated from shock damage from heavy impacts to the frontal armour.


The TPD-K1 comes with a neodymium doped glass infrared laser rangefinder. The sight is slightly unusual in that the laser rangefinder is installed inside the sight itself on the right hand side of the housing, but the rangefinder computer is installed outside the sight. This is apparently due to the fact that the TPD-K1 is essentially a modified TPD-2-49 sight, so the laser computer is practically an add-on module.

The two photos below show the detached rangefinder processing and readout unit.

The photo below show the rangefinder computer attached to the right side of the TPD-K1 sight module. There is a toggle switch on the computer that prevents the system from accepting ranges of less than 1,200 meters or 1,800 meters, and placing the toggle switch in the central position lets the system run normally. This is a system that filters out range readings that are below either 1,200 m or 1,800 m due to clutter or obstructions such as grass when ranging distant targets. The target can be determined to be further than 1,200 m or 1,800 m by estimation using the commander's stadia rangefinder or by simply using intuition.

According to the Indian Ordnance Factories website, the laser rangefinder uses an IR laser in the 1,060 nm wavelength. The rangefinder has an automatic range compensation mechanism for firing on the move known as a Delta-D system, whereby the rangefinder computer will automatically subtract the distance covered by the tank from the measured range. This system works by measuring engine revs to compute the distance traveled forward by the tank. The laser rangefinder has a maximum error of 10 m at distances of 500 m to 3,000 m. From 3,000 m to 4,000 m, the maximum error threshold increases to 15 m. The rangefinder may become unresponsive and highly inaccurate past 3000 meters, so it could be necessary for the gunner to manually dial in the range to the target by other methods. This limitation makes it infeasible to engage point targets at distances beyond 3000 m, but firing HE-Frag shells at targets further than 3,000 m is not a serious issue.

The two photos below show a TPD-K1 with the mechanical linkages that connect the sight to the cannon. The photo on the right is partly dismantled, exposing the internal PCBs in the laser rangefinder computer. The large mounting screw on the top of the sight housing is clearly visible in both photos, but the armoured cap is missing on the example in the photo on the right. The shock-damping bushing on the screw is visible in both photos.

The rangefinder computer has a digital display to show the measured distance, but range information is ported through to the range indicator dial on the top of the gunner's viewfinder, which the gunner can read. However, reading the range is generally not necessary since the fire control system will automatically calculate a ballistic solution. Knowing the range to the target is only necessary when using the coaxial machine gun, as the sighting complex does not automatically calculate a ballistic solution for it. To lase a target, the gunner must place the illuminated red circle over it and fire off the laser for 1 to 3 seconds, after which the firing solution is generated and the gunner can immediately lay the reticle on the target. If the target is mobile, it must be tracked within the boundaries of the red circle until the range is obtained. The rangefinder unit must take 6 seconds to cool down between uses.

BVD-2 Range input unit

Range information is automatically routed to the sighting unit, and the sight makes the necessary corrections and adjusts the reticle accordingly. The illustrations below shows what happens during the ranging process.

Firstly, notice the circle at or near the center of the viewfinder. That is where the target must go in order to initiate the rangefinding process. Once that is done, the reticle instantly lowers to account for ballistic drop, and the range indicator dial at the top spins to give a visual reference for the distance (with an accuracy of within 10 m). The lasing circle remains static for lasing the next victim. However, due to the lack of independent horizontal stabilization in the sight, the viewfinder is not adjusted horizontally to account for the parallax error created by the offset between the axis of the sight and the axis of the gun barrel. Because of this, the gun and sight can only be zeroed at a fixed distance by the gunner under non-combat conditions with no option to adjust it during combat.

The ranging procedure is completely normal in the realm of tank fire control systems, but one oversight is that the path for the laser beam is not merged into the same lenses for the main optic. Rather, the laser rangefinder has its own optical path parallel to the lens tube which the gunner uses. This is evident when you closely inspect the sight aperture:

As you can see, the mirror is divided into two halves by an opaque block in the middle. Underneath the mirror, you can see two apertures. One for the laser rangefinder, and one which the gunner sees out of. This means that the rangefinder circle is never directly on top of the reticle. The gunner must lay the rangefinder circle over the target, lase it, and then finish by laying the reticle on the target. There are a multitude of disadvantages to this. Instead of laying a reticle on the target once and letting the fire control system handle it, the gunner must conduct the laying process twice. This creates room for operator error and consumes precious time.

Without an optical coincidence rangefinder system installed, the optic tube that ran across the ceiling over the cannon breech block in the T-72 Ural is no longer present.

The TPD-K1 has a stadia-reticle rangefinder with markings for distances of 500 m to 4,000 m that can be used to gauge target distance if the laser rangefinder is malfunctioning. This and the manual gun laying drives allow the gunner to continue engaging targets even if all aiming systems have completely lost power. The sight will raise and depress along with the cannon when the stabilizer is off because the sight is linked to the vertical manual drive for cannon elevation via mechanical linkages.

All reticle lines can be illuminated (red colour) by an internal light bulb for better discernability in cloudy weather or at night.

The viewfinder of the sight includes graduations for firing the PKT machine gun to a maximum range of 1,800 m, for firing HE-Frag shells to a maximum direct fire range of 5,000 m, markings on either side of the center chevron for manually applying lead on moving targets, and an auxiliary stadia rangefinder for manually determining the distance to a tank-type target 2.7 m in height at distances from 500 m to 4,000 m. The stadia rangefinder is for emergency use only, as the laser rangefinder is far more convenient and accurate. On the top of the viewfinder is the range indicator dial which displays the measured range to the target. The dial is limited to a maximum range of 4,000 m. Once the gunner has lased the target, the range will be displayed here for the gunner's reference. A red light is illuminated above the dial when the fire control system is ready to fire.

The ammunition type is automatically inputted into sighting system via the autoloader ammunition selection dial. The silver coloured dial can be seen in the photo above, to the bottom left of the eyepiece of the TPD-K1. When the ammunition type is set, the autoloader begins loading the desired type and the gunner can proceed to lase the target during the loading cycle. Once the gunner has lased the target, the reticle in the sight automatically drops. All the gunner must do now is to raise the center chevron of the reticle onto the target and fire. Subsequent shots do not require the process to be repeated, even if the gunner changes ammunition types. All he must do is select a new ammunition type, and the sight will automatically adjust to the proper superelevation using the range information from the previous lasing. However, the system does not calculate a ballistic solution for the coaxial machine gun and the only provision for aiming it is the markings in the reticle. The gunner must use the range figure from a previous lasing and manually raise the reticle so that the target is aligned with the appropriate distance mark in order to open fire accurately. In practice, however, the gunner relies heavily on following the stream of tracers for fire correction.

As mentioned before, target leading is done manually by using the markings on either side of the center chevron. The gunner must estimate the lateral speed of the target by determining how long it takes for the target to move from the center chevron across the lead markings, and combine that information with the time of flight of the selected ammunition type to the measured range. Needless to say, this was not easy unless the gunner was highly experienced, and even so, the accuracy of manual lead estimation is invariably lower than an automatically computed lead solution. However, the significance of this deficiency against moving targets is counteracted by the immense speed (~1,800 m/s) of the APFSDS rounds fired by the T-72. At short ranges (<1 km), practically no lead is required to hit a moving tank-type target, and at medium ranges (1-2 km), there is a sufficiently large margin of error that a hit is still highly likely. Problems from the lack of an automatic lead computing system only arise when attempting to engage moving targets with HEAT rounds, which travel at half the speed of APFSDS rounds. HE-Frag rounds travel at a relatively low velocity as well, but as the main intended targets for HE-Frag ammunition tend to be static, this was not an issue.

The inputted ammunition type is indicated by one of three coloured signal lamps at the top left corner of the sight. If the fire control system is being operated manually or in a degraded mode, the sight can be set to the desired ammunition type by turning the dial next to the signal lamps. Otherwise, the ammunition type is automatically inputted.

T-72A, T-72B

1A40, 1A40-1, 1A40-1M Sighting Complex, TPD-K1M

The TPD-K1 evolved to become the 1A40 sighting complex, featuring some small improvements related to the fire control system, the most notable of which is the new manual lead calculator. According to Mikhail Baryatinsky, the T-72A received the 1A40 beginning in 1982, and the 1A40-1 came standard on the T-72B since its formal introduction in 1985. The 1A40-1 sighting unit features a slightly improved TPD-K1M primary sight and is distinguished from the 1A40 system by the improved laser rangefinder. The 1A40-1M was used in the T-72BA and differs from the basic variant by the possibility of manually entering corrections using data collected by an external wind sensor and by a cant sensor. Just as the 1A33 sighting complex includes the "Kobra" missile guidance system as an integral component, the 1A40 sighting complex includes the TPN-3-49 night vision sight and the 1A40-1 includes the 9K120 "Svir" missile guidance system.

The 1A40 sighting complex includes an additional eyepiece for the gunner's left eye for the UVBU lead calculation system. The new UVBU system is an add-on device that calculates the necessary amount of lead for a moving target and displays it in figures which can be manually applied by the gunner on the lateral scale in the TPD-K1M. It works by determining the rate of rotation of the turret as the gunner is lasing the target and then translates that information into mils, which is displayed in the eyepiece for the gunner to read. The gunner will then know which marking on the lateral mil scale on the reticle he should adopt as the new aiming point. The system will display the mil figure as a positive or negative integer to denote which the side that the gunner must use (negative for the scale on the left of the center chevron and vice versa). The use of an eyepiece rather than a separate digital display is so that the gunner does not need to break visual contact with the target. As the UVBU eyepiece displays a virtual number on a black background, the gunner can keep both eyes open to see the number floating in his vision, read the mil figure, and then apply it, all done without removing his eyes from the TPD-K1M eyepiece.

The precision of the UVBU unit is not high compared to the systems employed in more advanced fire control systems, as it can only calculate a difference in the angular velocity of the target compared to the tank down to ± 0.5 mils. This is more than enough even for medium range shooting, as a target tank travelling laterally across the sight would be presenting the silhouette of its profile which is a wider target than a head-on silhouette, but the system is insufficient for long range shooting. Other than that, its most serious drawback is the lack of automation. In the fire control system of an M60A3, for instance, lead for a moving target is calculated and automatically applied to the reticle by the ballistic computer after the target is lased, meaning that the sight automatically adjusts horizontally (via independent horizontal stabilization) so that the reticle has already compensated for lead. This allows the gunner of an M60A3 to press the trigger immediately after lasing as long as the reticle continues to track the target - no need to use secondary markings to engage. This is much faster than the system employed on the TPD-K1M. This is an inherent flaw in the 1A40 sighting complex as it cannot automatically adjust the reticle for lead, since it lacks independent horizontal stabilization. Overall, the system is somewhat crude, and could be considered technologically obsolete at the time that it was introduced. By the time the T-72B entered mass production in 1985, the 1A40-1 sighting complex family could be considered outdated. The 1A40-1 is shown in the two photos below.


The TPD-K1M sight itself differs from the TPD-K1 by the presence of mil values printed on the secondary chevrons for the UVBU leading system. Otherwise, the viewfinder is identical.

The photo below (credit to: ru-armor.livejournal) shows the markings more closely.

It is possible to use different shell models by simply twisting a dial on the UVP control unit, pictured below. Eleven distinct types can be selected from this unit. The UVP control unit was first used on the T-72B.

The UVP unit allows the gunner to instantly reset the sights for different types of each category of ammunition. It is also possible for the T-72 to use "exotic" ammunition this way. For example, one of the blank spaces on the indicator card for HE-Frag (labelled OF in the photo above) can be filled for flechette rounds. The gunner can then toggle the sight for the HE-Frag ammunition type, and then cycle the HE-Frag dial on the UVP panel to the flechette slot.

Unlike the handgrips for the fire control systems of previous Soviet tanks, the handgrips are permanently attached to the TPD-K1 sight. The handgrips have a protruding ledge at the base for the gunner's hands to rest on, and each handgrip has two buttons each. The left trigger button is for firing the coaxial machine gun and the left thumb button is resetting the laser rangefinder. The right trigger button is for firing the main cannon, and the right thumb button is for firing off the laser rangefinder. An ex tanker has remarked to the author that he found it difficult to operate the handgrips sometimes as it was rather confusing for him. He had previously operated a T-55 tank, and the thumb buttons were for firing the cannon and co-ax. With the T-72, the trigger buttons had not only moved, two more buttons were added!The price of progress is high indeed.



The TPN-1-49-23 is the gunner's night sight in the T-72 Ural and its variants, as well as almost all exported T-72 variants with the exception of the T-72S. The TPN-1-49-23 can be used in the passive image intensification mode or in the active infrared imaging mode, whereby infrared light emitted from the L-2AG "Luna-2" IR spotlight is used to illuminate the target. The "Luna-2" spotlight is mounted coaxially to the main gun so that all three devices - sights, cannon and spotlight - are adequately aligned. Like the commander's OU-3GA spotlight, the "Luna-2" spotlight is a simple tungsten filament lamp with a simple IR filter fitted in front of the bulb. Removing the filter transforms the IR spotlight into a regular white light spotlight. For passive observation, the sight is equipped with a 1st generation image intensifier tube. The image resolution is low, and the amplification factor is not high, so reliably seeing and identifying targets is very difficult in low light conditions. The inherent shortcomings of the 1st Generation system can be partially overcome with artillery-launched illumination rounds, but the low image resolution remains an issue.

Like the main sight, the TPN-1-49-23 is protected by a square-shaped armoured housing, with a bolt-on steel cover for the aperture. Beside the aperture is a single FG-125 infrared light, which is used only as a driving light and not for the TPN-1-49-23.

The sight has a maximum viewing distance of 500-800 meters in the active mode using the "Luna-2" IR spotlight. The passive image intensifier mode allows the same type of target to be spotted at ranges of up to 500 m if the ambient light is no less than 0.005 lux, which is the typical brightness of a moonless, starlit night with clear skies. Clarity and spotting distance improves with increasing brightness. Based on research into other 1st Generation night vision systems, the identification distance may be expanded to around 1000 m on moonlit nights, and it should be possible to spot tanks at distances of more than 1300 m during dark twilight hours, although low magnification and mediocre resolution complicates target observation beyond that range. The sensitivity of the image intensifier in the sight should be decreased to provide better image quality during twilight hours.

Soviet enthusiasm for image intensification technology gave the T-64 and T-72 a significant night fighting advantage over their Western counterparts, which relied solely on IR illumination technology for the entirety of the 60's and for most of the 70's. Case in point: The M60 received an image intensifier sight - a so-called "starlight scope" - only in 1977 with the M60A1 Passive modernization, and the original 1978 production M60A3 had the same passive nightsight. However, these new passive optics offered better performance than their contemporary Soviet counterparts, and the large investments in thermal imaging technology in the U.S allowed them to leapfrog over the USSR in night fighting technology by the late 70's. This is best exemplified by the use of AN/VSG-2 thermal imaging sights in the M60A3 (TTS) in 1979. The T-72A was introduced in the same year, but it still had the same TPN-1-49-23 night sight, and only upgraded to the TPN3-49 after some time into its production run, possibly at the turn of the decade.

Due to the short range of vision, the markings in the viewfinder of the TPN-1-49-23 are greatly simplified. All four ammunition types (including the coaxial machine gun) were represented on the markings, albeit not with complete exactness, but it was acceptable due to the short aiming distances expected. It's worth noting that the flat trajectory of subcaliber rounds like 3BM9 and 3BM15 made it infeasible to include them in the same set of markings as the other ammunition types, but their high muzzle velocity meant that it was fairly easy to hit a tank-sized target at short range anyway.

The end point of each line is an indicator for a certain distance for certain types of ammunition. A breakdown of the markings can be seen in the diagram below. This diagram is printed on an instruction panel on the sight itself, in case the gunner forgets the meaning of the markings in the heat of battle. As you can see in the diagram, the topmost point represents 100 meters for HE-Frag rounds, 200 meters for HEAT rounds and the coaxial machine gun, and 1,100 meters for subcaliber rounds. To fire subcaliber rounds at targets below 1,100 meters, the gunner would have to aim above the topmost point.

It has been frequently noted in various websites that the T-72 can be distinguished from the T-64 by the change in position of the spotlight from the left side of the cannon to the right side. This change was not arbitrary; the spotlight was placed next to the coaxial machine gun port so that the driver was physically blocked from sticking his head out of the hatch and in front of the machine gun when driving the tank during road marches, and so that the driver does not have to be in front of the machine gun when entering and exiting his station. Although accidental discharge was unlikely, there were some tragic cases, so the relocation of the spotlight was a serious safety consideration.

The sight cannot be used in daytime, because sunlight will overload the sight unit and possibly damage it. In accordance with this rule, the aperture has internal shutters linked to the firing trigger of the gunner's control handle. Upon firing, the shutters automatically close to shield the unit from the intense flash of cannon fire at night.


Most T-72B1 tanks and some late model T-72As have the TPN3-49 night sight installed. The TPN3-49 exists as a substitute for the 1K13-49 in the case of the T-72B1, and it exists as an upgrade over the TPN-1-49-23 for the T-72A. The photo above, taken from "T-72: Yesterday, Today, Tomorrow" by Sergey Suvorov, shows the gunner's station in a mock T-72A (imitator for training) with the TPN3-49 installed. TPN3-49 is more advanced than its predecessor in many ways, but still retains many of the key features and drawbacks of its predecessor, including a lack of independent stabilization. Like before, the vertical stabilization system of the TPN3-49 is based on mechanical linkages connecting the mirror head in the sight to the stabilizer of the TPD-K1, thus allowing the gunner to maintain a fully stabilized field of view in combat and throughout the loading process of the cannon. One of the improvements of the TPN3-49 over the TPN-1-49-23 is the ability to switch between different viewfinder markings for different ammunition types, giving the gunner better aiming precision.

Instead of a single universal set of markings with predetermined aiming points, the TPN3-49 features comprehensive range scales and a range dial for each type of ammunition.

The new night sight features a more sensitive electron-optical converter tube and amplifier system, giving the gunner a brighter and clearer image when using the sight in the active night vision mode. Another improvement came from the installation of the newer L-4A "Luna-4" spotlight, which is more powerful than the older L-2AG "Luna-2" spotlight. The image intensifier in the TPN3-49 is improved, but still belongs to the 1st Generation, so the maximum viewing distance is only around 500-800 meters at an ambient light level of no less than 0.005 lux. This is rather poor compared to the "starlight scope" in the M60A1 and M60A3, which used a 2nd Generation image intensifier and offered a viewing distance of 1,300 meters.

The L-4A spotlight can be distinguished from the L-2AG by the location of the power supply cable socket. The socket on the L-2AG was located on the right side of the spotlight, but the socket on the L-4A is located at the back, as you can see in the photo on the right. The photo on the left shows an L-2AG.

Even without delving into thermal imaging technology, the night fighting capabilities of the T-72A were still rather limited. In terms of active infrared imaging, the T-72 was behind both the M60A1 and the Chieftain. The M60A1 used the advanced AN/VSS-1 spotlight, featuring a motorized lens and occluder that enabled the gunner to remotely adjust beam width from 0.5-0.75 degrees in the narrow mode to 7 degrees in the wide mode as well as select between white light and infrared light on the fly. The AN/VSS-1 ran on 1 kW and had an output of 100 million candelas, but there was an overcharge mode that could bring the output up to 150 million candelas for a short time. Additionally, the M32 IR night sight installed in the M60A1 had an 8x magnification, giving the gunner better long range visibility compared to any T-72. The Chieftain was in a similar position of advantage, as it had an immensely powerful 2 kW spotlight with a large 570mm aperture, which greatly benefited the gunner in searching and engaging targets at longer ranges. The L-4A "Luna-4" spotlight is underwhelming in comparison, seeing as the spotlight ran on just 600 W and had an output of only 30 million candelas - between three to five times less than the AN/VSS-1. Additionally, the beam width from "Luna" was fixed at around 1 degree horizontally and 0.8 degrees vertically, making it exceptionally difficult to search for targets across open terrain. Furthermore, the lack of an occluder, otherwise known as a blackout shield, in front of the xenon arc lamp in "Luna-4" meant that only a part of the light was directed from the concave reflector. The rest of the light was emitted in a forward arc, illuminating the tank itself as well as the ground in front of it, making the T-72 an extremely prominent target once the spotlight was activated.

Despite these drawbacks, the TPN3-49 apparently allows a T-72A gunner to spot a target at a maximum range of 1,300 meters in the active infrared imaging mode. This is surprising when we consider the fact that the reported viewing distance for the gunner of a Chieftain Mark. 3 is only 1,000 meters, although that might be due to the relatively low 3x magnification of the Chieftain's No.33 IR night sight.


The 1K13-49 sight was created as the guidance control unit for new 9K120 "Svir" laser-guided 125mm gun-launched missiles but continued fulfilling the function of a night vision sight, having improved night vision capabilities using technologies derived from the TPN3-49. The electronic modulator and signal generator for the missile guidance system is in separate boxes located on other parts of the tank, but the laser emitter is installed inside the 1K13-49 sight itself. The system has an operating range from 100 m to 4,000 m. The 1K13-49 has the same functions as the 1G46 sighting complex and shares the same missile guidance technologies, but the 1K13-49 sight has lower magnification and has a reduced maximum range compared to the 5,000-meter range of the 1G46 which uses the slightly better "Refleks" missile guidance system. The first tanks that were assembled with the "Svir" missile system appeared in 1984 but the system only entered mass production in 1985 as an integral part of the T-72B tank which also entered mass production in the same year. The location of the "Svir" system components are shown in the drawing below.

The sight has a daytime channel that is normally used in conjunction with guided missiles, but having a daytime channel allows the 1K13-49 to be be used as a backup sight in case the TPD-K1 is non-functional. With a fixed 8x magnification in the daytime channel, the 1K13-49 can be an adequate replacement for the TPD-K1, but lacks a laser rangefinder and proper range scales for different ammunition types. The gunner is forced to make the most of the simplified markings provided in the viewfinder. The photo below shows the viewfinder for a variant of the 1K13 used in the BMP-3, so it may not be entirely accurate.

Its active infrared optoelectronic imaging system is also improved over the TPN-1-49-23. Now, the viewing range in the active mode is increased to 1,200 m. However, the image intensification system has not been improved, meaning that the 1K13-49 sight still only has an 800 m viewing distance under ambient lighting conditions of no less than 0.005 lux. The optical magnification factor remains at 5.5x, like the previous sighting systems.

The diagrams below show the markings in the viewfinder for the night vision mode. The same viewfinder is used for both the active and passive modes of operation. Like in previous night sights, the chevron and range indicator lines are adjusted up and down while the range scale remains static. To adjust for longer distances, the range indicator line is adjusted down until it lines up with the desired range, and the chevron will also drop down by the same amount. By laying the chevron onto the target, the gun is elevated by the necessary superelevation and the gunner can open fire.

The sight has a field of view of 5 degrees in the daylight setting or 6°4' in the nighttime setting. It is independently stabilized in the vertical plane, with +20° elevation -7° depression.

As usual, the sight aperture has two protective housings; one enclosing the sensitive optical workings of the aperture itself with a tempered glass window and a shock-proof shell, and another very heavy duty steel carapace covering that, along with a thick steel window shield.

Externally, the key differences between the 1K13-49 and the other night sights lie in its distinctly larger armoured housing, complete with a remotely opened armoured shield.

1A40-4 Sighting Complex, SOSNA-U

SOSNA-U is a multi-channel thermal imaging sighting complex with capabilities matching those of its contemporaries, giving the T-72 a much needed boost in target acquisition and engagement capabilities. SOSNA-U uses the French-designed 2nd generation Catherine-FC thermal imager. The SOSNA-U sighting complex features an internal ballistic computer that enables it to automatically detect targets, track them, and calculate a ballistic solution including lead using the data from its internal rangefinder, its image processing software and its internal gyroscope (to calculate cant). As you would expect, the sight is stabilized in two planes. The sight has a very limited 3x optical magnification with an equally disappointing 6x maximum digital magnification. Contemporary thermal imaging sights are typically capable of very high digital zoom with double digit magnification factors.

The sighting unit can be seen in the photos below.

The view through the eyepiece in the optical day channel can be seen below.

SOSNA-U can reportedly be used to identify and engage tank-type targets at a nominal distance of 5 km in daytime in the normal optical channel, and up to 3.5 km in either day or night through the thermal imaging channel, but this is extremely optimistic. For one, there is hardly any location that is flat and featureless enough that tanks can be spotted at such a distance, assuming that the weather is clear enough that tank-sized targets can be distinguished from the terrain. The 3x optical zoom of the sight is simply insufficient for anything more demanding than general observation. It is just not possible to spot and identify a tank-type target at 5 km through the daytime optical channel. Secondly, the limited 6x digital zoom of SOSNA-U makes it very difficult for the gunner to identify even a halfheartedly camouflaged tank at the claimed 3.5 km distance. When looking through the thermal imaging channel, any vehicle will appear more as a white blob on the screen at such long distances.

Like the 1K13-49 sight it replaces, SOSNA-U has a missile guidance unit that allows it to be used to guide existing gun-launched missiles as well as newly developed missiles. The automatic target tracking feature of the sighting complex would be quite beneficial when engaging moving targets with guided missiles.

The gunner has two means of looking through SOSNA-U - the eyepiece, which is for the right eye and comes with a very comfortable forehead pad, and the 640x480px (5.7 inch) flatscreen display.

In addition to the sight itself, the T-72B3 upgrade also comes with a new digital ballistic computer of unknown make, as seen below. The sight itself cannot accept data from peripherals such as anemometers, thermometers, muzzle reference sensors, and so on, so in order to make use of such data, a ballistic computer is necessary. The addition of a digital ballistic computer elevates the fire control system of the T-72B3 up to a level on par with, and quite possibly exceeding the T-90A. The addition of the flatscreen display and the digital ballistic computer eliminates the possibility of stowing ammunition in the turret on the wall behind the gunner, as the gunner's master control panel is now moved to a spot behind his left shoulder, and the ballistic computer housing occupies quite a lot of space behind his seat.

The SOSNA-U is considered the de facto main sight for a T-72B3 gunners, relegating the TPD-K1 to the back-up role instead. The UVBU lead calculator device installed parallel to the TPD-K1M sight has been removed as it is now totally obsolete, thus retrograding the 1A40-1 sighting complex into the 1A40. Unfortunately, the designers apparently didn't see it fit to swap the placement of these two sighting units, resulting in less than optimal placement of the SOSNA-U, which is only somewhat negated by the use of a separate flatscreen display. Another rather strange quirk is that the sight aperture window cover has to be manually opened by unbolting it, which seems to be a step backwards from the 1K13-49.

Also note the IR lamp mounted next to the sight housing. As SOSNA-U is a thermal imaging sight, this lamp is totally unrelated to its operation. To the contrary, this lamp is used to replace the normal driving headlights if they are submerged under water or plastered with mud, which could happen if the tank is fording a stream or driving through a swamp. This lamp is turned on and off by the commander.


Stabilizer precision and sensitivity is a crucial factor in overall engagement capabilities, especially when on the move. In a continuation of the endearing Russian tradition of naming military hardware after innocent, peaceful things, the stabilizers are named after flowers. The hydraulic pump and power supply system are located in the hull and the hydraulic turret rotation drive is at the turret ring in front of the gunner, behind the sights.

Turning on the stabilizer is done with the central toggle switch located just above the handgrips on the TPD-K1 sight. A well trained gunner would know not to keep the stabilizers on for too long, as it will overheat and wear out quickly.

2E28M "Sireneviy" (Lilac) Electric/Hydroelectric Stabilizer

The 2E28M dual-axis stabilizer is used in the T-72 Ural. The 2E28M stabilizer has the factory designation of VK1.370.058. The precision offered by this stabilizer is technically quite high for the era, but holistically the weapons system is still too imprecise to guarantee hits on the move at long ranges. Nevertheless, the precision and overall operating characteristics of "Sireneviy" is higher than previous tank stabilizers and offers an improvement in fire-on-the-move capability. It enables the tank to engage tank-type targets at average European combat distances or 1.5 km in a static position and when traveling at low speed with a reasonable degree of accuracy.

The stabilizer has two modes of operation: automatic and semi-automatic. The automatic mode is the primary mode for combat purposes; the stabilizer is at full operational capacity and will keep the gun aimed with maximum precision at the target when the tank is in motion. This mode is the default mode during combat, and is used when firing from all positions - stationary, while moving, and during short halts. The semi-automatic mode is an auxiliary operating mode as well as an emergency mode in the event of stabilizer failure. The stabilizer is set to this mode when combat is imminent in order to prevent overheating and to maximize the operating life of the stabilizer. In this mode, the vertical stabilizer system is disabled so that gun elevation reverts to a manual status, but the powered turret traverse and stabilization system remains operating, albeit at a reduced capacity; gun laying precision is greatly reduced, but the turret rotation speed is slightly increased. The semi-automatic mode is suitable when when the tank is stationary, or for firing during short halts. The slightly increased maximum turret traverse rate in the semi-automatic mode is ideal for tracking moving targets when the tank is in a stationary defensive position. In such conditions, the lack of powered gun elevation is not an issue.

The stabilizer is turned off entirely when combat is not expected, and the time needed to get the stabilizer to operational condition is 2 minutes. The stabilizer can remain activated for a maximum period of four hours, regardless of whether the tank is in combat or not.

The screenshot below (screenshot taken from this video) shows the location of some of the stabilizer components at the top left corner.

One of the components visible in the screenshot above is the gyroscope unit for the stabilizer.

Using this stabilizer, the turret is somewhat slow to turn at only 18° per second. It would take it a minimum of 20 seconds to do a complete 360° revolution. This has the effect of inhibiting the T-72's ability to react to the unexpected emergence of a dangerous target from different directions at close range. In the semi-automatic mode, the rate of rotation is increased to 20° per second, but this less irrelevant in combat because the semi-automatic mode is only used during road marches when combat is imminent, and not actually during combat. The traverse speed does not change from the minimum to the maximum rates smoothly. The turret traverse rate can be changed smoothly in the range of between 0.07° per second and 6° per second, but the transition to the maximum traverse rate of 18° or 20° per second occurs abruptly when the gunner turns his control handles to the maximum extent.

As usual, the stabilizer system revolves around the use of a pair of gyrostabilizers to measuring angular velocities in order to enforce corrections. Turret traverse is done electrically while gun elevation is accomplished using a hydraulic actuator. The hydraulic pump for powering the cannon elevation system is located under the cannon's breechblock, and the hydraulic motor for turret traverse is installed in front of the gunner, behind his TPD-K1 sight unit.

An inherent shortcoming of hydraulic components is the heightened risk of an internal fire in the event of a full turret perforation. Hydraulic fluid is highly flammable, and it would most likely cause and spread an internal fire very quickly. This is an especially serious concern to the T-72, since it has numerous shells in loose storage which can accidentally detonate from uncontrolled fires.

The hydraulic fluid used is MGE-10A, a type of mineral hydraulic oil with very low temperature sensitivity, having an operating range of between -65°C to 75°C. The entire system operates at 7.25 psi. This is quite dangerous, as with all hydraulic systems, because hydraulic oil may spurt out from burst tubes at high speeds, spraying large portions of the interior with the flammable liquid.

Automatic mode:


Maximum Cannon Elevating Speed: 3.5° per second
Minimum Cannon Elevating Speed: 0.05° per second


Maximum Turret Traverse Speed: 18° per second
Maximum Precise Turret Traverse Speed: 6° per second
Minimum Precise Turret Traverse Speed: 0.07° per second

Semi-automatic mode:


Maximum Turret Traverse Speed: 20° per second
Maximum Precise Turret Traverse Speed: 6° per second
Minimum Precise Turret Traverse Speed: 0.3° per second

Average time taken for complete rotation: 20 seconds

For a minimum traverse and elevation speed of 0.05° per second, the stabilizer should have an accuracy of 0.88 mils, equivalent to a stabilization accuracy (not mean deviation) of 0.88 meters at 1000 m. The speed of turret rotation is reasonable enough by Soviet standards, considering that earlier tanks like the T-55 were not very good. The turret of a T-55 with a "Tsyklon" stabilizer could spin around at 15 degrees per second, and the turret of a T-62 could do 16 degrees per second. Sirenevny is an improvement over earlier stabilizers in every possible way.

Combined, all of the components belonging to the stabilization system weigh a sum total of 319 kg, including the working fluid (hydraulic fluid). On average, the stabilizer system consumes 3.5 kW of power.

2E42-2 "Zhasmin" (Jasmine) Hybrid Electro-Hydromechanical Stabilizer

The 2E42-2 is a conventional stabilizer system as it combines an electric turret traverse and stabilization drive with a hydraulic gun elevation and stabilization drive. This stabilizer was being tested on experimental tanks in 1983 and only began to be installed in mass produced T-72 tanks since 1984. Naturally, the T-72B came with this improved stabilizer.

The T-72B obr. 1989 manual states that the maximum turret traverse speed is 16-24° per second, and mentions that the traverse speed when the target designation function is used by the commander is 16° per second. A rate of rotation of 24° per second is achieved under the "overcharge" condition. The gunner can get the turret to turn at this rate by turning his handgrips hard until it cannot physically go any further. The average power consumed by the stabilizer system is still the same as the 2E28M at 3.5 kW. The 2E42-2 has the same two operating modes as the 2E28M and previous Soviet tank gun stabilizers: automatic and semi-automatic.

Automatic mode:


Maximum elevating speed: 3.5° per second
Minimum elevating speed: 0.05° per second


Maximum Turret Traverse Speed: 16-24° per second
Maximum Precise Turret Traverse Speed: 3° per second
Minimum Precise Turret Traverse Speed: 0.07° per second

Semi-automatic mode:


Maximum Turret Traverse Speed: 16° per second
Maximum Precise Turret Traverse Speed: 6° per second
Minimum Precise Turret Traverse Speed: 0.3° per second

The turret traverse speed is improved to 24 degrees per second, enabling the turret to complete a full 360° rotation in just 15 seconds.

2E42-4 "Zhasmin" Electric/Hydroelectric Stabilizer

The 2E42-4 two-axis stabilizer is an improved modification of the 2E42-2 first used in the T-90. The 2E42-4 is installed in the T-72B3 and includes a much more powerful horizontal drive for faster turret rotation. According to Sergey Suvorov in "T-90: First Serial Tank of Russia", the 2E42-4 for the T-90 has an average stabilization accuracy of 0.4 mils in the vertical plane and 0.6 mils in the horizontal plane.

The 2E42-4 stabilizer offers a huge weight reduction of 120 kg over the 2E42-2 stabilizer, for a total weight of 200 kg. This is mainly because of the design simplification of the hydraulic gun elevation drive, the improved turret traverse motor, and the usage of solid state electronics in the digitized control systems. The screenshot below gives us a good view of the hydraulic pump for the gun elevation drive. The pump is mounted below the breech, and connects to the hydraulic elevator piston seen in the upper right corner of the picture.

Screenshot taken from the RT Documentary show "Tanks Born in Russia (E5) Kirill’s girlfriend reveals her biggest secret" (link).


Maximum elevating speed: 3.5° per second
Minimum elevating speed: 0.05° per second


Maximum turret slew speed: 40° per second
Minimum turret slew speed: 0.054° per second

The much faster turret traverse speed enables the turret to complete a full 360° rotation in 9 seconds.


Manual traverse and elevation is possible with all T-72 turrets through the use of two flywheels located behind the hand grips. There are two gear settings; "coarse" and "fine". The former allows the turret to turn as fast as the gunner can work the flywheel, whereas the latter allows the gunner to produce minute changes to the lay of the turret and gun. Gun laying with the manual traverse can be just as accurate as with stabilizers if not more so, though obviously much, much slower and nearly impossible to achieve on the move. The high level of accuracy is due to the extremely large gear ratio in the "fine" setting. In this setting, multiple cycles of the handwheel will shift the turret by only a few fractions of a degree, allowing the gunner to conduct extremely fine adjustments to the lay of the gun. The gun elevation flywheel has a solenoid button for firing the main gun, and the gun itself houses a manual trigger mechanism that can be used in the event of an electrical failure.

In many early tanks with powered turret traverse, the gun elevation drive still tended to be manually driven because a well-balanced gun could be elevated and depressed at sufficient speed with a handwheel and there was little need to quickly shift the gun more quickly. However, quick turret traverse with a manual drive became impossible when the mass of tank turrets grew sharply during and after WWII.

According to the proposal "Один Из Путей Повышения Надежности Комплекса Танкового Вооружения" ("One of the Ways of Improving the Reliability of the Tank Armament Complex") by A. I. Mazurenko and E. A. Morozov, the extremely low speed of turret rotation for a generic modern tank using manual controls makes it impossible to find and engage targets in combat conditions. Using the manual turret traverse handwheel, the speed of turret rotation on flat ground is 0.6-0.8 degrees per second and the speed of turret rotation when the tank is on a side slope of 15 degrees is halved to only 0.3-0.4 degrees per second. The amount of force required to work the flywheel is 1.6 kgf (15.7 N) when the tank is on flat ground and 20.0 kgf (196 N) when the tank is on a side slope of 15 degrees. Tracking a moving target is virtually impossible unless the target is exceptionally far away, and scanning for targets becomes impractical. From this, it is understood that the loss of powered turret controls can effectively bring a tank out of commission even if all other systems are operating at normal levels.


The T-72 first received a meteorological sensor unit with the T-72BA sub-variant. This manifested in the form of the DVE-BS unit, which can detect changes in wind speed and automatically register it in the ballistic computer. The maximum calculable winds speed is 25 m/s, which is equivalent to a 10 on the Beaufort scale. Such wind speeds are capable of uprooting trees and are very rarely experienced inland. The information gathered is synchronized with the automatic lead calculation unit found in the 1A40-1 sighting complex. The T-72B3 is also equipped with a DVE-BS unit.


The T-72 is equipped with the 125mm smoothbore D-81T cannon, otherwise known as the 2A26 and the 2A46. It can fire a wide range of shells including; APFSDS, HEAT, HE-Frag, and even guided missiles beginning from 1985 in the T-72B model. The cannon is loaded with two-part semi-caseless ammunition.

The cannon is partially derived from the U-5TS 115mm smoothbore gun, and the evidence of this heritage can be found upon close inspection. The recoiling mechanism is much more compact, which helps to reduce the volume of internal space taken up by the cannon in the small turret. Like the U-5TS gun from the T-62, the recoil buffer is located at the bottom of the breech block rather than on top of it, so despite the larger caliber and mass of the cannon, it was possible to create a very low turret with a very steeply sloped roof while still affording the cannon a reasonably range of vertical motion.

The gun can elevate +14 degrees and depress -6 degrees when facing the front, but elevate +17 degrees and depress only -3 degrees when facing the rear, over the engine compartment. This is generally sufficient for cross-country driving with lots of minor dips, dives and bumps, but the T-72 is unable to fully exploit the reverse slope of steeper hills to enter a hull-down position. This was not seen as a drawback because pre-prepared dug-in defensive positions were preferred for defensive operations. The lackluster gun depression compared to most NATO tanks has the potential to become an issue in highly irregular terrain when firing on the move, but it was considered to be not significant enough to warrant the corresponding changes to the turret height. Compared to previous Soviet tanks, the gun depression of the T-72 is slightly better than the typical -5 degrees.

Contrary to common perception, it is possible to extract loaded rounds from the chamber of the cannon despite the fact that the T-72 uses an autoloader, and in fact it is a basic necessity on all tank cannons. The cannon has a manual breech opening mechanism and an extractor. Spent casing stubs from the semi-caseless propellant charges are ejected from the chamber at a speed of 14-18 m/s after each shot, and the ejection of unfired propellant charges is done manually using the breech opening handle.

The end of the barrel has four shallow cuts in the shape of a cross.

This is for the gunner to align and tie two pieces of string into a crosshair over the end of the barrel for the purpose of zeroing the gun in the field. According to the manual, the gun and sights are zeroed by aiming both devices at the same point at a distance of no less than 1,600 meters.

The photos below show a crew member affixing the cross on the muzzle.

Worn out barrels tend to exhibit worse accuracy and usually reduce the muzzle velocity of the projectile due to the incrementally increased diameter of the bore. This was especially noticeable during the first Gulf War, where Iraqi T-72s often urgently needed barrel replacements because they had been used since the Iran-Iraq war. Because of the embargo on military equipment, they had no access to fresh barrels and they lacked the sensitive technology to ensure a steady supply from local production facilities. Firing APFSDS shells, especially the first generation ones that Iraq was supplied with (steel sabot with copper driving bands, and bore-riding projectile fins) was harsh on the barrel bore, especially since exported cannon barrels did not have a chrome lining. The 2A46M-2 on the T-72B could fire 220 contemporary APFSDS shells, but the latest 2A46M-5 gun barrel can apparently tolerate fire 500 of such rounds.

The reason that firing rounds out of a worn-out barrel is dangerous is because the bore has been eroded enough that there may be gaps between the driving band on the projectile and the surface of the bore. This permits propellant gasses to blow past the projectile, causing dangerous pressure fluctuations in the barrel. As the photo below shows, this can be very dangerous. Fracturing of the barrel is possible, but thankfully, the fuses of explosive ammunition like HE-Frag and HEAT shells exclude the possibility of premature detonation. Still, disintegrated fragments may potentially harm people and equipment in the vicinity.


The original Obyekt. 172 and Obyekt 172M prototypes use the 2A26M-2 (D-81T), and earlier T-72 Ural tanks was armed with the 2A26M2 as well. The 2A26M2 is a derivative of the 2A26 gun from the T-64A with minor modifications for the T-72. Like many guns from the 1960's, the 2A26M2 lacks a thermal sleeve. The recoil buffer is installed asymmetrically to the barrel axis at the bottom right hand corner of the breech block, and the recuperator is installed directly underneath the breech block. The asymmetric installation of the recoil buffer resulted in the unbalanced motion of the cannon during its recoiling cycle while the shell is still in the barrel, and the unbalanced motion generated somewhat more intense oscillations at the muzzle compared to a symmetric recoil system, resulting in larger shot dispersion. This configuration was carried over from the U-5TS gun.

The size of the 2A26M2 breech block is only slightly than the U-5TS gun. The height of the gun axis from ground level is 1,651mm. The length of the gun chamber is 840mm and the length of the barrel is 6,000mm (48 calibers). The maximum safe chamber pressure is 450 MPa.

The 2A26M2 cannon could have an electroplated chrome lining applied to the barrel bore but most if not all of the guns manufactured in the USSR and in Warsaw Pact nations had only steel bore and a correspondingly low durability, although specific details on this esoteric topic are very scarce. Depending on the source, the barrel could have a service life of up to 600 EFC (Effective Full Charge), equivalent to 600 shots with a full powered cartridge like HE-Frag. Less than a hundred rounds of APFSDS ammunition could be fired through the barrel and replacing it was not an easy task. The turret had to be lifted by a crane and positioned so that the gun assembly could be removed through the rear. This was a highly time consuming process that required specialized equipment. In the field, the crane would have been provided by recovery vehicles. In this regard, the Soviet tank gun design was very much behind their Western counterparts. For example, the 90mm gun on the M48 Patton (a 1950's tank) already featured a quick change barrel.

The photo below shows a T-72 Ural participating in exercises; notice the lack of a thermal shroud. Exported 2A26M2 cannons installed in T-72M tanks are known to lack a chrome lining, but it is unclear if all the barrels used in domestic T-72 Ural tanks had a chrome lining if at all.

A fume extractor is installed approximately 3/5 down the length of the barrel. It is a concentric type, and of an extremely simple design with two angled vent holes and two perpendicular intake holes with valves. When a shot is fired, the propellant gasses enter the fume extractor chamber through the four holes, and during the rapid pressure equalization inside the barrel (drops to 1 atm) after the projectile leaves the muzzle, the valves on the intake holes close and the pressurized gas accumulated in the fume extractor is forced to exit into the barrel through the two angled vent holes. The angled vent holes direct the gasses forward at a speed of around 500 m/s for a duration of 1-1.5 seconds, reducing the pressure inside the barrel by 3-5% compared to atmospheric pressure thus creating a suction force that removes lingering fumes in the barrel and ejects them through the muzzle.

The casing of the fume extractor on the 2A26M2 and all of its derivatives is made from steel, which is logical considering that the fume extractor is a temporary vessel for high pressure gasses. The length of the fume extractor is 840mm.


In 1970, the 2A46 was created as a modernization of the 2A26M cannon to rectify its most glaring issues. New technologies were mainly applied to the design and manufacturing of the barrel, but the rest of the cannon was not neglected. Various improvements increased the durability and accuracy of the new barrel, raising it to around 900 EFC. The maximum rated chamber pressure was not increased from the 2A26M2 and remained at 450 MPa, and the recoiling system remained essentially identical to the 2A26M2. The typical recoil stroke is 270mm to 320mm, and the maximum is 340mm. The full weight of the cannon is 2.4 tons, and the barrel and breech block have a combined weight of 1.82 tons. The barrel alone weighs 1.156 tons.

To manually open the breech, the commander has to pull on the breech opening handle in a single stroke, which is not easy in the confines of the tank as the pull weight of the handle is 245 N.

According to "Increasing Firing Accuracy of 2A46 Tank Cannon Built-in T-72 MBT", 81% of 3BM-15 APFSDS shots fired from an unmodified 2A46 cannon will land within 0.5 meters of the aiming point in the vertical axis at a distance of 1 km, and 43% will land within 0.5 meters of the aiming point in the horizontal axis at the same distance. From this, it was calculated that the probability of hitting a T-72 tank target with the 2A46 at a distance of 2 km is 57%. Keep in mind that this is the mechanical accuracy of the cannon alone. Errors from the rangefinder and fire control system of the T-72 tank will most definitely reduce the probability of hit, especially since the dispersion of shots in the vertical axis are higher than in the horizontal axis. Rangefinding errors will exacerbate the high vertical dispersion. The higher vertical dispersion is due to the asymmetric recoil buffer assembly inherited from the 2A26M2, which in turn was inherited from the U-5TS.

The barrel of the 2A46 cannon was enveloped by a plastic thermal shroud in order to eliminate barrel bending issues from meteorological conditions. For example, the heating of the barrel from sunlight only occurs on the surface of the top of the tube, so the metal on the top part expands whereas the metal on the bottom part does not. This causes the barrel to bend downwards. Bending may also occur when the barrel is cooled unevenly, such as by light rain or by a crosswind. A thermal shroud solves this issue by insulating the barrel from sudden external temperature changes. However, this has the side effect of increasing the heat retention of the barrel from firing and from prolonged solar heating, so the barrel would overheat faster and the sustained firing rate of the cannon would have to be curbed. To address this issue, the thermal shroud on the 2A46 is built with an air gap between the barrel and the plastic jacket and with holes in the bottom of the jacket to allow hot air to flow out, as shown in the photos below (credit to Azrael Raven from Because the holes are located at the bottom, rain water cannot enter the jacket and cold air from wind cannot enter in significant amounts.

The thermal shroud is comprised of four segments, almost covering the entire length of the barrel except for the muzzle and a section of the base of the barrel in front of the armoured gun mask where the barrel is thickest. The lack of a shroud on the base of the barrel is mainly because of the need to account for the recoil stroke of the cannon, as the clearance between the gun mask and the gun barrel is very narrow. The length of the uncovered section at the base of the barrel is 540mm. The length of each thermal shroud segment is not the same: counting from from the base and ending at the muzzle, the first two sections are 836mm in length and the other two sections are 632mm in length. The addition of the thermal shroud is a source of increased accuracy, although it tends to be only noticeable in real world conditions and not during controlled bench testing where environmental conditions are always kept constant.

The T-72A began its service life in 1979 with the 2A46 and the T-72M (Obj. 172-E2) export model also used the 2A46. Several years later, the 2A46 was replaced by the newer and more accurate 2A46M, but its service life did not end there. The 2A46 was reverse engineered by the Chinese and an improved copy of the gun called the ZPT-98 is currently in production to equip the current generation of Chinese main battle tanks, including the Type 96, Type 99 and even the latest VT-4. A video of the assembly of a Chinese 2A46 is available on YouTube as part of a documentary on the VT-4. The screenshots below are taken from the video and show a disassembled ZPT-98 cannon.

One of the differences between the ZPT-98 and the 2A46 is the gun cradle. The cradle on the ZPT-98 closely resembles the one on the 2A46, but the trunnion is derived from the D-10T gun mounted on the Type 59 tank. Other than that, the ZPT-98 is built with features derived from the 2A46M including a quick-change barrel and various other maintenance-oriented amenities.


In 1974, NII Stali mastered several advanced material processing technologies, which were subsequently transferred to the production of new cannons. These new technologies included electroslag remelting steel, differential isothermal quenching and improved thermomechanical processing. When the decision to modernize 125mm tank cannons was made in the late 70's, these technologies were implemented into the 2A46M cannon, among other things. According to Mikhail Baryatinsky, new T-72A tanks from 1981 onward were produced with the 2A46M. It is written in "Combat Vehicles of Uralvagonzavod: T-72 Tank" that the T-72 with the 2A46M cannon passed acceptance tests in 1978. The 2A46M was installed with the T-72B since its introduction in 1985. The 2A46M is also noteworthy for having the necessary electrical components that render it capable of launching guided anti-tank missiles.

The barrel life was substantially improved by the use of a new, more durable chrome lining to reduce wear from new high-energy APFSDS shells. Thanks to the new chrome lining, the barrel life was increased to 1200 EFC. Accuracy was improved by a very impressive 50% due to the completely revised recoil system. The photo below (credit to Dmitry Derevyankin) shows the symmetric installation of two recoil buffers at the top right and bottom left corners of the breech block, and the retention of the recuperator at its original position directly underneath the breech. The symmetrical installation of two smaller recoil buffers greatly reduces the moment (the turning effect of a force) experienced by the cannon during the recoiling cycle, and thus reduces the oscillations at the muzzle while the shell is still in the barrel. The 2A46M has a typical recoil stroke of 260mm to 300mm, and a maximum recoil stroke of 310mm.

The cannon in the photo below is actually a 2A46M-1 for the T-64BV/80BV, but the breech block is otherwise identical to the 2A46M. The only differences are in the shape of the breech guards and in the presence of an electric motor for raising and lowering the shell casing stub ejection mechanism.

Furthermore, the method of seating the barrel to the gun cradle was changed. According to a marketing presentation by UVZ, the seating of the barrel was changed from the combination of the breech ring and support from a single contact point with the cradle to purely cradle support with two contact points. The small gap between the cradle support and the barrel (less than 1mm) in all D-81T pattern guns is a free expansion zone to minimize the impact on the ballistics of the barrel from thermal expansion after multiple shots are fired. By having the free expansion zone, the number of contact points between the cradle support never changes and the performance of the gun is more consistent.

Assuming that the increase in the mechanical accuracy of the 2A46M over the 2A46 is indeed 50%, then the probability of a 3BM-15 round fired from the 2A46M hitting a T-72 tank at 2 km would be 85.5% instead of 57% for the 2A46, if all other conditions are equal.

The 2A46M was also a milestone product in another way: the new mounting system for the barrel enabled quick replacement in the field from the outside of the turret by pulling it out from the front, without needing to remove or shift the turret. The procedure reportedly takes around 2 hours, but it is not clear if this is for an operation done in a depot or in the field. The maximum rated chamber pressure was increased to 500 MPa in accordance with the appearance of high energy APFSDS shells. Due to the relocation of the recoil buffers, the manual breech opening mechanism was redesigned, but remained principally identical. The oil level in the recoil buffers and recuperator can be checked without the opening of the stopper caps, making it much simpler and quicker to perform routine maintenance on the gun. Previously, tank crews referred to a schedule to record and determine if the buffer and recuperator in the 2A46 required a top up. If that information is not available, then the cannon would need to be retrieved from the turret and it would have to be inspected on a testing mount and tested with a pullback winch.

The overhaul of the design of the cannon also brought improvements to some of its less major components. Most notably, the manual breech opening mechanism at the top left corner of the receiver had to be redesigned because of the location of the new recoil buffers. The new mechanism has a ratchet so that two tugs on the lever are needed instead of one, and the pull weight for the lever was decreased considerably. This makes it much easier to manually open the breech within the confines of the tank. All of the modifications increased the mass of the cannon to 2.5 tons.


The T-72B3 builds upon the T-72B with the inclusion of the 2A46M-5 gun (D-81TM-5), which was first introduced in 2005 and used in the T-90A. The 2A46M-5 can be considered the most perfect of the entire series thus far.

The dynamic balancing of the barrel during the firing procedure (while the shell is still in the barrel) have been better tuned, thus minimizing oscillations at the muzzle. The barrel itself was improved, now having 11% greater rigidity than the 2A46M barrel. The final result is a further reduction in shot dispersion. The maximum rated pressure in the barrel was increased to 608 MPa. According to the manufacturer, dispersion of all shell types by an average of 15% to 20%, and the accuracy when firing on the move has been increased 1.7 times, thanks to the greatly decreased vibration of the gun the tank is in motion over rough ground. Overall, the estimated probability of hit in combat was increased by 20-29% for APFSDS ammunition, 4-12% for HEAT ammunition, and 21-38% for HE-Frag ammunition.

The 2A46M-5 follows the 2A46M with the inclusion of a quick-replacement barrel. Like before, the barrel is released from the gun chamber and receiver assembly by twisting it by 45 degrees fitting a special hexagonal wrench on a hexagonal part of the barrel. The threads that lock the barrel to the receiver are seen in the screenshot below, taken from a news tour on the No. 9 Factory which builds these guns.

As you can see in the photo below, the location of the recoiling mechanism elements remained unchanged from the 2A46M. The main differences were not so obvious. More photos of the 2A46M-5 are available on Stefan Kotsch's website.

The drawing below (from here) showcases the location of the recoil buffers and the recuperator in relation to the axis of the cannon barrel. The drawing is probably valid for the 2A46M as well.

Furthermore, the 2A46M-5 is provisioned with special notches at the muzzle of the barrel, which are used for boresighting. Using the sights, the gunner aligns special markings to the notches and calibrates the sights to the gun from the recorded angle; a process that takes only 1 minute.



The T-72 uses the AZ electromechanical carousel-type autoloader with a 22-round capacity. All 22 rounds are stored in individual cassettes arranged radially around a central hub which houses the carousel rotation motor and drive as well as the power supply for the turret. The autoloader was modernized in the T-72B to missiles to be carried. The new autoloader had higher reliability, and could also store longer projectiles. We will first examine the original Ural autoloader (known as the AZ-172), and then examine the newer T-72B autoloader (known as the AZ-184) in the context of improvements to the original. The patent for the T-72B autoloader (Russian Patent No. 2204776) is available hereAs mentioned earlier in this article, the gunner controls the autoloader from a control unit located on the TPD-K1 sight.

According to the memoirs of Leonid Kartsev, the AZ-style autoloader was developed independently by the UKBTM design bureau after Kartsev inspected a T-64 during its trials and felt "trapped" by the ring of ammunition surrounding the turret ring, noting that the crew in the turret had no form of contact with the driver besides the intercom system. Upon his return to Nizhny Tagil, he ordered the creation of an autoloader system with a different scheme of ammunition stowage to surpass the MZ autoloader of the T-64, resulting in the AZ autoloader. This new autoloader was first fitted to the Object 167M before being implemented in the Object 172 and then finalized in the T-72 Ural as the AZ-172. The AZ autoloader traded the larger 28-round capacity of the MZ autoloader for an all-electric motorized system, a slightly reduced 22-round capacity and a lower profile. The nuances of this autoloader will be the topic of discussion.

Each shell and propellant charge stored within the carousel is housed within a two-tiered steel ammunition cassette with extended bills to properly line up the shell or propellant charge with the gun chamber. The diameter of the carousel spans the width of the hull. Below, you can see the ammunition cassettes being dropped in place on a T-72B3 autoloader. The carousel rotates independently of the turret and the armoured bulkhead/false floor on top of it during both normal and manual operation.

The notch on the edge of the central hub marks where the tray lines up flush with the trapdoor on the carousel cover. The notch allows projectiles that are physically longer than the ammo cassette to pass through the trapdoor.

The diameter of the autoloader carousel is around 1,800mm and the height of the carousel is around 450mm. The maximum length of each cassette is 680mm: just 2mm longer than the HEAT projectiles carried by the T-72 like the BK-14 and BK-18, and only 5mm longer than HE-Frag shells like the OF-19. The APFSDS ammunition supplied to tanks during the Cold War was the shortest among the three main ammunition types, and even the 3BM-46 "Svinets" projectile fits easily into the ammo cassette.

Modified cassettes are used in the T-72B carousel in order to accommodate guided missiles. The modified cassettes have special latches on both sides accommodate guided missiles and to prevent the stabilizing fins of the missile from accidentally deploying when the missile is rammed into the cannon.

Although the new cassettes are designed to accommodate guided missiles, the length of the cassettes remain at 680mm, so the 9M119 guided missiles (695mm long) supplied to T-72B tanks will overhang the cassette by 15mm. The main factor enabling guided missiles to be carried in the autoloader carousel was not the modified cassettes, but the reduction of the diameter of the hub of  the carousel. Thanks to the provisions for accommodating anti-tank missiles, APFSDS rounds of higher elongation are compatible with the AZ-184 autoloader.

The autoloading cycle requires the gun to be locked at a pre-programmed elevation of +3°30' which is done so automatically as the cycle begins. It is claimed in the memoirs of Leonid Kartsev that this was superior to the T-64A as the spent shell stub was a significant source of propellant fumes from smoldering residue inside the stub, and that disposing of the stub reduced the concentration of fumes in the fighting compartment. There is truth to this claim, as video evidence has shown that even when there is virtually no escape of fumes from the cannon breech after firing (indicating that the fume extractor is working well), the spent shell stub may still pollute the fighting compartment until the propellant residue is completely burnt. This video is a good example of this. We can also see from  videos of artillery crews  that this is also true for fully cased unitary ammunition, so the immediate ejection of the spent shell stub from the turret of the T-72 is clearly beneficial to the working environment of the crew.

Steven J. Zaloga claims that there were some problems with zeroing the sighting system and the cannon because the sight was independently stabilized, and the vertical stabilizer for the cannon would sometimes fail to synchronize with the stabilizer unit in the sight as the cannon resets to its original position when finishing its loading cycle. However, it is doubtful if this issue truly exists, because the stabilizer for the cannon is slaved to the independently stabilized TPD-K1 sight so the stabilizer will always attempt to lay the cannon as close as possible to the aiming point of the sight. The alignment will never be perfect, because the weapons stabilizer is less precise than the stabilizer for the sight. Zaloga may have mistakenly labelled the small alignment error between the two cross-linked systems as a design flaw instead of the mere technical limitation that it actually is.

During the reload cycle, a cassette is elevated to the ramming position by an electric chain-driven elevator, and the two-part ammunition is rammed into the gun breech. First the shell is loaded, and the propellant charge follows. Because the cannon automatically elevates by +3°30' degrees at the beginning of the reload cycle, the top half of the autoloader elevator is slightly tilted by the same angle to bring the cassette into alignment with the breech. The slight tilt is visible in the diagram below. The diagram shows a T-72 Ural type autoloader. The T-72B has a modified carousel, but the autoloader is otherwise identical.

The use of an electric ammunition cassette elevation mechanism instead of the hydraulic system of the MZ autoloader enables the volume of flammable hydraulic fluid located close to the ammunition to be reduced, thus indirectly increasing the survivability of the tank.

Propellant charge casing stubs are automatically ejected by a stub catcher (seen in the drawing above labelled '4') through a small port at the rear of the turret, visible below:

The stub catcher mechanism registers the presence of a casing via a switch installed in a paddle-shaped arm located behind the stub catcher, shown in the drawing on the right. The paddle-shaped arm also keeps the stub from falling out the back of the stub catcher. When a casing stub is ejected from the cannon, it hits a pin (4) at the back of a paddle-shaped arm which lifts a conductive flat spring (5) off an electrical contact (6), breaking the circuit and prompting the system to register the presence of a propellant charge casing stub in the stub catcher mechanism.

When the gunner activates the autoloader to load the cannon, the stub catcher mechanism is lifted to the ejection port in the turret roof by a linear actuator installed underneath the breech block of the cannon, and the stub is ejected by spring-loaded ejectors that are charged by the recoil of the cannon via a mechanical linkage connected to the breech block, as seen in the drawing on the left. The paddle-shaped arm is swung away to the side (in the direction of the gunner) by the ammunition cassette elevator to make way for the ramming mechanism.

When loading the cannon manually in the total absence of electrical power, the spent casing stub held in the stub catcher mechanism is removed by manually swinging the paddle-shaped arm away. After that, the casing stub can just be pushed out the back of the stub catcher and thrown away. The spent casing stub must be removed after every shot when firing the gun manually. If not, the casing stub from the next shot would hit the previous stub and bounce off, possibly hitting someone in the face. Bear in mind that these stubs are made from steel, weigh 3.45 kg, and are ejected from the cannon at a speed of 14-18 m/s.

Unlike what most people assume, it is completely possible to unload the cannon after a shot is already loaded. Although not always necessary, this may be useful when two different threats appear one after another. For instance, enemy infantry may appear after the T-72 has finished dealing with an enemy tank, so switching to HE-Frag will be necessary. If an APFSDS round is still loaded in the cannon, it will have to be unloaded either by returning it to the autoloader or firing it off, perhaps into the previously knocked-out enemy tank to ensure that it is truly disabled. If it is desirable to conserve APFSDS ammunition, then the loaded round should be unloaded from the back end of the cannon rather than the front. This can be done by simply using the commander's autoloader control box. The empty ammunition cassette is raised from the carousel, whereby the commander can open the breech manually using the breech operating handle. The ejection of the propellant charge is done automatically by an ejector built into the cannon. The action of pulling the breech operating handle opens the breech and returning the breech operating handle to its stowed position manually activates the ejector to push the propellant charge out of the chamber. The chambered projectile is not ejected; it must be extracted by hand. The propellant charge and projectile are inserted into the empty cassette and returned to the autoloader carousel. After this, the gunner can continue the target engagement process as normal. The amount of time needed to complete this operation is non-trivial, but it is absolutely possible if the situation calls for it.


The autoloader is able to recognize the position of each round stored in the carousel using the carousel storage memory unit, as shown below (AZ-172 model). Three ammunition types can be indexed into the carousel in this model.

To load ammunition into the autoloader, the commander must use his control box to cycle between cassettes. After loading a cassette, he must input the ammunition type into the memory unit by pushing one of three buttons, one for each type of shell: HEAT (К), APFSDS (б), or HE-Frag (O). When any one of the three buttons is pushed, the memory unit records the corresponding ammunition type and returns the cassette to the carousel, then the carousel spins to the next empty cassette.

The memory unit indexes the type of ammunition on a data disc stored inside the circular housing. The type of ammunition is identified using a binary system on the data disc. There are twelve radial magnetic rings on the surface of the disc divided into three groups of four rings and twenty two sectors; one sector for each round stowed in the autoloader carousel. In each group, four of the radial rings are for recording the type of ammunition, four are used to determine when to brake the carousel rotation motor, and four are used to determine where to stop the carousel in order to line up the ammunition to the trapdoor. As there are twenty two sectors with twelve data recording elements capable of storing one bit each, the data disc is considered to have a storage capacity of 264 bits.

Recording the ammunition information is done by three current carrying pins connected to the magnetic rings via electrical contacts. Data storage is done by changing the polarity of the sector groups to either positive or negative using the electrical contacts. The electrical contacts are spring loaded to keep them in contact with the magnetic rings to ensure that reading and writing the data is still possible even while the device is experiencing strong vibrations such as when the tank is on the move over rough ground or after the tank is hit with the shockwave of an explosive blast. However, the constant pressure wears out both the magnetic ring and the electrical contact over time leading to a loss in the ability to record and read data, and the conductive metallic dust produced by the rubbing of the contact on the ring surfaces can contaminate other parts of the unit, causing reading errors. Such errors could prevent the autoloader from accepting new ammunition when loading it, or cause the autoloader to lose track of where ammunition is stored or even to "forget" when to stop rotating the carousel if it is already in motion, so that it rotates indefinitely. Even if the recording surfaces are not worn out, it is also possible for the device to fail from the accumulation of dust and grime over time. At this point, it is possible to either replace the electrical contacts the hard disc, or replace the entire memory unit. This is an easy task that can be done in the field as long as a spare memory unit is available, as the replacement of the unit only requires that the old one has its electrical cable unplugged.

The rotation of the data disc is not powered by an internal motor, but by the carousel rotation motor via a driveshaft passing through the bottom of the memory unit. When the autoloader loading procedure is activated by the gunner, the carousel motor receives the command to rotate, but it does not know when to stop until the memory unit reaches the appropriate ammunition type, so if the gunner selects HEAT rounds, the carousel will rotate until the system reads the appropriate binary code on the data disc whereupon the command to brake and stop the carousel motor is read and processed by the autoloader.

In other words, the system does not know what the shortest route to the selected ammunition type is. This system limits the carousel to rotating in only one direction even though the motor is actually capable of rotating in both directions. Due to the system limitations, the reverse rotation is only activated when braking the rotation of the carousel. This is unlike the MZ autoloader of the T-64A and T-80 which has n autoloader memory system that is able to dynamically read the positions of all ammunition stored inside the carousel and display it digitally on a special circular device. However, the MZ autoloader is also not capable of rotating in both directions. Instead, the MZ autoloader has a "sequence" loading mode where the carousel will automatically spin to the next round of the type selected by the gunner immediately after loading. This eliminates the delay caused by the slower carousel rotation motor of the MZ autoloader.

After the round is loaded and the ammo cassette returns to the carousel, the memory unit instantaneously rewrites the data to a zero value to represent the empty status of the cassette so that the autoloader will ignore empty cassettes when loading. Conversely, the autoloader will only stop at empty cassettes when replenishing the carousel.

The design of the data disc is clearly an extremely simple and archaic form of a hard disc storage with an extremely low storage capacity. The lack of sophistication, however, is completely justified by the lack of a need to store large volumes of data and the high robustness required of the system. The simple design of the memory unit grants high resistance to shock and mechanical damage, and its self contained housing facilitates quick replacement if it is damaged. Apparently, one of the most common source of autoloader malfunctions is the memory unit.

Due to the limit of three ammunition types, this memory unit is not used in the T-72B, as there is a new type of ammunition: guided missiles. According to the patent for the T-72B autoloader (Russian Patent No. 2204776), the memory storage was upgraded to accept a fourth ammunition type; missiles. The upgraded memory storage unit was also improved for better reliability.

The upgraded memory storage unit had a rotary dial instead of three buttons. The dial has four positions for the four ammunition types. To select and index an ammunition type, the dial is turned to one of the four positions, and then pressed. A closer look at the dial is available in this video at (3:08). The new memory unit can be seen in the screenshot below.

The photo below shows a T-72 Ural or T-72A, as evidenced by the welded appliqué armour plate on the upper glacis. Note the T-shaped box with wires coming out of it to the left of the blue torsion bars, at the center of where the carousel would be.

The T-shaped box is a VKU-330-4 power distribution unit to supply power to the tank turret. The VKU-330-4 is shown below.

The permissible length of projectiles in the T-72B autoloader carousel was increased by reducing the size of the central hub. This was done by redesigning the hub and replacing the VKU-330-4 power distribution unit installed on top of the carousel rotation motor with the VKU-1 unit. The photo below shows a T-72B3 with a VKU-1. Note the three protruding arms instead of a T-shape.

This modification enabled the 695mm-long 9M119 guided missile to be used with the autoloader carousel. The T-72B1 uses the AZ-172 autoloader and memory system since it is a low cost version of the T-72B without the missile firing capability, so the upgraded autoloader is not necessary.

In the T-90A autoloader, the system was revised and digitized. The information on the type and location of the ammunition in the carousel is stored digitally in a separate device, and the shortest distance to reach the ammunition is determined by an algorithm. The absence of the old disc-type memory unit is confirmed in the photo below, although the crankshaft housing from the carousel that would have rotated the hard disc is still present as a "vestigial tail" of sorts. This is evidence that although the control system was overhauled, the T-72B carousel was retained. The issue of two-way rotation is resolved by the implementation of a sufficiently sophisticated control system. The carousel rotation motor itself is reversible and has always been capable of both clockwise and anti-clockwise rotation since the original version in the T-72 Ural, but due to the rather crude ammunition retrieval system, the reverse function of the motor had only been used for braking until then.

There are some claims that the T-72B3 uses the autoloader from the T-90A, and that this allows the T-72B3 to use more elongated APFSDS rounds. Currently available evidence shows that this is not the complete truth. A T-72B3 with the old T-72B memory unit (Red) and commander's control box (Yellow) can be see in the photo below.

This shows that the ammunition indexing and retrieval system is still based on the older T-72B, so the carousel must also be from the T-72B. However, it is clear that the system has been revised. Note that the old ammunition selector dial has been replaced with a new one. The photo below - this time showing the T-72B memory unit (Red) in a T-72B3 obr. 2016 - supports this theory. Even in 2016, the T-72B3 is evidently still using the old T-72B carousel, and even the same control box (Yellow) is used.

The T-72B3 obr. 2016 is indeed modified to accept longer APFSDS rounds. More specifically, it is designed to accommodate Svinets-1 and Svinets-2. This is according to a saved copy of order document No. 31603190542 by Uralvagonzavod corporation in a government registry of purchase documentation which contains two mentions of modernizing the autoloader to enable the use of products designated as "S-1" and "S-2".

One similarity between the T-72B3 and the T-90 autoloaders is in the sequence of actions of the stub ejection port hatch which is observed to momentarily open and close immediately after firing without actually ejecting a shell casing stub, presumably to evacuate the fumes. This feature was first seen in the T-90 as displayed in this videothis video and this video and many others. The fact that the T-72B3 also has this feature indicates that it shares something in common with the T-90 autoloader. As we have seen, there is evidence to show that the T-72B autoloader can load longer projectiles than the Ural autoloader, but there is nothing concrete that indicates that there were any further upgrades to projectile length after the T-72B. It is often assumed that the carousel is to blame for the limited projectile length, there is evidence that other factors are to blame for this limitation.

In June 2005, a patent (Patent No. 2300722) filed by UKBTM for a method of increasing the permissible length of projectiles usable in the autoloader was filed. The patent describes a modified autoloader elevator design wherein the ammo cassette is pulled backward to avoid the cannon breech as it is elevated to the ramming position. It is hinted in the patent that the main restriction on the projectile length is not the carousel, but the cannon. To be more specific, the patent states that a possible method of increasing the permissible length of projectiles involves moving the cannon forward, and that this would require significant reworking of the turret, and it would disrupt the balancing of the cannon. The carousel is not mentioned at all. It is not clear if this patented system was actually implemented in new production tanks or implemented at all, but since the carousel is not the main limiting factor, it is absolutely possible that the T-72B3 can simultaneously have the old T-72B carousel installed and still be able to fire the same shells as the T-90A.

The photo below shows the location of the memory unit for the Ural and the carousel trapdoor through which the two-piece ammunition passes through.

This scan comes from the book "T-72/72M/72M1 in detail", from preview pictures available on (link).

The time taken per shot is around 7 seconds. This enables the tank to achieve a maximum rate of fire of 7 to 8 rounds per minute. The cyclogram below shows the chronological order of the steps in the autoloading process.

The cyclogram gives a total loading time of around 7.7 seconds, but this is because the cyclogram includes the rotation of the carousel over two ammunition cassettes instead of transferring directly to the next one to represent a specific ordered arrangement of ammunition in the carousel or to represent a the time taken when switching ammunition types. The cyclogram also includes the firing and recoil of the cannon after the loading cycle, so the total time taken represents the time taken between shots rather than the time taken by a reload cycle. With a time between shots of 7.7 seconds, the AZ autoloader is on par with the MZ autoloader which achieves a speed of 7.5 seconds under the same circumstances (loading the third round in the sequence). The difference of 0.2 seconds is simply imperceptible.

As you can see in the cyclogram, the last second of the loading cycle is taken up by the release of the cannon from hydrolock and by the automatic laying of the cannon back into the last previous aiming position and then onto the new aiming point, so the gunner can open fire immediately, which is represented by the tag "Recoil of the cannon" that represents the firing of the cannon immediately after loading is concluded. This is possible because of the independent vertical stabilization of the gunner's primary sight and the separation of the turret traverse system from the rotation system of the autoloader carousel, so he can conduct ranging and aim at a new target during the loading cycle. This is no different from any other modern fire control system. The biggest drawback of the AZ autoloader is that it requires two ramming cycles instead of ramming the entire two-part cartridge in one go alá MZ autoloader.

The AZ autoloader carousel is very compact, as you can see in the photo below. Based on this official UKBTM drawing of the cross-section of a T-72, the carousel occupies around half of the internal height of the hull, so its height is around 450mm, including the top cover. Excluding the top cover, the carousel has a height equivalent to a standing propellant charge (408mm), including the struts that mount the carousel to the belly of the tank. Note that the carousel in the photo below has a cylinder attached to the central hub, indicating that this carousel is for a T-72B.

There is some additional equipment installed on top of the carousel cover. The silver box you see near the center of the carousel cover is a KR-175 relay box. It connects to the VKU-330-4 power distribution unit and supplies power to the turret.

The T-72 does not have a significant disadvantage when compared to human loaded counterparts, which include the majority of NATO tanks. Most examples can achieve a 4 to 5 second loading time - when their tank is immobile. However, it's a whole different story on rough terrain. An advantage to the autoloader is that a bumpy ride, change of direction or slope traversal will never affect the autoloader's operation in any way. It can maintain its normal cyclic loading rate in whatever condition or orientation the tank is in. In manually-loaded tanks, the whole vehicle will pitch and dive as it drives over ruts and mounds while the gun - which would be disconnected from the stabilization system in tanks like the Abrams when the loader drops the safety lever - will move up and down on its own volition, making it less straightforward for the loader to get the shell aligned with the chamber to ram it in.

Firing on the move is usually done at a low cruising speed or at a crawl in order to maximize accuracy, but a tank speeds up and performs evasive maneuvers in between shots in order to avoid enemy fire, before slowing down again to return fire. The stressful time between shots is when the loader must perform his duties, and it would generally be harder to load the cannon during that time. This video illustrates this point perfectly. At 1:08 and 1:31 in the video, the movement of the gun delays the loader by around a second, extending his loading time to 7.9 seconds and 8.2 seconds respectively (loading time is defined as the time between dropping the loader's safety lever and moving back to a position away from the path of recoil). This would not be an issue for a tank furnished with an autoloader, but to be fair, this is also not an issue for tanks installed with a loader's assist system where the gun automatically raises by a few degrees and fixes the breech in detente, placing it at the optimum loading angle for the loader. The earliest tank to have this feature was the T-54B, followed by the T-62. Later on, tanks like the Leopard 2 and the Merkava 4 featured similar loader's assist systems.

The autoloader can maintain its cyclic loading speed throughout an extended engagement until the carousel is exhausted. On the other hand, the speed of a human loader is affected by the location of the ammunition and he may become fatigued long before the ammunition is exhausted or even before combat even commences, whether it be due to excessive heat, excessive cold, shortage of food, shortage of water, or any other imaginable fact of life for a soldier fighting on the front lines. For instance, it was mentioned earlier in this article the personnel heater in the M60A1 and M60A3 was astonishingly unreliable and frequently caused the fire extinguisher system to discharge accidentally, and some tanks like the Chieftain did not have a heater at all. The efficiency of a human loader would be affected by this, whereas an autoloader would not.

All in all, the T-72's autoloader is entirely satisfactory for generating a sustainable rate of fire for realistic encounters. While NATO tanks with human loaders were intended to put out as many shots as possible on huge formations of approaching Soviet tanks while staying stationary behind cover, the T-72 never had such a requirement. In modern shoot-and-scoot combat where tanks rarely stop moving or risk getting hit themselves, the advantage of human loaders become much less apparent. In this sense, the T-72's autoloader is not a hindrance at all, but an advantage, if the system is not at least on par with its Western counterparts.


The overhead cover on top of the carousel acts as a false floor for the turrets' occupants. Here is a better view of the cover.

This close up of the surface of the autoloader carousel reveals that it is actually made of thin sheet steel, but it is covered in a layer of thick, rigid matting. The matting is actually a layer of "Podboi" anti-radiation lining with a thickness of around 30mm. "Podboi" is known to be an effective spall liner, so it serves as additional protection for the ammunition inside the carousel. The anti-radiation lining carried over from the T-72 Ural to the T-72A, T-72B and the T-72B3, but was removed in the T-90 and compensated by thickening the cover. A good view of the matting is visible in the picture below (screenshot taken from TV Zvezda series "Made In the USSR", episode "T-72 Main Battle Tank").

The sheet steel cover is bent down at the edges for structural stiffness, so the cover you see in the screenshot above does not represent its true thickness. In fact, the sheet steel cover seems to be too thin to be helpful against anything but the lightest spall fragments.

Externally, it appears that there is only one layer of "Podboi" matting on the top of the cover, but there is an additional layer on the underside of the cover on most sections, as shown in the drawing below. The total thickness of the two layers may exceed 50mm. This not only gives the turret occupants another layer of security from radiation, but also improves the protection of the ammunition.

The perimeter of the carousel is protected by sheet steel guards at certain places, as shown in the photo below. In other places, the perimeter of the carousel intersects with conformal fuel tanks. The thickness of the guards was increased for the T-72B autoloader. One of the original perimeter guards can be seen in the photo below (open image in new tab and zoom in). Note the two reinforcement ribs pressed in to the plate - this indicates that the plate is quite thin and flimsy.

This photo gives us a closer look at the guard. The sheet is quite thin, so it is more likely that its main function is to help prevent unintentional interactions between the driver and the carousel and not to protect the ammunition from spall and fragmentation.

The sheet steel guard can also be seen on the left of the picture below. Screenshot taken from TV Zvezda series "Made In the USSR", episode "T-72 Main Battle Tank".

In the T-72B, these ribbed steel guards were replaced with thick solid armour plates, as seen in the photo below of a late model T-72B undergoing repairs at the 103rd Armoured Repair Plant in the Far East (photo credit to darkbear-ru). It is very unlikely that the tank in the photos below is a T-72B3 model because the delivery of the very first T-72B3 tanks only began in 2013, whereas the photos below were uploaded to darkbear-ru's livejournal in December 2012. Also, the tank has clearly seen some use, as shown by the worn rubber rims of the roadwheels.

The locations of the armoured steel guards around the perimeter of the carousel did not change. One plate is located between the driver and the autoloader carousel and another plate is located between the air supply unit and the autoloader carousel. Thus, the carousel is protected in the 10 o'clock to 12 o'clock sector and in the 4 o'clock sector. The front right hull fuel tanks offer protection in the 1 o'clock sector and the rear conformal fuel tank offers protection in the 5 o'clock to 8 o'clock sector. The remainder is unprotected, as there is no equipment between the carousel and the sides of the hull.

The lower left corner of the screenshot below grants us a closer look at the steel guards for the T-72B3 carousel. It appears that the steel guard plate was not changed from the T-72B to the T-72B3.

The T-90A appears to have the same steel guard plate as the T-72B and T-72B3, as shown in the photo below (credit to twower). Note that there are two fire extinguisher canisters clipped to the guard plate. The same clips are seen in both of the photos of the T-72B and T-72B3, likely indicating that all three tanks have the same armoured plate installed in front of the carousel. The rather large gap seen in the photo below is only due to the top-down perspective of the photographer. When viewed horizontally, the armoured plate almost fully covers the side of the carousel.

As T-72B1 uses the autoloader of the T-72 Ural, it also retains the same thin sheet steel guard as shown in the picture below. Screenshot taken from a video by user Khercrit, titled "T-72: how a driver-mechanic crawls into the turret". You can get an idea of how thin the sheet steel cover atop the carousel really is in the screenshot below.

The type of steel used for the perimeter guards are not known, but the thickness of the older T-72 Ural-style sheet steel guards would be insufficient for real ballistic protection. The thick plate in the T-72B can be considered a serious armoured plate, and would probably have a positive effect on the survivability of the tank under some circumstances. It is important to point out that it is known that the guards are made from steel and not aluminium because rust is observed on the surface of the sheets.

This article translated by Peter Samsonov details the post-penetration effects of 125mm APFSDS ammunition. The original pages of the Russian document were first shared on Andrei Tarasenko's blog. The document featured in the article pertains to a lethality analysis done on 3BM-9, 3BM-15, 3BM-22 and 3BM-26. These four rounds will all be examined more closely later on, but for now, it is only necessary to summarize that the 3BM-9 is an all-steel "torpedo" projectile, while the 3BM-15 and 3BM-22 are composite shells with a a tungsten carbide core at the front of the projectile, and the and 3BM-26 has a tungsten carbide core in its tail. All of the shots were for a 60 degree obliquity impact, and the velocity of all of the shells corresponds to their velocities at 2 km.

According to the article, the vast majority of fragments expelled behind the armour plate are smaller, low energy particles that are only capable of penetrating 3-6mm of aluminium sheeting at a distance of 0.5 to 1 meters. Keeping in mind that the overmatch factor used in the experiments was in the range of 100mm to 300mm, these figures simply cannot be considered realistic if the same or equivalent ammunition was fired at a T-72 tank, but assuming that a composite shell managed to overmatch the front hull armour of the T-72B by 100mm to 300mm, most of the fragments will definitely not be able to penetrate the steel guard around the perimeter of the carousel, especially not after passing through the anti-radiation lining (which doubles as a spall liner) lining the interior walls of the tank. This is important, because igniting or detonating ammunition requires a certain amount of energy. Very low energy fragments that can barely pierce a millimeter of steel would have no hope of igniting the ammunition, and more energetic fragments may lose enough energy from impacting the carousel perimeter guard that they may fail to ignite the ammunition. The thick armour plate in the T-72B may even be able to protect the carousel from fragments that are capable of penetrating 30mm of aluminium or more, of which there are comparatively few. It is not known what type of aluminium alloy was used for the plates in the post-perforation analysis, but is is likely to be structual aluminium and not armour-grade aluminium. This is because the equipment in Soviet tanks (radios, control boxes, relay boxes, sights, etc.) is encased in a thick die-cast aluminium housing. We can safely say that the armoured plate, which appears to be around a centimeter thick, is equivalent to around 30mm of structual aluminium or more.

It is worth mentioning that the inefficient composite construction of Soviet APFSDS rounds like the aforementioned four models makes them exceptionally prone to disintegration and fragmentation after passing through armour plates. Early 105mm APFSDS also relied on composite projectiles, but later on, more efficient long rod ammunition was deployed, and such ammunition would produce much fewer but much more powerful fragments given the same degree of overmatch. So unless the penetrator barely makes it through the armour of the tank, long rod ammunition has a much better chance of penetrating the armour plate around the carousel than composite penetrators, even if the composite penetrator achieves a greater degree of overmatch somehow. All in all, the chances of reaching - let alone igniting - the ammunition in the carousel is rather low, even in the event of a hull penetration. Fragments from a turret penetration would most likely fail to even reach the carousel.

In short, only the T-72 Ural and T-72A use the original autoloader and original carousel with minimal side protection. The T-72B used a different autoloader carousel with revised ammo cassettes in order to fit missiles, and the size of the central hub was reduced in order to fit projectiles that exceeded the length of the ammo cassettes. The armour protection for the carousel was also upgraded by installing a bona fide armoured plate in front of the carousel, behind the driver.

Now, having examined the protection of the autoloader carousel from the front and from above, it is worth examining its protection from the side. There is no additional armour between the carousel and the side armour of the hull, but the side hull armour is split between a thicker upper half and a thinner lower half, and the upper half is lined with a thick layer of "Podboi" anti-radiation lining whereas the lower half is not.

The two drawings below show that the shells in both autoloaders are behind the thinner "tub" armour, but the propellant charges is behind the thickest part of the side hull armour in both designs. The drawing on the left shows a T-64B and the drawing on the right shows a T-72.

The difference in armour protection is difficult to quantify because there are various factors at play. One of the factors is the height of the lower side armour "tub" which is only around 250mm from the floor of the belly to the top of the "tub" on both the T-64 and T-72. Simply put, the lower side armour an exceptionally low target and the tank is not likely to be hit there. The lower side armour is afforded some additional protection by the large diameter roadwheels, but the lower side armour does not have a "Podboi" anti-radiation lining so there is nothing to stop spall.


The carousel rotates independently of the turret. It can rotate to line up new shells at a nominal speed of 70 degrees per second, but as mentioned before, it can only rotate in a counterclockwise direction. This needlessly prolonged the loading cycle in some circumstances, but it is entirely possible to avoid this issue by practicing smart ammo placement. If APFSDS ammunition is stowed to the right of HEAT ammunition, and HEAT ammunition is stowed to the right of HE-Frag ammunition, the time needed to load anti-armour rounds can be greatly reduced at the expense of greatly increasing the time taken to reach the HE-Frag rounds. This way, the gunner can start with APFSDS, and then switch to HEAT without delay when APFSDS is exhausted, or switch to HEAT quickly to deal with IFVs when the high priority tank targets have already been knocked out. Switching to HE-Frag from APFSDS takes longer, but if the target is supposed to be engaged with HE-Frag, then it can be assumed that it is a lower priority threat. In general, sorting the ammunition this way is simply logical, as the time taken to switch ammunition types only increases when switching to ammunition designed for less dangerous threats. In this case, the hierarchy of danger would be: Tank, IFV, and Infantry or other.

One of the techniques developed by a T-64A tank company commander during the 1970's was to load the ammunition in repeating sets of alternating groups so that the time needed for the carousel to reach each round would be equal, and that less time would be spent switching ammunition types. For example:


By doing this, switching from APFSDS to HEAT would take less time than loading the next APFSDS round. This solved the problem of increased loading time when switching ammunition types, but created the additional problem of increasing the time needed to load ammunition of the same type. However, this was considered an acceptable compromise due to the slow carousel rotation speed of the MZ autoloader of the T-64 and T-80 - only 26 degrees per second. It would take an unbearably long time to scroll through large parts of the carousel simply to reach the desired ammunition type. This technique became institutionalized and was a typical method of sorting ammunition among tankers. However, it is not known if T-72 tankers were taught this technique, as it would not have been very useful. The carousel of the AZ autoloader spins almost three times faster than the MZ autoloader, so this problem is much less serious and the flaws of this sorting technique become rather more pronounced. For one, neither the T-64 or the T-72 carry an equal distribution of all three ammunition types, especially not when missiles became a part of their repertoire.

For instance, the generic combat load of a East German T-72M (according to an ex-GDR tank commander) would have 9 APFSDS rounds, 2 HEAT rounds and 11 HE-Frag rounds in the autoloader carousel. It is not possible to arrange these rounds in such a way that the three ammunition types alternate in repeating sets, and it would not be desirable to do so. When engaging tanks, it is much quicker to have the APFSDS rounds grouped together so that the next round is loaded as quickly as possible, allowing the gunner to rapidly fire a potentially decisive second shot. Arranging the ammunition in alternating groups takes away this capability.

Here is a video of a demonstrator autoloader carousel spinning:

In the summer of 1969, a comprehensive test cycle conducted on a number of Object 172 tanks in Central Asia and in the South-Western regions of Russia revealed that the air purification system, engine cooling system, the autoloader and the T-64 suspension had insufficient reliability. These issues were partially eliminated on the subsequent batch of Object 172 tanks. Work on these tanks continued until February 1971, and by then, most of the subsystems in the tank were working within acceptable parameters. The reliability of the autoloader at that point was excellent, having a loading failure rate of only 1 per 448 loading cycles (Baryatinskiy 2010). This roughly corresponded with the barrel life of the 2A26M-2 cannon of 600 EFC. 600 EFC equates to 600 rounds of ammunition with an EFC rating of 1 like HE-Frag or HEAT, but harsher and high pressure APFSDS rounds which erode the barrel quicker have a higher EFC rating of 4 to 5. As such, the rule of thumb is that the autoloader should undergo maintenance or light repair work whenever the gun barrel is in need of replacement. Periodic inspections and testing would greatly benefit the longevity of the autoloader. The newer T-72B autoloader has improved reliability, but the magnitude of the improvement is not known. If troubleshooting is not successful or if individual components cannot be repaired from inside the tank, then the replacement of the entire carousel can be done in the field with the help of an engineering/recovery vehicle like the BREM-1 or at any garage with a hoist large enough to detach the turret. Replacing the rest of the autoloader requires the turret to be partially dismantled.

According to page 17 of "The Soviet T-72 Tank Performance" from 1982, user testing revealed that the autoloader reportedly loaded 3,000 rounds without a malfunction. Since the document uses information from a wide variety of U.S Intelligence sources (apparently also including a T-72M manual) and is not based on actual testing of a T-72 in the possession of the U.S military, this astonishing figure may or may not be correct. Nevertheless, some improvement in the autoloader can be expected since the testing of the Object 172 prototype in February 1971, so the MTBF (Mean Time Before Failure) of the autoloader is likely to be above 448 rounds for a production model T-72 and probably improved with time.

The gunner has a full set of autoloader controls for selecting ammunition to fire, or to replenish the autoloader. In order to fill up the autoloader carousel, the commander uses his autoloader control box to manipulate the autoloader and fill empty ammunition cassettes. According to the manual, reloading the carousel with a full stock of ammunition from an external supply takes 4 to 5 minutes. During the reloading process, the commander remains inside the tank while the driver and gunner pass rounds from outside the tank to the commander through the commander's hatch and the commander sequentially fills the ready ammunition cassettes, as shown in the photo below.

The autoloader carousel of the T-72 can be considered quite easy to load because the position of the ammunition cassettes from the autoloader carousel is always the same. Thus, the task of loading the carousel is repetitive and predictable so that the crew may easily develop a rhythm, unlike some tanks which have multiple ammunition racks in different positions and may require the turret to be turned to certain positions to access certain racks.

The ability to quickly replenish the ammunition of the T-72 was a contributing factor to the success of the "tank carousel" tactic used during the war in Chechnya. Indeed, the quick turnaround time of a T-72 is not a universal trait of all tanks with autoloaders but is actually a byproduct of the straightforward and rugged design of the AZ autoloader, seeing as the T-64A (and its successors) requires 13-15 minutes to replenish its MZ autoloader carousel according to the manual. Of course, the MZ autoloader carries 6 more rounds than the AZ autoloader, but this is not the culprit of the large difference between the two systems: it takes between 28 to 32 seconds to load each round in the MZ autoloader whereas it only takes 11 to 14 seconds for each round in the AZ autoloader; reloading each round in the T-64 takes 2.4 times longer than the T-72.


Aside from the carousel itself, ammunition is stored in racks located throughout the interior of the tank in various nooks and crannies with varying degrees of accessibility. While much of the ammunition in stowed in fairly secure conformal fuel tanks, there are a few rounds of ammunition that are placed out in the open. For the T-72 Ural, 17 rounds are carried in loose stowage. The stowage layout was revised in the T-72A, leading to an increase in the number of shells carried in loose stowage to 22. This enabled the tank to carry two full complements of ammunition into battle and fully replenish the autoloader carousel in the absence of resupply trucks or other sources of ammunition. The stowage layout in the T-72B is slightly modified from the T-72A, allowing an additional cartridge to be carried. The abundance of ammunition stowed outside of the autoloader carousel allowed the T-72 to exceed the ammunition capacity of the T-64A and its successors (37 rounds) despite the lower capacity of the AZ autoloader.

Restocking the T-72 with a full load of ammunition including both the autoloader and the rounds held in loose stowage reportedly takes 13 to 15 minutes. By comparison, the T-64A (and its successors) requires 25 to 27 minutes to do the same despite the much smaller complement of ammunition stowed outside of its autoloader, probably because the rounds are stowed outside the autoloader carousel "basket" and access to the hull from the turret cabin is extremely limited.

The stowage layout for the T-72A is presented in the diagram below (44 rounds in total). The diagram is from a T-72A manual.

The layout for the T-72B is shown in the diagram below (45 rounds in total). The diagram is taken from a T-72B manual. As you can see, the stowage layout is largely identical to the T-72A with only a few small changes. The main difference is the relocation of the ammunition stowed on top of the autoloader carousel behind the gunner's seat to the front left corner of the hull. The ammunition in the front left corner of the hull are easily accessed by the gunner if he is obligated to load the gun manually. Two shells are stowed in racks on the turret wall, on the gunner's side.

Almost all of the propellant charges - the most vulnerable half of the two-part ammunition - are stowed in cylindrical slots inside the conformal fuel tanks. There are twelve slots in the large fuel tank behind the autoloader carousel for propellant charges. Due to the excellent location, the charges are almost completely safe - the carousel would always be in the way instead unless the tank was hit from behind, and it is extremely difficult to hit this fuel tank from above due to the location of the autoloader elevator mechanism and the crew seats. The drawing on the right below shows the propellant charge inside one of the slots. Note that the only the steel casing stub is exposed while the combustible charge casing is fully enclosed by the fuel tank. The drawing on the left shows the arrangement of slots in the fuel tank. The depression at the left hand corner of the fuel tank is made to help accommodate the shells clipped to the side of the hull.

The location of the fuel tank at the rear of the fighting compartment of a tank without the autoloader carousel is shown in the photo below. Note the small red TD-1 flame detection sensor at the lower left corner of the photo, next to the fuel tank.

This stowage method minimizes the risk of immediate deflagration from open flames inside the tank because only the non-flammable steel casing stub is exposed. Of course, this arrangement is not completely fireproof, but it may give the crew enough time to evacuate the tank or extinguish the fire before it becomes too serious. The commander can access the propellant charges stowed in the fuel tank by either swinging the backrest of his seat forward (in the T-72 Ural and T-72A) or by pivoting the backrest of his seat to the side (in the T-72B) as shown in the photo below. The newer pivot arm mechanism on the T-72B allows the commander to stay seated as he access the ammunition at the back of the tank and loads it into the cannon, thus reducing fatigue and possibly increasing the rate of fire of the tank when using manual loading.

The right hull fuel tank on the right hand side of the driver has slots for three propellant charges and four shells plus a single exposed propellant charge stowed in a circular cup at the back of the fuel tank. The right hull fuel tank is shown in the photo on the left (rotated to represent the actual orientation of the fuel tank in the hull) and a cross-section of the propellant charge slot in the fuel tank is shown on the right side. Note that the rim of the propellant charge is not laid flush to the surface of the conformal fuel tank slot unlike the conformal fuel tank at the back of the fighting compartment. This is because the charges are held horizontally so there is very little danger of burning liquid flowing into the slot.

Cross-sections of the slots for the shells in the conformal fuel tank are shown in the diagram below. The slots are designed with the dimensions of HE-Frag shells in mind, so they are large enough to accommodate the two other ammunition types, which were shorter, even the long rod APFSDS ammunition appearing late in the Cold War. The shells are held in place by a simple crescent shaped rotating cover.

The location of these racks makes it convenient for the commander when loading the cannon manually, assuming that the turret is oriented forward or to the left. As you can see in the two photos below, the height of the propellant and shell compartments is just above the autoloader carousel. To use these racks, the commander should turn his TKN-3M/MK periscope to the right, lean forward to pull out a shell, ram it in into the cannon chamber, and repeat the motion to load the propellant charge. The main ergonomic issue with the arrangement is that the large ammo box for the coaxial machine gun is in the way, so the commander has to lean underneath it unless the turret is turned to the left.

More shells and propellant charges are stowed on top of the carousel cover. Some of the propellant charges and shells are clipped to the cover, and others are placed vertically and clipped to the turret ring. There is one position at the 11 o'clock sector of the carousel cover where a single shell can be clipped onto the cover lying down. This shell may obstruct the driver from moving to the gunner's position, or the gunner from pulling the driver out of the tank through the turret.

The circular "ashtrays" at the back of the carousel at either side of the trapdoor are where the shells and propellant charges are placed upright. The shells are secured using clips attached to the turret ring. The "ashtrays" can be seen in the photo below, but the diagrams from the manuals are much more useful. It is shown that two pairs of shells and propellant charges are stowed to the left of the carousel trapdoor, behind the backrest of the commander's seat. Here is one "ashtray" between the autoloader cassette elevator mechanism and the gunner's cannon breech guard. This is the same one as seen in the photo above, to the right of the carousel cover trapdoor. The conformal fuel tank at the rear of the fighting compartment can be seen in the background.

To access the ammunition behind the seats of the gunner and commander, they must scoot forward in their seats and swing the backrest forward. Only then can the shells be unclipped and extricated. Since there is ammunition on both sides of the turret, both the commander and gunner can manually load the cannon if the situation calls for it.

The two pairs of shells and propellant charges stowed on the racks on the carousel cover are located behind the commander's seat. The clips that secure them to the turret ring can be seen in The Challenger's video review of a Czechoslovakian T-72M1 tank. These rounds are very easy to access. The commander only needs to swing or pivot the backrest of his seat out of the way, and then he can directly load the cannon or replenish the autoloader carousel in short order.

The degradation of the propellant charges stowed out in the open atop the carousel cover is reduced by the inclusion of a protective sleeve that fits over the exposed combustible nitrocellulose-impregnated cardboard charge casing just over the metal casing stub. The sleeves are meant to protect the combustible charge from environmental damage, mostly from moisture. The combustible propellant charge casings may swell in highly humid environments, potentially preventing it from fitting inside the cannon chamber. These sleeves also offer a modicum of protection from open flames.

Eight shells can be stowed on the engine compartment bulkhead, on top of the conformal fuel tank behind the autoloader carousel. Three more shells are clipped to the wall on the side of the hull, on the gunner's side. The shells are secured to the ammunition racks using clips. All of the stowage spaces on the engine compartment bulkhead and on the side hull wall are visible in the screenshot below (T-72B type tank hull on display). The gunner can easily access the shells clipped to the side of the hull, but only if the turret is turned slightly to the left.

The drawing below shows the racks. These racks can accommodate all ammunition types, including missiles as the drawing on the right shows (taken from a T-72B manual).

The commander can freely access the shells clipped to the wall at the back of the fighting compartment. Thanks to the lack of a turret basket on the T-72 turret, the commander can simply swing the backrest of his seat out of the way, unclip one of the shells, and then bring it up to the cannon to ram it in. He can then lean down and extract one of the propellant charges from the conformal fuel tank behind the autoloader carousel. The ease of accessing the ammunition in loose stowage heavily contributes to the relatively high rate of fire when loading manually and also reduces the time needed to replenish the autoloader carousel.

The screenshot above gives us a good view of the ammunition from the driver's perspective, so while it may appear that a shell penetrating the hull armour would seriously jeopardize the ammo, this is not the case. Due to the highly cluttered fighting compartment and the very large distance from the upper glacis armour to the ammunition mounted to the wall (more than two meters), the ammunition has a very good chance of avoiding any damage whatsoever. The photo below, for example, is the same view taken from the same angle, but it is clear that the engine compartment bulkhead is completely obscured behind the stabilizer components underneath the cannon and behind the seats of the commander and gunner. If the tank hull is penetrated from the front, the tank will likely be knocked out by a firepower kill via damage to the stabilizer or some other component of the gun control system, but an immediate ammunition detonation may be avoided and the crew may survive.

Although the loose ammunition stowed in the turret and hull is quite vulnerable, it is important to recognize the fact that some attention was paid to their protection. Most of the vulnerable propellant charges are held inside the conformal fuel tanks and those that are stowed in the open are at least protected by a fire retardant sleeve. These measures offer some protection from open flames. The projectiles held in loose stowage are not protected by similar measures, but they are already fire resistant to a large extent with the exception of the APFSDS rounds.

Furthermore, it is also clear that the ease of access to the ammunition held in loose stowage was a major ergonomic consideration during the design of the tank. The location of these rounds and the amenities provided to access them facilitate the speedy loading of the cannon and replenishment of the autoloader carousel. In the event that the commander is incapacitated or dead, the ammunition on the gunner's side of the turret is laid out in such a way that he could also load the gun manually at a reasonable speed in an emergency.


All in all, there can be between 17 to 23 additional cartridges stowed outside the carousel for a total of between 39 to 45 rounds of ammunition depending on the specific model of T-72. However, in practice, crews tend to ignore certain spaces such as the shell stowage rack on top of the carousel cover (as seen above), and some crews may decide not to have any ammunition in loose stowage at all, so the actual sum total of loosely stowed ammunition can be anywhere from 22 to none. Nevertheless, from a design standpoint, the fact that the T-72 can have a total ammunition capacity of 45 rounds when the older T-54/55 with a 100mm cannon had only 43 and the T-62 with a 115mm cannon had just 40 is a highly noteworthy achievement and a grand step forward in tank design efficiency. Compared to the Leopard 2 (42 rounds) and M1A1 Abrams (40 rounds), the T-72 carries more ammunition overall for a cannon of a similar caliber. Considering how many online tank enthusiasts espouse the higher ammunition capacity of early Cold-War era NATO tanks as decisive advantages over the Soviet T-54 and T-62, the fact that the T-72 carries more ammunition than its contemporaries is curiously ignored. 

Surprising as it may be, the T-72 also carries more ready ammunition than its two most modern NATO counterparts: 22 in the carousel compared to 18 and 15 in the bustle ready racks of the M1A1 Abrams and Leopard 2 respectively. The ammunition capacity of the autoloader carousel is the same as the autoloader bustle of the Leclerc, and the Challenger 2 carries slightly more ready rounds (25). Generally speaking, this is not an issue for any of these tanks because it is rare for a tank to expend so much ammunition in a single engagement. There is typically a lull in the fighting, which is when the loader in any tank would take the time to replenish his ready racks from the less convenient stowage racks. In the case of the T-72, the commander and gunner will replenish the carousel using the loose ammunition stowed on board the tank in the turret and hull. In an Abrams, the commander will open the bustle door on his side of the turret and pass cartridges over to the loader who will then stow them in his ready bustle racks. In the Leopard 2, the loader will take ammunition from the front hull racks and stow them in the bustle ready racks.

However, ammunition carried in loose stowage can be a huge liability in battle as it has been proven to be the main cause of irrecoverable or catastrophic tank losses. Some of the loose ammunition stowed in the hull is still somewhat secure, but the ammunition in the turret constitutes a significant risk to the survival of the crew. As the diagram below shows (diagram taken from Tank-Net), only 2% of shots land at a height of one meter from the ground in the 60 degree frontal arc of a tank. This is good news for the carousel autoloader, but the diagram shows that 65% of shots hit the turret. As such, the benefits of the low placement of the carousel may be completely undone by loose ammunition stowed in the turret. 

However, it should be understood that the distribution of hits fluctuated somewhat over the years in various conflicts. In the Second World War, the majority of hits sustained by tanks were on the hull. It is commonly thought that this was due to the fact that the hulls of the tanks of the era tended to be much larger than their turrets and also quite tall due to the placement of the transmission at the front such as on the Panther and M4 Sherman. Later on, combat in Korea and in the Middle East showed that more hits were being taken on the turret than on the hull, creating a more even distribution between the turret and hull. It is worth noting that the average combat distance in Korea was very short - only a few hundred meters - due to the nature of the terrain. Later on, it was observed by Dr. Manfred Held that in Kuwait during Operation Desert Storm (ODS), the vast majority of shots landed on the turret. The diagram below, taken from "The Main Battle Tank of Russia: A Frank Conversation About The Problem of Tank Building", shows the distribution of hits in the vertical plane by percentage for the 1967 Six-day war, the 1983 Yom Kippur war and ODS in 1991.

The three black bars in the diagram indicate (from top to bottom) the bottom of the turret, the belly of the tank, and ground level. The bottom of the turret - the turret ring - is considered to be 1.5 meters above ground level, and the belly of the tank is considered to be around 0.5 meters from ground level. The hull is therefore considered to be around 1 meter tall. This is a representation of a Soviet main battle tank like the T-72, T-64 or T-80. All of these tanks have a ground clearance of 0.48 meters, a hull with a height of 1.0 meters, and a turret ring located 1.48 meters above ground level. Considering that the AZ autoloader carousel has a total height of 450mm (including armoured false floor), the total height of the carousel from ground level is 940mm.

As you can see, even though the distribution of hits was not entirely consistent across the three conflicts, the fact that the lower half of the hull (between 0.5 to 1.0 meters) statistically sustained the fewest hits was universally true for all cases, and therefore, the autoloader carousel would sustain very few direct hits. The turret ring sustained the greatest number of hits statistically, which makes sense as the turret ring is the center mass of any tank. From this, it can be seen that the majority of damage inflicted to the carousel would be from projectile fragments or secondary fragments coming from above - from the upper sections of the upper glacis or from the turret. These fragments would be attenuated by the "Podboi" anti-radiation liner on the upper glacis which has a thickness of 50mm on the upper glacis and 20mm on the turret. The air gap between the surfaces of the upper glacis and the turret further reduces the energy of fragments.

Furthermore, it can be seen that omitting some of the ammunition carried in the turret appears to be beneficial to the survival of the tank in the event that the turret armour is pierced. As mentioned before, this was widely practiced by Russian tank crews during the war in Chechnya.

Another noteworthy aspect to consider is the mine resistance of the autoloader carousel. The carousel of the AZ autoloader is mounted to the floor whereas the carousel of the MZ autoloader of the T-64 and T-80 is suspended from the turret and the autoloader itself is housed in the turret cabin. Only the rotary power supply unit connects the turret cabin to the hull floor. The extent of the difference in durability is difficult to determine, but it seems unlikely that a mine blast underneath the tank will disable the autoloader as a conventional anti-tank mine with a tilt-rod fuse would detonate under the driver's station and an anti-tank mine with a simple pressure fuse would detonate under the tracks at the first roadwheel. The shock from the explosion may potentially be a source of failure, but it would only be fair to point out that the power supply unit (for any tank) may be damaged as well as it is mounted to the belly. In that case, the power supply to the turret would be cut off and the loss of autoloader function becomes a much less serious issue by comparison.

If the autoloader elevator malfunctions, it is still possible to operate the elevator mechanism manually using a crank wheel (pictured). The commander and the gunner must take turns to load the cannon depending on the location of the ammunition in loose stowage. The commander would obviously load the cannon using ammunition close to him, and vice versa. Some of the ammunition requires the turret to be oriented in a specific direction to access. The T-72A manual has a full table detailing the locations of the ammunition, the orientation of the turret needed to access it, and whose responsibility it is to load that ammunition. The benchmark time for a complete manual loading cycle is 26 to 30 seconds. The corresponding time for a T-64A as dictated in a manual is 1 minutes 40 seconds for the first shot and 1 minute for subsequent shots. In other words, a T-72 is able to sustain a rate of fire of around 2 rounds per minute with manual loading whereas a T-64 would struggle to attain a rate of fire of 1 round per minute. However, according to official Soviet norms (battle drills), the minimum acceptable times for loading in a T-72, T-64, and T-80 are all the same. The "minimum" grade for manual loading using ammunition from the ammunition carousel is for 1 minute, the "good" grade is for 50 seconds and the "excellent" grade is for 45 seconds.

If the carousel fails, it is possible to manually crank the carousel and access the ammunition inside using a cranking lever located underneath the commander's seat. However, the commander would have no idea where the desired ammunition type is located in the carousel, so it may be more feasible to simply use the ammunition in loose stowage.

Both the gunner and commander have free access to the autoloader from their respective stations. As such, it is possible for the crew to troubleshoot the device without leaving the tank or requiring the separation of the turret from the hull. The old Soviet promotional image below shows a tanker in the gunner's seat with his hands on the autoloader elevator and the stub ejection mechanism.

Manual loading is something to be done in emergencies only, not only because it is much slower than normal automated loading, but because it also forces one of the two crew members to abandon his usual duties. In reality, autoloader failures are exceedingly rare (but not non-existent), so there is little need to worry about manual loading. The propensity for autoloaders to malfunction either from wear and tear or from a knock on the turret is greatly exaggerated.


There are four main types of ammunition for the 125mm gun. A typical loadout for a general infantry support mission would see that HE-Frag shells are loaded in large quantities, for example, while more HEAT and APFSDS shells would be loaded for ambushes where armoured vehicles are expected. A standard ammunition load would comprise of an equal mix of high explosive rounds (HE-Frag) and anti-armour rounds (APFSDS, HEAT) with APFSDS rounds being much more numerous than HEAT rounds. The ammunition load in the autoloader carousel would weigh around 564 kg.



125mm ammunition for the D-81 gun series is split into two parts: propellant and projectile. Each propellant charge is contained within a thin TNT-impregnated pyroxylin-cellulose (plastic) casing that is consumed upon firing, and the entire assembly is embedded into a steel casing stub shaped like a cup, much like a shotgun shell. All of the propellant charges have a total length of 408mm.

The steel casing stub is the only part of the cartridge assembly left intact after firing. In the T-64 and T-80 series, this stub is returned to the autoloader mechanism by a mechanical arm, but in the T-72, this stub is ejected from the tank via a small hatch at the rear of the turret roof. The diameter of the casing stub is 138mm and the diameter of the rim is 150mm. The length of the casing stub is 140mm. It weighs 3.45 kg. The purpose of this item is to obturate the breech and ensure a complete seal so that no flashback occurs.

The use of a relatively large steel casing stub as opposed to a small primer unit like on the bagged charges of the L11 cannon or the much smaller casing stub on the cartridges for the Rh 120 L/44 and M256 cannon is a very minor design flaw due to the fact that the internal surface area of the casing stub allows a relatively large quantity of incompletely-burnt propellant residue to linger. This smoldering residue becomes a minor source of unpleasant fumes until the casing stub is ejected from the turret. The weight of the casing stub also needlessly increases the mass of the propellant charge, contributing to the ~10 kg weight of each complete propellant charge without any tangible benefits.

The GUV-7 electric/percussion primer is used in all of the three propellant charges designed for the D-81, giving the option to either fire the shell normally using the fire controls on the gunner's hand grips, or using the button on the manual traverse flywheel, or using the manual lever-operated striker pin incorporated into the breech block of the cannon.


Original propellant charge designed for the D-81, used since the T-64A. It uses the 15/1TR VA propellant compound. The mass of the propellant charge itself is 5.66 kg, while the steel casing stub weighs 3.45 kg. The remainder is taken up by the combustible plastic casing and the primer.

Mass of Complete Assembly: 10 kg
Propellant Charge mass: 5.66 kg


Newer general-purpose propellant charge. It uses 12/7 VA propellant compound. 4Zh52 is completely interchangeable with 4Zh40.

This model of propellant charge has completely replaced the Zh40 in frontline use. Here is a video of the Zh52 propellant charge being opened up (link). Nowadays, HE-Frag and HEAT rounds are fired almost exclusively with Zh52 except during live fire exercises in order to use up obsolete charges.

Mass of Complete Assembly: 10 kg
Propellant Charge mass: 5.786 kg


High-energy propellant charge to launch APFSDS shells at a greater muzzle energy than possible with previous propellant charges. It uses 16/1TR VA propellant compound. It is used with newer APFSDS rounds which also contain 16/1TR VA propellant in the incremental charge, but it seems that there is nothing to stop it from being used with older APFSDS rounds. It would presumably propel older APFSDS at a muzzle velocity far in excess of 1,800 m/s.

Mass of Complete Assembly: 10 kg
Propellant Charge mass: 5.3kg



Two part superquick, distance armed piezoelectric fuse. Point-detonating design that has provisions for graze initiation to allow detonation despite steep angles of incidence. It is distance-armed by inertia at a distance of 2.5 meters from the muzzle.


The V-429E fuze is point-detonating, distance armed and with variable sensitivity settings. It has two settings - superquick and delayed. The superquick setting detonates the shell with a 0.027 second delay and the delayed setting detonates the shell at 0.063 seconds. Superquick action guarantees reliable detonation in snowy or swampy ground, and delayed action gives a small time allowance for the shell to penetrate its target before detonating. The shell is set to the Fragmentation mode when the fuze is set to the "O" position. HE mode is set when the fuze is set to the "O" position but the safety cap is left on. Delayed HE or "bunker busting" mode is set when the fuze is set to the  "З" (a Cyrillic "Z") position, and the safety cap is left on. The additional delay enables the shell to penetrate more deeply into hardened targets.

The fuze is armed by inertia; the shell experiences a momentary braking effect from the unfolding of the stabilizer fins 5 to 20 meters from the muzzle, and this is used to arm the fuze.


The V-429V fuze is an updated version of the V-429E fuze. The safety cap has been replaced with a safety pin with a protruding ribbon. To deactivate the safety "cap", the ribbon is pulled to tug the pin out. This is much faster than unscrewing the old safety cap.


The T-72 normally carries 12 HE-Frag shells in the autoloader, although this will almost certainly vary by situation. These shells have traditionally been predominant in Soviet armoured tactics, where tanks were regarded as the tip of the spear during breakthroughs. Bunkers, ATGM teams and troop concentrations - not tanks - were the bane of any and all armoured targets, and thus became high priority targets. HE-Frag shells were therefore a critical part of the combat ammunition loadout of the T-72.

The V-429E fuse gives 125mm HE-Frag shells a great deal of flexibility. When attacking infantry in the open or in covered positions, such as anti-tank teams, advancing troops, or machine gun nests, the fuze should be set in the "superquick" mode, giving it a delay of 0.027 seconds to ensure that the shell will detonate instantly upon meeting soft ground like mud and snow, allowing it to exploit its thick steel shell to its fullest as shrapnel.

When attacking reinforced concrete targets like bunkers and pill boxes, the shell should be set in the "delayed" mode, giving it a delay of 0.063 seconds, allowing the shell with its thick steel casing to travel a fair distance into target material before detonating. This is great for bunker busting because the impact of the big, heavy shell creates fractures, cracks and fault lines in concrete, making it easier for the explosive charge to shatter and blow apart the entire structure. If targeting non-hardened buildings like houses, the shell could pass through cinder block or brick walls and explode on the other side of the wall. It is also possible to fire a shell in the delayed mode at the low angle of incidence with the ground at the target area in order to skip the shell, causing to explode above ground level and produce an airburst effect.

With that in mind, HE-Frag rounds should not be mistaken as a purely anti-infantry or anti-structure munition as they may even be used as a substitute to more specialized anti-armour shells like APFSDS and HEAT against heavy armour under certain circumstances, such as when all other ammunition has run out, or if effective destruction cannot be achieved by other means. A direct hit with the fuse set to the superquick setting will likely result in the debilitating disability of the cannon or the destruction of aiming devices, the destruction of the driver's vision blocks or the destruction of the tank suspension (although probably not all at the same time), producing a firepower and mobility kill. In many cases, the driver of a modern tank has an unsettlingly high probability of being killed or at least severely injured by a hit to the turret or glacis due to insufficient blast attenuation. The explosion of a large caliber HE round on the turret ring will most certainly send spall and fragments shooting down into the driver's neck through the thin hull roof. The T-72 is very capable of inflicting this type of damage on most foreign tanks, which often do not have spall liners. This makes it exceptionally easy for a 125mm HE-Frag shell to kill, maim, and injure the crew behind the armour of all-steel tanks like the M60, Chieftain, Leopard 1, AMX 30, and so on. However, modern tanks sporting composite armour arrays and spall liners may not be vulnerable to the same degree.

When set in the HE mode, 125mm shells are extremely deadly to lightly armoured vehicles. For example, 76mm HE shells fired from the F-34 cannon of a T-34 were also found to be able to perforate the armour of tanks like the Pz.III. There is comparatively little information available on the internet on this topic, but Peter Samsonov's translation of a report on the effects of 76mm HE-Frag shells at tanks with a variable fuse is especially enlightening. Here is a fascinating paragraph from that report:

"When firing [76mm] HE shells from mod. 1931 guns consider that they can penetrate 45 mm of armour at 30 degrees from 500 meters, and 50 mm of armour under the same conditions can be penetrated from 300 meters or closer."

Like any other anti-aircraft gun intended for engaging high altitude targets, the mod. 1931 gun fired relatively high velocity rounds (~800 m/s), but such velocities are pedestrian for modern guns like the 2A46. It is also worth noting that the shape of HE shells is ogived, so the efficiency of such shells on sloped armour is very low compared to a blunt-nosed APBC shell like the 100mm BR-412B. If 76mm HE shells are capable of defeating 45mm of armour plate angled at 30 degrees at 500 meters, a much larger 125mm HE-Frag shell travelling at the same velocity would be able to penetrate much more or at least achieve a more destructive effect. Indeed, Hungarian testing of 125mm HE rounds on T-34-85 tanks showed that frontal hits could perforate the upper glacis armour and detonate inside the tank with catastrophic effects as shown in the photos below.

Interestingly enough, the 0.063-second delay of the V-429E fuse is identical to the delay of other post-War fuses like the RGM-6 point-detonating fuse for 100mm HE-Frag shells, and this delay is longer than what WWII and pre-war fuses for artillery rounds provided, which was 0.03-0.05 seconds. This was most likely related to the increased standards of bunker fortifications and the disappearance of concrete piercing rounds from the arsenal of the Red Army shortly after WWII, but it is also possible that the increased delay was deliberately aimed at increasing the anti-armour capabilities of high velocity tank-fired HE-Frag shells.

The side armour of some NATO tanks like the Chieftain and the Leopard 1 are probably vulnerable to 125mm HE shells from combat distances (1.5 km or more) and it is likely that the thicker side hull armour of an M60A1 may not be enough to resist 125mm HE at closer distances. Thin side skirts may offer too little resistance to set off the V-429E fuse, even steel skirts which are found on tanks like the Centurion and Chieftain. The Leopard 1 is noteworthy as the armour on the side of its turret is only 40mm thick (angled at 30 degrees), so even the aforementioned 76mm HE shell fired from the 1931 anti-aircraft gun may potentially defeat the Leopard 1 from 500 meters. The addition of a 30mm spaced appliqué plate on the turret in later Leopard 1 variants might still not be enough to defend it from a 125mm HE shell, and even if the shell was successfully stopped, the explosion might still be powerful enough to split open the base armour plate. Against a heavily armoured IFV such as the M2A2 or M2A3 Bradley, the 38mm layer of steel appliqué armour on the hull will be entirely insufficient to stop such a shell even at long distances. Modern IFVs designed with relatively heavy armour to combat the ubiquitous RPG are probably still extremely vulnerable to 125mm HE.

As such, even though the T-72 carries more HE-Frag shells than anti-armour shells, it can be seen that this is not a problem as HE-Frag shells have a very substantial multi-role capability. It is not wrong to say that the combat value of an individual HE-Frag round easily exceeds that of an APFSDS round or a HEAT round.

As a side note, it is interesting to note that HE-Frag shells are relatively kind to gun barrels. They have an EFC rating of 1, meaning that if a barrel was rated for 1,000 EFC, it would be able to fire 1,000 HE-Frag shells before bore erosion or other factors render it unsafe or otherwise unfit for use.


Regular shell with copper driving bands. The shell has the shape of an ogive. It is interesting to note that this shell has a length of 675mm, making it the longest projectile among the three ammunition types carried by the T-72. This only changed with the inclusion of guided missiles in the T-72B, and high elongation long rod projectiles later on.

Complete Shell Mass: 23 kg
Complete Shell Length: 675mm
Wingspan (deployed): 356mm
Muzzle velocity: 850 m/s

Explosive mass: 3.148 kg
Explosive composition: TNT

It's worth noting that TNT is a relatively sensitive explosive compound. The risk of an ammo detonation is significantly higher if these shells are present.


Improved HE-Frag shell with compressed explosive charge of a different composition designed to provide added incendiary effect. The use of compressed explosives meant that the density of the explosive charge could be increased and thus increase the explosive power of the shell. This shell uses plastic driving bands instead of copper ones in an effort to reduce barrel wear.

Maximum Chamber Pressure: 3432 bar

Total Length: 676mm
Total Shell Mass: 23.3 kg
Muzzle velocity: 850 m/s

Explosive mass: 3.4 kg
Explosive composition: A-IX-2 (Phlegmatized RDX + Aluminium filings)

A-IX-2 is much less sensitive than TNT. The risk of ammo detonation is much lower if these shells are stowed.

Practice HE-Frag

Practice HE-Frag shell that emulates the ballistic characteristics of live HE-Frag shells. Contains a 200-gram TNT charge to produce a bright flash that acts as a visual hit marker for the trainee gunner.

Maximum Chamber Pressure: 3432 bar

Total Length: 676mm

Total Shell Mass: 23.3 kg
Muzzle velocity: 850 m/s


The T-72 carries a few HEAT rounds in the autoloader carousel for its flexibility. They are powerful enough to pierce contemporary armour in most cases and the explosive charge allows them to be used against light or unarmoured vehicles with a much better result than APFSDS shells. HEAT shells may also be used against hardened concrete bunkers or simple earthen fortifications with good results, and it is entirely feasible to engage personnel owing to the very thick steel case containing the charge.

Against thickly armoured targets, HEAT shells produce deep but small holes. The secondary methods of destruction aside from the cumulative jet itself (which is the primary one) is the blast of the explosion of expanding gasses rushing through the hole in the armour, the flash of heat (capable of causing flash burns) and the spray of high velocity fragments of armour and shaped charge material following perforation, which can set internal equipment alight and injure the crew. It is difficult to kill crew members without a direct hit by the shaped charge jet unless there is a very significant armour overmatch, forcing HEAT shells to rely mostly on knocking out essential equipment or causing internal fires. But still, due to the enclosed nature of tanks, there is a high likelihood of striking at least one crew member if one could score a hit on the occupied sections of the tank. Outside the tank, the blast and shrapnel produced from the explosive charge and thick casing of the warhead can kill or injure dismounted infantry and external equipment, including periscopes, sights, and so on.

HEAT shells also retain a characteristic advantage over APFSDS shells in that they wear down the barrel at a greatly reduced rate. One HEAT shell is only equivalent to one EFC, whereas an APFSDS shell can be equivalent to 3, 5 or even 7 EFC. This makes them the preferred choice of training ammunition during live fire exercises, besides HE-Frag shells. Training with APFSDS is not held quite as often, as scoring a hit with hypervelocity shells is obviously not quite as challenging as doing the same with shells that are travelling at almost half the speed. HEAT ammunition is also more expendable than APFSDS ammunition during live fire exercises, as it is now almost entirely useless against modern tank armour.

Vasily Fofanov's old website popularized the thinking that Soviet HEAT ammunition was more accurate than their APFSDS ammunition. One paragraph in particular has been frequently quoted:

"HEAT-FS rounds were also substantially more accurate than APFSDS (which might also be surprising to a Western reader). This is reflected in the Soviet deviation criterion, which was more strict for HEAT rounds (0.21 mil) than for APFSDS rounds (0.25 mil). However, in practice HEAT-FS rounds were even more accurate. As control trials of a random mass-production T-64A held in the 70s (the details of which were made available to the author) indicated, while APFSDS rounds hugged the outer bounds of acceptance criterion, HEAT-FS rounds actually demonstrated the average deviation of well under 0.1 mil!"

This may or may not be true, since Fofanov has admitted that the information presented in his website is mostly outdated and unreliable. Thankfully, there are more solid sources for our perusal. According to firing tables for 3BK-14M provided by Stefan Kotsch, the HEAT shell has a horizontal deviation of 0.19 m and a vertical deviation of 0.19 m. On the other hand, the firing tables for 3BM-15 also provided by Stefan Kotsch show that the APFSDS shell has a horizontal dispersion of 0.20 m and a vertical dispersion of 0.20 m - the difference in dispersion is only a centimeter. The gap in probable deviations remains minor even at 2 km. At that distance, 3BK-14M has a horizontal dispersion of 0.38 m and a vertical dispersion of 0.39 m, whereas 3BM-15 has a horizontal dispersion of 0.4 m and a vertical dispersion of 0.4 m - only a centimeter or two. Evidently, there is some truth to the claim that Soviet HEAT ammunition was more accurate, but the enormity of the gap between the two types has clearly been highly exaggerated. Furthermore, these firing tables only represent the mechanical accuracy of the projectile and the cannon it is fired from, and does not represent the accuracy of the entire weapon system, including the aiming and gun laying devices.

In practical real world conditions, the expected hit probability of HEAT ammunition is vastly lower than that of APFSDS ammunition for a variety of factors as a rule. This has been confirmed by Soviet studies on the hit probabilities for T-64 and T-72 tanks during live firing exercises. The most major factor in the advantage held by APFSDS rounds is the interference of crosswinds and head or tail winds, and another factor to consider is ranging errors, particularly at longer distances. The high velocity of APFSDS ammunition (3BM-15 has twice the muzzle velocity of 3BK-14M) also makes it much easier to score hits on moving targets at any distance. Furthermore, the accuracy advantage of HEAT ammunition inevitably declined when sabots made with tighter tolerances and more reliable petal separation became available.


Wave Shaper: Object or device that infleunces the propagation of blast waves in a way that is beneficial to jet formation. Typically composed of an inert material with low sound speed.

A-1X-1: Phlegmatized RDX, consisting of 96% RDX and 4% wax.

OKFOL: Explosive compound composed of 75% HMX and 25% RDX.

Standoff Probe: Extended structure to increase the distance between the shaped charge cone and the target material, i.e, standoff.

Explosive Pressing: The process of increasing the density of explosive compounds by high-pressure mould pressing. The result is more explosive mass per volume, translating to more energy.

All of the information presented below are backed by either photographic or videographic evidence, or official documentation.


3BK-12, 3BK-12M

First 125mm HEAT shell, originally for complementing the T-64. By the time the T-72 emerged, 3BK-12 had long been replaced by the 3BK-14 but large quantities were probably stocked and continued to be stocked. The 3BK-12 is the low cost veriant with a steel shaped charge liner, and the 3BK-12M is the more expensive variant with a copper liner. ("M" stands for "med", which means "copper" in Russian). The use of a copper liner grants improved penetration performance, but at a slightly higher price.  The shell is characterized by the rather thin walls of the standoff probe, straight standoff probe, and a "house" shaped wave liner, as seen in the diagram above.

Projectile weight: 19.8 kg
Total Projectile Length: 678mm

Muzzle velocity: 905 m/s

Explosive Charge: A-1X-1
Explosive Charge Weight: 1,760g

Shaped Charge Cone material: Steel
Shaped Charge Cone diameter: 105mm

Penetration (at all distances):
420mm RHA

The purpose of the "teeth" around the edges of the projectile surrounding the standoff probe is not known. It is possible that they are designed to ensure that the projectile is smoothly chambered when it is rammed into the gun breech.


3BK-14, 3BK-14M

Updated HEAT shell with similar dimensions as the BK-12, but with minor internal differences. It is characterized by distinct knurls around the top edge of the main body surrounding the standoff post, possibly to enable the shell to fuse even on extremely high obliquity impacts by tilting the tip towards the armour plate on a glancing blow. This shell uses a cylindrical wave shaper with a slight taper, and the standoff probe now has a slight taper. As before, the "M" suffix for BK-14 denotes that it has a copper shaped charge liner.

3BK-14 and 3BK-14M are noteworthy for being the most advanced HEAT round supplied to client nations operating T-72 tanks as well, including East Germany. As such, it has proliferated more than any other HEAT round that entered service in the USSR.

Maximum Chamber Pressure: 2,900 bar

Projectile Weight: 19.8 kg
Total Projectile Length: 678mm
Warhead Casing Length: 296mm

Muzzle velocity: 905 m/s

Explosive Charge: OKFOL
Explosive Charge Weight: 1.76 kg

Shaped Charge Cone material: Steel
Shaped Charge Cone diameter: 105mm
Shaped Charge Cone angle: 36°
Shell wall thickness: Tapering from 7mm (front) to 17.5mm (base)

Standoff probe diameter: 65mm tapering to 44mm
Standoff probe length: 206mm

Penetration (BK-14):
450mm RHA at 0°

Penetration (BK-14M):
480mm RHA at 0°

Of course, it must be noted that the penetration for BK-14M is most definitely a very conservative value that may not necessarily even represent the average penetration of the shell. Considering that the 115mm BK-4M shell fired from the U-5TS gun of the T-62 has an average penetration of 499mm, a minimum penetration of 418mm and a maximum penetration of 559mm, the performance of BK-14M cannot be lower.

Data on the fragmentation of BK-14M can be found in "Осколочное Действие Кумулятивных И Осколочно-Фугасных Снарядов При Взрыве На Броне Танка" ("Fragmentation of Cumulative and High-explosive Explosive Shells during the Explosion on Tank Armour") by Yu. A. Mikheev. Scans of the entire article as published in the "Вестник Бронетанковой Техники" ("Bulletin of Armour Technology") specialized magazine are available on Andrei Tarasenko's blog. The table below contains the relevant data:

The detonation of the BK-14M warhead produces an average of 500 fragments in a forward cone of 38-47 degrees. 10% of the fragments are capable of perforating an aluminium plate with a thickness of 60mm, which translates to an average of around fifty fragments. It should be noted that the side armour of the aluminium-hulled of an M113 tracked APC only has a thickness of 38mm. Thus, there would be a great deal of armour overmatch. As such, the use of a BK-14M shell on a typical APC or IFV target would not only guarantee the penetration of its armour by shaped charge attack, but also inflict heavy damage on its internal components via the fifty heavy steel fragments sprayed in a relatively wide angle.


3BK-18, 3BK-18M

The BK-18 is a further improved design. It is visually identical to previous designs, but differs in that it features an aluminium shaped charge cone. The fragmentation effect of BK-18 is similar to BK-14M.

Unlike the lightly tapered wave shaper of the BK-14, it has a cylindrical one, which coincides with the usage of a different cone material with different physical properties. Like its predecessors, it has distinct knurls around the top edge of the main body.

The BK-18M is a variant of the BK-18 using a copper cone. Both models are very widespread in current Russian Army stocks.

The BK-18 is characterized by its thickened walls, both for the warhead casing and for the standoff probe. This presumably translates to a significant improvement in the anti-personnel capabilities of the shell compared to earlier designs, and the more robust standoff probe may be beneficial for other reasons.

Maximum Chamber Pressure: 2,900 bar

Total Length: 678mm
Projectile Weight: 19.8 kg
Muzzle velocity: 905 m/s

Explosive Charge: OKFOL
Explosive Charge Weight: 1,760 g

Shaped Charge Cone material: Aluminium
Shaped Charge Cone diameter: 105mm
Shaped Charge Cone angle: 36°
Shell wall thickness: Tapering from 7mm (front) to 17.5mm (base)

Standoff probe diameter: 65mm tapering to 45mm
Standoff probe wall thickness: 7.5mm

Penetration (3BK-18):
500mm RHA at 0°

Penetration (3BK-18M):
550mm at 0°



Relatively recent (late 80's) shell with tandem warhead configuration primarily to aid in penetration of complex armour arrays and to defeat ERA-equipped targets. Despite being heavier than its single-charge predecessors, it somehow travels slightly faster.

The precurser shaped charge is located halfway down the standoff probe and may be rightfully considered a fully-fledged warhead all on its own, having a considerable explosive charge and complete with its own standoff accounted for. The use of a precursor warhead makes it effective against the special armour of NATO tanks from the early 80's, prior to the implementation of exceptionally heavy armour in the later variants of the Leopard 2A4 and M1A1.

This shell is characterized by the lack of "teeth" on the front edges of the primary warhead case, and the new fuse, which is fully conical in shape. The shell uses a hemispherical wave shaper. Both charges have base fuzes.

The 3BK-29M is same shell as its parent, but with a copper liner. Whether both the precursor and the main charge use copper is unclear, but it is likely.

Projectile Weight: >20 kg
Muzzle velocity: 915 m/s

Explosive Charge: A-1X-1
Explosive Charge Weight: (?)

Precurser Explosive Charge: A-1X-1
Precurser Charge Cone material: (?)
Precurser Charge Cone diameter: 40mm

Primary charge penetration (without precurser/after reactive armour): ~620mm at 0°
Primary charge penetration (after precurser/without reactive armour): ~820mm at 0°

BK-29 should be considered the most important development in tank cannon-fired HEAT ammunition in recent times, as it is the only example of a tandem warhead HEAT tank shell in service. Now that we know of the armour composition of the M1 Abrams, we can speculate on how BK-29 would fare against it. One of the things that we can be certain about is that the front armour plate protecting the NERA array is relatively thin. The precursor charge of BK-29 could probably punch through the armour plate, enter and leave holes in several of the NERA plates and leave an open channel through which the main warhead can then pass through and defeat the main armour, since the precursor and the main charge are coaxial. With the extra spacing created by the large and mostly hollow armour array, the main charge could easily reach its optimum standoff distance and reach its maximum penetration power. Following this train of thought, it is probable that BK-29 can defeat the front hull armour of the M1A1 Abrams, but probably not the turret, since the thickness of the new turret composite armour was substantially increased over the M1.

However, it will not be able to defeat the turret cheeks of a T-72B. The Kontakt-1 boxes that encapsulate the composite armour of the T-72B will reduce the effect of the precursor charge of BK-29 (and indeed, any tandem warhead projectile) so much that it will be easily stopped by the relatively thick front wall of the armour cavity in the turret, meaning that the primary warhead will still have to deal with the untouched NERA armour contained within, and the warhead will most probably fail to defeat the entire array.



Enigmatic and ingeniously designed triple-charge HEAT shell. It is probably not in service at present. It can penetrate 800mm of steel armour with a hardness of probably about 280 BHN, as demonstrated by a cutaway.

Total length: 665mm

Penetration: 800mm RHA (No reactive armour)

From Vasily Fofanov's website

Practice rounds


Single-charge inert HEAT warhead designed to exactly emulate the ballistic trajectories of the 3BK-14 and 3BK-18 shells. There is a 200-gram squib inside the warhead that acts as a visual hit marker for the trainee gunner.

Total Length: 678mm
Projectile Weight: 19.8 kg
Muzzle velocity: 905 m/s


Training round imitating the exact flight characteristics of the 3BK-29 shell.


Despite pioneering APFSDS shells with the introduction of the 2A20 115mm smoothbore gun, the Soviet Union never had the technology to mass produce true long rod tungsten or depleted uranium projectiles until the mid-80's, whereas the Americans had already fielded the M774 DU APFSDS round since the mid to late 70's. Even then, the long rod APFSDS rounds fielded by the Red Army had sheathed penetrators when American APFSDS rounds were already monobloc. Their best APFSDS rounds were composites composed of steel projectiles with small embedded cores. This type of composite shell was incredibly economical due to the very small quantity of tungsten used in each projectile, but it was limited in scope and growth potential. The last and most optimal composite APFSDS shell created on the basis of the subcaliber tungsten carbide core philosophy entered service in 1983. Long rod shells made with tungsten and depleted uranium alloys entered service just two years later after protracted development and troubleshooting as a direct response to intelligence on new Western multi-layered composite armour.

During the past few decades, modern Russia has fielded several long rod APFSDS rounds while retaining a large stock of older composite APFSDS ammunition. However, there was a serious deficit in truly modern examples of APFSDS ammunition due to a variety of factors. The requirement for the T-72B3 obr.2016 to be able to use Svinets-1 and Svinets-2 ammunition was a belated attempt to remedy this problem.

In terms of armour penetration capability, the D-81T had great promise, but the reliance on steel projectiles well into the 80's betrayed the fact that the Soviet munitions industry was not fully capable of producing heavy alloy long rod penetrators. Before 1985, all 125mm APFSDS ammunition followed the basic principle as the 3BM-3 introduced in 1961 in one form or another. The use of steel caps on steel long rod penetrators meant that the performance of these less-than-stellar APFSDS shells at high obliquities was somewhat lower than at lower obliquities, whereas it was the exact opposite with heavy alloy long rod rounds like the M111 "Hetz". Incremental improvements reduced the severity of the issue over time, but the problem was only truly solved when long rod tungsten alloy or depleted uranium shells were introduced in significant quantities in the late 80's. Producing high-quality weapons-grade tungsten carbide and other tungsten alloys in slug form was difficult and expensive, and manufacturing heavy metal alloy rods of sufficient strength for anti-armour purposes was not a trivial task. The equipment simply did not exist in the USSR.

The main defeat mechanism of APFSDS rounds against armoured targets is by damaging internal equipment and killing crew members with shards of broken armour (spall) and fragmentation from the body of the APFSDS projectile itself, but a secondary mechanism is setting internal equipment alight. The huge kinetic energy and extreme forces imparted during armour defeat results in some of that kinetic energy being converted to heat energy, which results in a flash of heat and a shower of high velocity sparks from particles of both armour material as well as penetrator material. This becomes an ignition source for flammable fuel and hydraulic fluid as well as the furniture in the tank.


Besides the enhanced armour penetration capability compared to contemporary foreign KE rounds, 125mm APFSDS was very competitive in terms of precision. According to a Soviet evaluation of the 120mm ammunition for the L11A5 of the Chieftain Mk.5R, the dispersion of the L15A5 APDS round is 0.12 mrads in the vertical plane and 0.15 mrads in the horizontal plane. According to the firing tables for the 3BM-15 shared by Stefan Kotsch, the probable deviation of the shell when fired from the 2A46M at 3000 meters is 0.7 m in both the vertical and horizontal planes, meaning that the dispersion of shots should lie within a 1.4 m circle and the angular dispersion is 0.23 mrad. However, information shared by the online user "Fu_Manchu" states that the spread of L15 APDS rounds at 3000 meters is 1.925 m in the vertical plane and 1.375 m in the horizontal plane. Both rounds could be considered roughly similar, and because L15 is almost identical to L52 APDS in the 105mm caliber, the precision of early Soviet APFSDS could be considered to be on par with contemporary APDS in both the 105mm and 120mm calibers.

According to a firing table provided by Stefan Kotsch, 3BM-15 suffers a 138 m/s drop in velocity over 1 km from the muzzle, and a 127 m/s drop from 1 km to 2 km. This is slightly greater than the drop suffered from the 115mm 3BM-3 shell, which makes sense since the stabilizing fins have a larger diameter and the projectile itself is wider. The ballistic characteristics of 3BM-15 can be used as a surrogate for every other APFSDS shells from the 3BM-9 to the 3BM-22, as they all share the same torpedo shape. Compared to Western 105mm and 120mm APFSDS models, Soviet projectiles suffer from very high drag - more than twice as high - but compensates for this by its exceptionally high muzzle velocity. In general, a higher velocity projectile will have a higher probability of hit against any given target compared to a lower velocity projectile of identical characteristics. This is because the higher velocity projectile will have a shorter time of flight and will therefore be less affected by random factors such as meteorological variables.

The relationship between hit probability and projectile velocity is explored in "Hit Probability of a High Velocity Tank Round" by Fred Bunn. Computer simulations were done using data from the Armament Research Laboratory (ARL) and Army Systems Analysis Agency. It was found that for a "modern fire control system", there is a negligible increase in hit probability when the muzzle velocity of four different types of tank shells is increased from 1,600 m/s to 3,000 m/s when firing a static target from a stationary tank from distances of 1 to 4 kilometers. Four different projectile types were simulated: a conventional finned APFSDS round design, two flared cone-stabilized APFSDS round designs, and a HEAT round design. It was concluded that doubling the speed of an APFSDS round does not improve hit probability against a stationary target by any substantial amount.

Increasing the velocity of the projectiles had a much larger impact on the simulated probability of hit for moving targets. The three charts below show the different types of moving tank targets used in the simulations: the STAGS target, ATMT target and TEMAWS target. The STAGS target has a regular zigzagging movement pattern, the ATMT target has a randomized lateral movement pattern, and the TEMAWS target has an irregular zigzagging movement pattern. The TEMAWS target was considered the easiest and the STAGS target was considered the hardest.

The simulated probability of hit for all three targets at different projectile muzzle velocities is shown in the three graphs below (click to enlarge). It was found that doubling the speed of an APFSDS round increased the hit probability against a moving target by an average of 30% to 35% at a distance of 1 km and by 55% to 60% at a distance of 2 km.

From the results of the simulations, it can be seen that increasing the velocity of a finned APFSDS projectile from 1,600 m/s to 3,000 m/s will increase the probability of hit at all distances and that any increase in velocity will translate into an improvement in hit probability. The difference between a 1,400 m/s APFSDS round and a 1,800 m/s APFSDS round is not drastic, but still contributes toward a higher overall accuracy if all other factors are equal.

The selection of APFSDS ammunition available to the T-72 gave it the upper hand in any engagement with any of NATO's heaviest tanks until the new generation rolled out in the early 80's. Before the introduction of the Leopard 2 and M1 Abrams, the M60A1, M60A2 and Chieftain were the most heavily armoured tank designs used by NATO, but since it is well known that these tanks lack sufficient protection to resist 125mm APFSDS, later variants like the Chieftain Mk.10 are worth more scrutiny.

In 1986, the Chieftain Mk.10 was introduced and Stillbrew armour made its debut. It is rather difficult to find a credible source that describes Stillbrew with any amount of useful detail, but online user "Volketten" has done some research into the topic and claimed in an article published on Rita Sobral's blog that Stillbrew was a form of composite armour that worked like NERA. However, it appears that this is not the case.

The thickness of the frontal projections of the bare Chieftain turret are well known. The thickness across the cheeks on both sides is an average of 125mm at 60 degrees, and 158mm at 35 degrees at the base. The effective thickness ranged from a minimum of 240mm to a maximum of 280mm, and the average thickness is 250mm. In terms of total thickness, the average value for a bare Chieftain turret is less than the NATO Single Medium target which is 130mm of RHA steel at 60 degrees, and the cast steel of the Chieftain would also be less effective than rolled steel. There is some difficulty in determining the exact thicknesses of Stillbrew over the various curves and bumps of the turret, but on page 58 of the book "Chieftain Main Battle Tank: 1966 to the present" it is stated that the total thickness of the armour including Stillbrew is around 500mm, although the armour does not reach this thickness consistently due to a variety of factors causing the thickness to fluctuate from 480mm to 540mm. Also, it is stated that each Stillbrew panel is required to be able to survive at least two hits, contradicting other information claiming that the panels are designed to fall off in one hit.

U.S Patent 4848211 is the patent for "Stillbrew" armour. If any doubts persist, note that one of the inventors listed is a "John H.T Brewer", and that the priority date for the patent is 4 June 1986. The patent states that the armour is a composite armour, but makes absolutely no mention of the movement of the add-on plate, and it is easy to see why. The add-on cast armour plate is simply too thick and too heavy to move and the plate is securely screwed to the turret by large countersunk stainless steel screws. The use of countersunk screws is the most significant piece of evidence, because the beveled rim of the screws (for example) pins the plate down and prevents the plate from sliding up the shaft of the screw. Furthermore, large thicknesses of rubber does not necessarily mean that more energy is transferred to the cast add-on plate because the energy of a penetrator absorbed in the rubber causes it to expand radially, so much of the energy is wasted as it is transmitted transversely along the rubber layer and not into the plate.

One of the confirmed roles of the rubber interlayer is to absorb some of the force from an impact landing on the add-on cast plate and to prevent the plate from falling off in much the same way as the rubber mounting blocks found on the Leopard 1A1A1. The function of the rubber is almost certainly as a simple sandwich material in the same way silica was used in early siliceous core armour and glass textolite was used in the sandwich armour in the upper glacis of the T-64A and the T-72. The 50mm rubber interlayer has an effective thickness of 100mm when sloped at the 60 degree angle of the turret, so the effects of the rubber layer are not negligible when attacked with a HEAT warhead. There are some claims that the rubber acts like a spring because it is apparently compressed during the installation procedure of "Stillbrew" and is designed to release its compressed energy when attacked. However, this photo of a damaged Chieftain used as a range target seems to disprove this theory as it shows that the thickness of rubber behind the untouched "Stillbrew" panel is the same as the thickness of exposed rubber behind the damaged panel, denoting a lack of expansion.

Nevertheless, the addition of "Stillbrew" armour would have made the turret immune to typical Soviet single-charge 125mm HEAT shells (like BK-14M) as well as any APFSDS round from before 1985, probably down to point blank range. "Stillbrew" can be considered a crude and cheap method of increasing the protection of the Chieftain against certain threats to an acceptable standard, but it must be emphasized that this was achieved by simply adding more steel on top of steel, not by the use of a unique mechanism.

Having said this, it must still be understood that Soviet 125mm APFSDS ammunition would never have truly had any real trouble defeating NATO tank armour from the same era. "Stillbrew", for example, was introduced in 1986 but by 1985, the "Vant" depleted uranium shell with a claimed penetration of 500mm RHA at 60 degrees at 2 kilometers became available in the Soviet arsenal. It would presumably be capable of defeating "Stillbrew" at combat distances of 1-1.5 kilometers. Even non-NATO tanks like the Strv 103 from 1968 would have fared very poorly against Soviet 125mm guns. Rickard Lindström reports in this article that during a Swedish test in the early 90's involving an Strv 103 and a T-72 tank (purchased from the ex-GDR after the collapse of the USSR), 3BM-22 rounds fired at the famous S-tank proved to be so powerful that it went all the way through the entire tank. This is hardly surprising given that 3BM-22 is capable of piercing the turret roof of the T-72B tank (45mm angled at 78 degrees) from a distance of 3,700 meters when the upper glacis armour of the Strv 103 is equivalent to around 50mm at 78 degrees.

The new generation of NATO tanks undoubtedly gained an increased resistance to the older style of composite cored APFSDS ammunition for the 125mm gun, but even then, the hull armour and the sides of the hull and turret remained inconveniently soft, so Western engineers had to find creative solutions. The Leopard 2 shielded the crew compartment from the side with three heavy 110mm ballistic modules (consisting of two 50mm plates each separated by a 10mm air gap) bolted to the side of the hull just over the first two roadwheels (source), This armour was used since the first batch of tanks in 1979 up to the sixth batch in 1985 where it could be found on the first hundred Leopard 2A4 tanks. From an incidence angle of 30 degrees from the longitudinal axis of the tank, this arrangement was worth 200mm RHA in spaced sloped armour. The Abrams implemented a similar design with its 65mm composite sideskirts. These were thinner than the heavy ballistic plates on the Leopard 2 and the base hull armour of the Abrams is also thinner, but the larger skirts of the Abrams provided much greater coverage and also probably offered better shaped charge protection as the skirts incorporated a pair of thin NERA sandwiches. These protective measures may have been somewhat useful until the introduction of Soviet monobloc long rod APFSDS rounds in the mid-80's, which is not far from the introduction dates of the Leopard 2 and Abrams.

The Soviet standard for calculating the penetration limit of armour piercing projectiles is V80, meaning that the expected consistency of achieving full armour perforation given a certain projectile velocity must be 80%. In formulas, V80 must replace V50 (50% chance of armour perforation). For example, if a certain projectile has to penetrate 500mm of steel, then at least 80% of all projectiles of that type must achieve that standard. Also, the Soviet criteria for a full armour perforation dictates that 80% of projectile mass must be recorded on the other side of the target plate as opposed to U.S Army criteria which only requires that a hole is produced in the armour such that light can be seen from the other side. Overall, Soviet standards were not only stricter, but the steel they used for targets was sometimes of a greater hardness than NATO targets. In reality, the given penetration data may be an underrepresentation of the actual achievable penetration of these shells.


3BM-10 (3BM-9 Projectile)

The 3VBM-3 round was adopted in 1968 for the T-64A, but it was quickly relegated for use in training due to its obsolescent design and lack of future prospects. The 3BM-9 projectile is an upscaled modification of the 115mm 3BM-6 but uses 60KhNM maraging steel (310 BHN) as opposed to 35KhZNM tool steel (600 BHN). The 3BM-9 projectile features a steel penetrator and a ballistic cap but does not have a soft steel armour piercing cap like the 3BM-6 that it was based on. The design of the round was probably an attempt to conserve tungsten carbide while achieving the same performance as the 115mm 3BM-3 tungsten cored round, as 3BM-9 can penetrate the same armount of armour as 115mm 3BM-3 (300mm) tungsten core rounds at 0 degrees at a distance of 1 km but without the use of any tungsten carbide. The use of an inferior steel is probably responsible for the reduced performance of this much speedier shell compared to the 115mm 3BM-6.

Nevertheless, 3BM-9 was more than enough for any NATO tank of the time, including the Chieftain. According to a Soviet analysis of an Iranian Chieftain captured by the Iraqi army during the early part of the Iran-Iraq war, available here on Andrei Tarasenko's website, the Chieftain Mk.5 had totally insufficient protection even from the 3BM-9. The frontal part of the entire turret, hull upper front plate and lower front plate could all be defeated at 3 km or more. This essentially means that the T-72 Ural can defeat one of NATO's toughest tanks at any reasonable combat distance with zero expenditure of valuable tungsten - an extremely profitable situation for the Red Army in the event of a large scale war.

Besides the materials used in the making of the projectile itself, 3BM-9 is also quite remarkable for being the first service munition to be fired at truly hypersonic speeds (Mach 5+). This shell used a steel "ring" type sabot with a copper driving band. Sabot construction is critical to shooting accuracy, and the steel "ring" type sabot was perfectly fine compared to any other APDS sabot at the time. Plus, any deficiencies in accuracy from the sabot would be unnoticeable given the gobsmacking speed of the projectile coming out of it.

This is just speculation, but it is possible that 3BM-9 might have been what T-72 Urals were given for their first years of service as an intermediary before the supply of 3BM-15 shells (introduced one year before the T-72 Ural) was assured. T-64s should be the first to transition to the newer ammunition, and the T-72 might have had to wait.

Muzzle velocity: 1800 m/s

Mass of Projectile: 3.6 kg
Mass of Sabot: 2.02 kg
Total Mass: 5.67 kg

Total length of projectile: 518mm
Length of steel penetrator only: 410mm

Projectile maximum diameter: 44mm
Projectile minimum diameter: 30mm
Average Diameter of Projectile: 36mm

Penetration at 2.0 km:

245mm at 0°
185mm at 45°
140/150mm at 60°

Penetration at 1.0 km:

300mm at 0°
160mm at 60°

(According to a Soviet GRAU document and a comparison chart)

Penetration at 2.0 km:

290mm at 0°
140mm at 60°

It should be noted that the stated penetration is only 245mm of penetration at 2 km when the 115mm 3BM-6 has 240mm of penetration at 2 km despite having a significantly lower impact velocity (1358 m/s vs 1540 m/s). In terms of dimensions, the two projectiles are close: the 3BM-9 has a steel penetrator of 410mm in length and 36mm in average diameter (min 30mm, max 44mm) whereas the 3BM-6 has a steel penetrator of around 400mm in length (including the armour piercing cap) and 36mm in average diameter (min 30mm, max 42mm). The only difference is that the 3BM-9 uses a relatively soft but strong maraging steel (treated to 310 BHN) whereas the 3BM-6 rod uses a high hardness high strength steel (35KhZNM tool steel, 600 BHN). The high strength and ductility of maraging steel combined with the ease of processing the steel probably made it quite cheap to manufacture ammunition like the 3BM-9, although the low hardness compared to tool steel clearly had an large negative impact. The ductility of maraging steel probably influenced the decision to forego a soft armour piercing cap, which was present on the 3BM-6. The 3BM-12 round that was developed after the 3BM-9 used a 35KhZNM tool steel penetrator and had a soft steel armour piercing cap, and the penetration of this improved round is reported to be 350mm at 2 km.

Nevertheless, it should be noted that the achievable armour penetration of 3BM-9 is somewhat higher than its certified penetration capability due to the use of V80 ballistic limit in Russia as opposed to the V50 ballistic limit as well as due to the different definition of armour perforation. The post armour penetration effects of the shell are very powerful, due to the large hole created by the inefficient steel penetrator.


3BM-16 (3BM-15 Projectile)

The 3BM15 has a steel penetrator made from 35KhZNM tool steel with a small VN-8 tungsten carbide core. The VN-8 core is a cemented tungsten carbide with an 8% nickel binder matrix. 3BM-15 was introduced in 1972, just one year before the T-72 Ural. It is externally identical to the 3BM-9 projectile, but structurally, it bears a non-trivial similarity with Soviet APDS ammunition that entered service in the late 60's and even has connections to vintage APCR designs. Although decently hefty and very speedy, the shell primarily relies on a small tungsten carbide subcaliber core for a large part of the penetration period. A ballistic windshield was crimped onto the soft 30KhGSA steel armour piercing cap. The purpose of the shock absorber cap was to improve the performance of the shell on sloped armour and to reduce the shock of the impact experienced by the brittle tungsten carbide core. The impact of the steel armour penetrating cap creates the distinct large entry cavities that were typical of Soviet tungsten-cored steel rounds. Tag number 5 in the photo below marks the impact of a 3BM-15 shell into the left turret cheek of a T-72M1 in Finland, Parola. Photo by Andrej Smirnov.

All this doesn't mean that it cannot go through incredible thicknesses of steel, however.

The photo on the left shows the entry and exit of a 3BM-15 projectile into a 200mm steel armour plate at an impact angle of 0 degrees and an impact velocity of 1,280 m/s, corresponding to a distance of 4,000 meters. The photo on the right below shows the penetrated cavity inside a 200mm plate from a 300mm combined armour block (100mm + 200mm) after the penetration of a 3BM-15 round at an impact angle of 0 degrees at the impact velocity of 1,198 m/s, corresponding to a distance of 4,600 m. The velocities and corresponding distances were traced using a firing table for 3BM-15 made available to the public courtesy of Stefan Kotsch.

The nature of the composite construction of 3BM-15 makes it immensely powerful against homogeneous steel targets, though only in the case of perpendicular or near-perpendicular impacts. The two graphs below show the penetration depth for three types of Soviet armour piercing ammunition on a 0 degree target at a fixed impact velocity of 1,500 m/s. The smooth line represents 122mm 3BM11 and 100mm 3BM8 (the two projectiles share the same core. The dashed and dotted line represents 3BM6 steel penetrator. The dashed line represents 3BM-15 steel and tungsten carbide core penetrator. The graph on the top (a) shows the penetration in physical plate thickness. The graph on the bottom (b) shows the penetration in line-of-sight (LOS) plate thickness after dividing by the cosine of the impact angle.

The impact velocity of 1,500 m/s corresponds to a distance of 2,100 meters for the 3BM-15, and as you can see, the penetration at 0 degrees is around 460-470mm RHA. The penetration falls dramatically at 15 degrees and falls to around 280mm between the angles of 30 degrees and 50 degrees, but rises to 300mm at the impact angle of 60 degrees. The penetration unexpectedly rises at an impact angle of more than 60 degrees and increases to a maximum around 360mm at an impact angle of 80 degrees before dropping off sharply, presumably due to projectile ricochet. The drop in penetartion at the 15 degree critical angle was investigated and attributed to the location of the tungsten carbide core.

"При взаимодействии с наборной (без зазора) или монолитной преградой лидирующий сердечник, сохраняющий благодаря всестороннему обжатию относительную целостность, снижает интенсивность расходования наседающей массы стального корпуса, экономить запас кинетической энергии снаряда и повышает его бронепробивное лействи в целом. Разрезы по поражениями в толстых наборных преградах показывают, что стальной корпус снаряда, разрушаясь путем трещинообразования, по  расходуется мало; расходование его массы происходит экономно преимущественно за счет "стачывания" по боковой поверхности. По этой причине снаряды типа 3БМ-15 в диапазоне углов 0 ... 15 град, при которых сердечник функционирует нормально, обладают значительно более высоким бронебойным действием по монолитной броне, чем цельнокорпусные снаряды типа 3БМ-6."


"When interacting with a stack of plates (without gaps) or a monolithic target, the leading core, which preserves the relative integrity due to a uniform compression, reduces the intensity of the consumption of the mass of the steel casing, retains the kinetic energy of the projectile, and raises its armor-piercing capability as a whole. The cutaways of penetration channels in thick targets show that the steel shell of the projectile, being destroyed by fracturing, is consumed little; the expenditure of its mass occurs economically mainly due to "grinding" along the lateral surface. For this reason, projectiles of type 3BM-15 in the range of angles of 0-15 degrees at which the core functions normally have a much higher armor-piercing action on monolithic armor than solid-shell projectiles of the 3BM-6 type."

Being installed at the front of the projectile, the tungsten carbide core strikes the target plate first and creates an entry channel. For low obliquity impacts, the steel penetrator behind the core is able to follow the core into this entry channel and only the outer parts of the steel penetrator rod are ground off due to the larger diameter of the steel penetrator compared to the diameter of the core. Thus, very little steel is eroded during the penetration of very thick blocks of armour. The high kinetic energy of the combined penetrator assembly (core + steel rod) is maintained due to the lack of erosion of the core and the high kinetic energy of the assembly, primarily from the mass of the steel penetrator. To put it another way, the projectile behaves like an arrow: the metal tip penetrates the flesh while the wooden shaft simply follows. The red zones in the drawing below shows the parts of the outer edges of the steel penetrator that are "ground off" during penetration:

The penetration channel created in an armour plate has a certain shape that is characteristic of the torpedo form of the 3BM-15 projectile. Referring to the photo below, it can be seen that the penetration channel at the surface of the plate is largest. This is due to the large 44mm diameter of the steel penetrator near the tip of the round. The taper of the projectile means that the diameter of the steel penetrator decreases throughout the penetration process, and as a result, the diameter of the penetration channel decreases accordingly.

This use of a steel rod behind a tungsten carbide core allows huge thicknesses of steel to be penetrated in an extremely efficient manner with minimal expenditure of valuable tungsten. However, 3BM-15 is not the first example of this type of shell. There exists a variant of the BR-354P APCR round in the 76x385mm caliber that works on the same principle, shown below (right). In this variant, the tungsten carbide core is held in a soft metal "arrowhead" projectile with a steel plug placed behind it. This enabled deeper penetration into armour as well as greater beyond-armour damage without a large increase in the size of the tungsten carbide core of the basic BR-354P shell (left), which had a core of a similar size but no steel plug. This was an economical alternative to having a larger tungsten carbide core. Indeed, early Soviet APFSDS rounds like 3BM-15 share a surprising number of similarities with vintage APCR. More information on BR-354P is available in Tankograd's PT-76 article.

At an angle of between 0 to 15 degrees, the projectile behaves in this manner and is able to penetrate a huge thickness of steel. At higher angles, the tungsten carbide core separates from the steel penetrator due to misalignment and the penetration of the armour plate is done by the steel rod alone. The graphical curves for penetration thicknesses achieved by 3BM-15 and 3BM-6 intercept at an angle of around 30 degrees. Due to this phenomenon, 3BM-15 can penetrate very high thicknesses of steel at flat angles but does not necessarily perform better on sloped plate compared to a steel long rod penetrator. At armour slopes of beyond 30 degrees, the 3BM-6 tool steel penetrator is able to perform better than 3BM-15 at the same impact velocity. The only advantage of 3BM-15 over 3BM-6 in this respect is that its muzzle velocity is much higher, so that the impact velocity of 3BM-15 will always be higher than 3BM-6 for any given distance.

The mechanisms at play on monolithic plates are complicated on their own, but the behaviour of 3BM-15 on multi-layered armour is even more complicated:

A stacked 300mm plate consisting of a 100mm plate and 200mm plate without an air gap in between was found to be 4.6% more resistant than a monolithic plate of the same thickness at an impact angle of 0 degrees. The introduction of an air gap between two plates proved to be an effective method of severely degrading the penetration of 3BM-15, but only for perpendicular impacts. In the table below, the first column from the left shows the impact angle and the next three columns from the left list the spaced armour configurations: b1 and b2 denote the thickness of the first and second plates in millimeters, and L denotes the size of the air gap in millimeters. The fourth column from the right lists the velocity limit of 3BM-15 for the spaced described armour configuration, and the third column from the right lists the velocity limit for a monolithic plate of the same thickness in steel (b1 + b2). The difference in the velocity limit is listed in the second column from the right. The first column on the right shows the difference in the velocity limits between the spaced armour configuration and a monolithic plate in percentage points, and also represents the improvement in mass efficiency.

The best performing spaced armour configuration was the 100-1000-200 array, where a 48.1% improvement in mass efficiency was recorded. The velocity limit of 1,740 m/s corresponds to a distance of 300 meters. Of course, the large 1-meter air gap between the two spaced plates is rather large and impractical for tank armour purposes. The effectiveness of the spaced armour is dependent on the size of the air gap to a large extent as demonstrated by the increasing effectiveness of a spaced armour configuration with a 50mm front plate and a 200mm back plate as the size of the air gap is increased from 70mm to 300mm to to 1,000mm, but the optimal air gap size was 480mm. The improvement in mass efficiency was 46.4% for the optimal configuration - slightly higher than the array with the 1,000mm air gap. Of course, there are no examples of tank armour with completely vertical facings, so the results from the 15 degree and 30 degree impacts are much more interesting. The 30 degree impact cases are particularly interesting because the turret cheeks of the Abrams tank are known to be sloped at 21 and 30 degrees for each side, and the arrangement of plates likely follows the hull and mantlet armour modules where a pack of NERA panels is located behind a thin front plate, and an air gap is placed between the NERA pack and the thick back plate. The closest equivalent for this is the configuration described in the first row for 30 degree targets: a 50mm front plate and a 150mm back plate with a 330mm air gap. For the turret of an M1 Abrams, the air gap should be larger and some small effect from the NERA pack is expected, but overall, the mass improvement in efficiency from such a spaced armour configuration appears to be very small at only 2.4%. The distance corresponding to the velocity limit for the spaced array is 3,100 meters.

The table below shows the data for 3BM-15 on spaced armour at 60 to 65 degrees obliquity. The two new columns on the right lists the penetration depth into the second spaced plates (b2) using 4-6 test shots for the data, and the average penetration depth. The penetration depth is usually larger than the LOS thickness of the second spaced plates because the penetrator is heavily deflected as it travels through the air gap, so a relatively shallow penetration channel is gouged along the surface of the plate by the penetrator. Nevertheless, at the higher angles of 60 degrees and 65 degrees, spaced armour is not significantly more effective than a monolithic armour plate unless air gaps of more than 1 meter are used. For instance, the best result for spaced armour at 60 degrees was achieved with a 1,700mm air gap. Taking the 60 degree angle into consideration, the total size of the air gap is 3.4 meters. The best result for spaced armour at 65 degrees follows the same pattern. Even though 3BM-15 failed to fully defeat most of the targets (rows marked with "-" indicate full armour perforation), the impact velocities were extremely low, corresponding to a distance of around 6 km.

The 100-1430-50 spaced armour configuration at 60 degrees is very interesting because it broadly represents the toughest part of the side armour of the Leopard 2A0-A4, albeit with a much bigger air gap. According to the table, 3BM-15 is capable of defeating this target at an impact velocity of 1,590 m/s, corresponding to a distance of 1,450 m. The actual velocity limit for a more correct representation of the side armour should be lower than the listed value as the distance between the heavy 110mm ballistic plates and the 50mm side armour of the Leopard 2 should around 650mm, based on the known width of the D570F tracks of the Leopard 2 (635mm). The difference between a ~650mm air gap and a 1,430mm air gap is not negligible, to put it mildly, so in other words, the 3BM-15 should be capable of defeating the most well-protected part of the side hull armour of the Leopard 2 at typical combat distances from a side angle of 30 degrees.

The photo below shows the defeat of a 100mm + 200mm RHA plate by 122mm 3BM11 tungsten-cored APDS at a 0 degree impact angle. The perforated plate on the left was achieved at an impact velocity of 1,272 m/s and the partial perforation on the right (you can see the remnants of the penetrator) was done at an impact velocity of 1,246 m/s (normal muzzle velocity from M-62 is 1,575 m/s).

From comparing the penetration paths of 3BM-15 and 3BM11, it can be seen that both projectiles create a relatively slender penetration channel through the plate but leave a large crater on the surface of the plate. However, the 3BM-15 projectile leaves an extremely deep crater. The 3BM11 projectile has a soft steel armour piercing cap of around 60mm in length in front of the heavy 50x120mm tungsten carbide core. From what we can see in the cross section of the plates, this soft steel cap is responsible for producing the shallow crater, but for the 3BM-15, the soft steel armour piercing cap over its 20mm tungsten carbide core is only 20mm in length so the source of the deep crater must be the steel penetrator behind the core.

3BM-15 was designed with adherence to the same design principle as the older 115mm 3BM-3 round, and 3BM-3 was intended to have armour penetration capabilities similar to or exceeding that of a contemporary APDS shell without incorporating as much tungsten carbide in its construction. 3BM-3 was highly successful in this regard, as it managed to achieve more penetration than BM-8 APDS for the 100mm D-10T using only a tenth of the amount of tungsten carbide in its core.

The fact that 3BM-15 is able to penetrate an astonishing ~470mm at 2,100 meters with a small 0.27 kg tungsten carbide slug when the large 2.78 kg core of the 3BM11 manages 400mm at the same velocity despite having more than ten times the mass is extremely interesting, to put it mildly. The difference in muzzle velocity between the two rounds should not be neglected, of course, and when we consider that the 3BM11 has a muzzle velocity of 1,575 m/s when fired out of an M-62 cannon whereas 3BM-15 has a muzzle velocity of 1,780 m/s, it is clear that the 3BM-15 APFSDS design is simpler superior to equivalent APDS for a low obliquity impact. However, because of the fact that tungsten carbide penetrators practically do not erode while penetrating through armour plate, a heavy 2.78 kg core with a 50mm diameter will still generate a larger spray of core fragments, armour fragments and spall when it perforates an oblique plate compared to the combination of a small 0.27 kg core and a 3.63 kg steel rod. The lower post-perforation lethality of the 3BM-15 APFSDS shell on oblique plates is one critical drawback of the small-cored high velocity design, the other being the greatly reduced performance at the range of angles from 15 to 60 degrees. As usual, there is no "free lunch", so to speak.

Nevertheless, a T-72 issued with 3BM-15 rounds was lethal to all NATO tanks of the time, but it was exceptionally lethal to tanks like the AMX-30 and Leopard 1 which had particularly light armour that would not be enough to stop the round, yet offer too much resistance to prevent the steel projectile from fragmenting after perforating the armour plate. The only caveat is that the 3BM-9 round would perform even better against such armour configurations and do so without any expenditure of tungsten at all. 3BM-9 would also perform better on high obliquity targets so it would be more effective against tanks with heavy sloped armour plating like the Chieftain and M60A1, essentially meaning that the high-performance 3BM-15 lacked a niche during the 70's. 3BM-15 may have gained some usefulness when new NATO armour emerged in the early 80's as the armour of the new generation tanks like the Leopard 2 and M1 Abrams were largely flat, allowing the superior performance of 3BM-15 to show.

3BM-15 uses the same steel "ring" type sabot as the 3BM-9. The photo below is from a Rheinmetall brochure on PELE ammunition, demonstrating a modified 3BM-15 PELE round in flight and the airflow around the components of the round. The sabot was unmodified.

An incremental propellant charge is wrapped around the projectile body.

Mass of Incremental Charge: 4.86kg
Maximum Chamber Pressure: 4440 bar

Muzzle velocity: 1785 m/s

Steel body maximum diameter: 44mm
Steel body minimum diameter: 30mm
Armour piercing cap diameter: 20mm
Core diameter: 20mm

Total length of projectile: 548mm
Length of steel penetrator only: 435mm
Length of armour piercing cap: 20mm
Length of core: 71mm

Mass of Steel body: 3.63kg
Mass of Core: 0.270 kg

Total Mass of Projectile: 3.83 kg

Penetration at 2.0 km: 
400mm at 0°
200mm at 45°
150mm at 60°
These figures come from a Soviet GRAU document and are corroborated by the penetration data presented in "Particular Questions of Terminal Ballistics" 2006 (Частные Вопросы Конечной Баллистики) published by Bauman Moscow State Technical University on behalf of NII Stali, except for the penetration at 0 degrees which is much less than the 470mm obtained in the testing.

Penetration at 2.0 km
400mm at 0°
180mm at 60°
These figures come from page 587 of the "Textbook of Means of Defeat and Ammunition" 2008 (Учебник Средства Поражения И Боеприпасы) published by Bauman Moscow State Technical University.

Penetration at 1.0 km
425mm at 0° 
This figure comes from "Kampfpanzer: Die Entwicklungen der Nachkriegszeit" by Rolf Hilmes. 

Penetration at 2.0 km
200mm at 38°
150mm at 60°
120mm at 67° 
These figures come from a marketing brochure for the M88 round, a Serbian copy of 3BM-15.


The steel "wedge" in front of the tungsten carbide slug is a soft steel armour piercing cap to protect it from shattering at the moment of impact with a steel target, and to improve performance on sloped armour plate. Compared to the tungsten carbide core of 3BM-8 (far left), the core for 3BM-15 (far right) is incredibly tiny, and yet 3BM-15 penetrates far more armour. As mentioned before, this was achieved solely by the higher launching velocity of 125mm APFSDS compared to 100mm APDS.

(Credit for photos to PzGr40 from

(Sourced from,


3BM-18 (3BM-17 Projectile)

This shell is essentially identical to the 3BM-15 externally, but it lacked the tungsten carbide core of its parent and had a modified armour piercing cap, presumably to fill in the gap of the absent core and to improve the performance on sloped armour plate. The steel penetrator is made from 35KhZNM steel as well.

Muzzle velocity: 1780 m/s

Armour piercing cap diameter: 30mm

Total length of projectile: 548mm
Length of Projectile only: 435mm
Length of armour piercing cap: 50mm

3VBM-9 (Zakolka)

3BM-23 (3BM-22 Projectile)

3BM-22 is a derivative of 3BM-15. It began mass production in 1976, but only formally entered service in 1977. It features an enlarged and improved armour piercing cap in front of the tungsten carbide core to further improve performance on sloped armour plate. The projectile is shorter than the 3BM-15, and it retains the steel ring-type sabot. The clearest difference between the 3BM-22 and the 3BM-15 is use of a tungsten alloy armour piercing cap made from VNZh-30MT alloy. Tungsten alloy generally has greater yield strength than tungsten carbide and increased toughness, giving it better resistance to shattering compared to tungsten carbide and tool steel. This enhances the performance of 3BM-22 on angled armour plate as the tip of the projectile is much more resistant to deflection. The armour piercing cap has a larger diameter than the core and protrudes beyond the steel body, so that the armour piercing cap invariably strikes the target before the rest of the projectile even on a very high obliquity target. An additional benefit to the use of a large tungsten alloy armour piercing cap is that spaced armour may be substantially less effective against 3BM-22 as the cap is much more resilient than the steel caps of earlier designs. The tungsten carbide core at the tip of the projectile is the same design as its predecessors and it is made from the same VN-8 tungsten carbide. The steel body is made with the same 35KhZNM steel as earlier APFSDS rounds.
3BM-22 is notable for its treatment as a surrogate for foreign APFSDS rounds during the 70's by Soviet scientists and engineers when evaluating tank armour and in related topics. It was also used in the evaluation of prospective reactive armour designs.

Mass of Incremental Charge: 4.86 kg
Maximum Chamber Pressure: 4440 bar

Muzzle velocity: 1785 m/s

Steel body maximum diameter: 44mm
Steel body minimum diameter: 30mm
Armour piercing cap maximum diameter: 30mm
Armour piercing cap minimum diameter: 27mm
Core diameter: 20mm

Total length of projectile: 558mm
Total length of penetrating elements: 453mm
Length of steel penetrator only: 400mm
Total length of tungsten armour piercing cap: 88mm
Length of tungsten armour piercing cap in front of steel penetrator: 35mm
Length of tungsten carbide core: 71mm

Total projectile mass: 4.485 kg
Mass of core: 0.270 kg

Penetration at 2.0 km:
470mm at 0°
220mm at 60°
These figures come from page 587 of "Учебник Средства Поражения И Боеприпасы" 2008 (Textbook of Means of Defeat and Ammunition) published by the Bauman Moscow State Technical University.

Penetration at 2.0km:
420mm at 0°
170mm at 60°
Mentioned by Mikhail Rastopshin in an article. Also listed in a fact sheet from his article "Наши танки в реальной войне обречены?" (Are Our Tanks Doomed in a Real War?) published in 2010.

As usual, it is important to treat the armour piercing capabilities of 3BM-22 with some care and consideration. Its performance on monolithic homogeneous steel armour is likely to be the same as advertised (aligning with the different numbers provided by different sources), but its performance on complex composite armour of similar mass will invariably be less. It has been claimed that the M1 Abrams is equivalent to 400mm of RHA steel, but it is obviously not possible for this to be true for the simple reason that composite armour is simply too complex to be distilled into a single figure in RHA equivalence.

The "400mm" figure might be true for a certain type of round, but even if it could be guessed with a reasonable level of certainty that a certain type of round was used as baseline to represent the probable Soviet threat, the fact that the document specifically identifies the armor as belonging to "one version of the M1 turret armor" makes it doubtful that any meaningful conclusions can be made. This was most likely intentional, of course.

Existing stocks of 3BM-22 are currently being expended in live-fire exercises, for which older projectiles are favoured since they are less harsh on the gun barrel.

3VBM-11 (Nadezhda)

3BM-27 (3BM-26 Projectile)

Officially entering service in 1983, the 3BM-26 projectile is the most optimum Soviet APFSDS shell that is still based on the composite design principle of a small tungsten slug placed in a steel long rod body. Unlike previous shells that used the steel "ring" type sabot, the 3BM-26 projectile rides on a "bucket" type sabot made from a lightweight aluminium alloy. The new design of the "bucket" type sabot interfaced with the projectile via fine threads as opposed to six large threads as found on the "ring" type sabots, and this significantly contributed to the improved accuracy of the shell, although the magnitude of the improvement is not known. This round was the first to use the high-energy Zh63 propellant charge, allowing the heavier 3BM-26 projectile to reach the same muzzle velocity as previous models despite the increase in projectile mass at the cost of accelerated barrel wear due to the higher pressure. Additionally, 3BM-26 differs from its predecessors in that it uses a VNTs (tungsten-nickel-zinc) tungsten alloy core instead of a tungsten carbide core. Direct information on the composition of this alloy is scarce, but nickel and zinc are commonly alloyed with metals to increase hardness. Thus, it is likely that the incentive behind the transition from a cemented carbide to an alloy was to eliminate the brittleness of cemented carbides while remaining as hard as possible.

Behind the tungsten alloy core is a short steel rod with a length approximately equal to the core itself (71mm) and a diameter exactly equal to the core (20mm). This layout is exactly the same as a late variant of the BR-354P APCR round. The steel rod fulfills the same function as the steel rod behind cores of the earlier 3BM-15 and 3BM-22 projectiles - increasing the kinetic energy of the core with a minimal expenditure of tungsten. Behind the rod is the tracer.

Unlike the 3BM-22 and 3BM-15 that preceded it, the core is located at the tail of the projectile body and not at the tip. This means that the core will only begin to come in contact with the target material only when the steel body in front has been completely eroded during penetration. There is an air space forward of the core to give it room for forward travel under its own inertia as the rest of the body decelerates within the target material. This is to allow the core to retain the same impact velocity as the rest of the projectile at the moment of contact with the armour plate in spite of the erosion and deceleration of the steel penetrator in front of the core, so if, for example, the 3BM-26 projectile impacts a armour plate of infinite thickness at 1520 m/s, the steel rod will penetrate the target until it is completely eroded inside the armour plate so that its final velocity is 0 m/s, but the core at the rear of the projectile retains the same impact velocity. Immediately after the steel rod is eroded, the core impacts the surface of the armour cavity and begins penetrating the target. The use of a tungsten alloy instead of a carbide vastly reduces the likelihood that the bare core will shatter when it impacts the plate after the steel rod is fully eroded.

At the very front of the projectile is the ballistic cap, crimped onto the armour piercing cap, which is slightly larger than the one in the 3BM-22 projectile. The increased length of the armour piercing cap did not significantly affect the performance of the 3BM-26 on oblique targets, as evidenced by the fact that the penetration of this shell for a plate at 60 degrees is the same as the 3BM-22. To the contrary, the relative performance of the 3BM-26 projectile on sloped plate appears to be slightly worse seeing as the penetration on a perpendicular plate was increased but the penetration on sloped plate was not.

Out of all Soviet cored steel projectile designs, this one has the best prospects against the newly emerging NERA armour of the M1 Abrams and Leopard 2. The projectile is still as vulnerable to fracturing and disintegration when strong lateral forces are imparted onto the steel body by the bulging plates of a NERA array, but due to the location of the tungsten alloy slug, 3BM-26 can retain its most potent component upon reaching the back plate of the armour array. In previous designs, the tungsten carbide slug could be dislocated from the rest of the projectile after emerging from behind a spaced armour plate whereby it could yaw and shatter upon impact with the base armour - a consequence of the brittleness of tungsten carbide - while the intact steel tail of the projectile may continue to impact the back plate with slim chances of defeating it. Generally speaking, steel long rod projectiles will perform very badly against NERA armour or sloped spaced armour relative to monobloc tungsten alloy or depleted uranium alloy projectiles due to the comparatively low yield strength of steel. As explained in Part 2 of this T-72 article, NERA plates and sloped space armour will defeat long rod projectiles via the destruction of the rod through the application of lateral stresses. In short, the lower yield strength of steel makes it more susceptible to structural failure when it experiences strong lateral stresses as it exits the back of a sloped spaced steel plate or as it passes through a NERA array. The tail usually survives the experience, but when the front part of the rod is broken up, the penetration of the tail may not be enough to defeat the back plate of the armour array. For a more holistic understanding of the concept, please visit Part 2.

In the M1 Abrams, the back plate is a little more than 100mm thick. If we consider a scenario where an older steel round like the 3BM-15 were to penetrate the armour array of an M1 Abrams, the defeat of the back plate requires at least a third of the steel penetrator to survive the interaction with the NERA array and the front plate. Given that the steel penetrator in previous APFSDS rounds is 435mm in length and made from tool steel, success is a distinct possibility but the probability of succeeding is still not very high. With a tungsten alloy core in the tail, the probability of success for the tail of the 3BM-26 penetrator is much higher. The armour piercing cap and the steel penetrator have a combined length of around 255mm, not counting the hollow tail of the penetrator which houses the core, so the penetrator can be expected to penetrate around 130mm of RHA steel on its own. The remaining thickness of armour is handled by the core.

It must be observed that the use of a small tungsten alloy core instead of a more elongated penetrator like on the older American M735 was already an anachronism by foreign standards in 1983, but at least it was still a highly economical solution considering the lackluster technological capabilities of the Soviet munitions industry at the time. An additional economic advantage to 3BM-26 is that its core shares the same dimensions as the earlier pattern of tungsten carbide cores. It is possible to retrofit older stocks of ammunition with the more effective core.

Mass of the sabot: 2.2 kg

Mass of the projectile only: 4.8kg

Total length of projectile: 558mm
Length of (partly hollow) steel penetrator only: 395mm
Length of armour piercing cap: 115mm
Length of core: 71mm
Diameter of core: 20mm

Maximum diameter of the projectile: 36mm
Maximum diameter of armour-piercing cap: 36mm
Minimum diameter of armour-piercing cap: 27mm

Muzzle Velocity: 1720 m/s

Penetration at 2.0 km:
490mm at 0°
230mm at 60°
These figures come from page 587 of the "Textbook of Means of Defeat and Ammunition" 2008 (Учебник Средства Поражения И Боеприпасы) published by Bauman Moscow State Technical University.

There appears to be some discrepancy in the penetration data, as it is claimed in some Russian sources that 3BM-26 "Nadezhda" has 18% more penetration than 3BM-22 "Zakolka". If the data of the certified penetration for 3BM-22 is used as a reference point (380mm at 2 km at 0° and 170mm at 2 km at 60° according to Fofanov), then the certified penetration of 3BM-26 at 2 km must be 448mm at 0° and 200mm at 60°. Only the penetration figures at 60° matches the claimed 18% increase (200mm vs 170mm), but the penetration on flat plate does not show an 18% difference. This inconsistency may be explained if we take a closer look at the definition of certified penetration, which dictates that 80% of the projectile mass must be on the other side of the target plate with a consistency of 80% for a given velocity. Due to the requirements set out in the definition of this term, the minimum permissible penetration of 3BM-22 can never be lower than its certified penetration of 380mm at 2 km at 0°, so the only explanation is that the actual average penetration (410mm according to Fofanov) of 3BM-26 must be much higher than its certified penetration would suggest. However, this requires us to make the assumption that Fofanov's figures are infallible and that the 18% figure is not directly based on Fofanov's own figures (which have been spread far and wide).

Despite total the obsolescence of this round as a front line anti-tank munition, "Nadezhda" is still used in reserve units in the Russian army to this day. Generally speaking, their fate is to be expended in live firing exercises. High readiness units in the Western military district have gotten rid of this shell long ago.

3VBM-13 (Vant)

3BM-33 (3BM-32 Projectile)

Having being informed of new Western developments of advanced composite armour, GRAU set forth new requirements to defeat future tank armour in the mid-70's. In 1977, work began on new APFSDS projectiles to accomplish this. The new projectiles would be based on totally new design concepts in order to avoid the limitations imposed by the previous cored designs. The first result of the development process was "Vant".

3BM-32 "Vant" is a depleted uranium projectile introduced in 1985. The depleted uranium-nickel-zinc alloy penetrator rod has a monobloc construction, and the projectile is aesthetically similar to the 120mm DM13 APFSDS shell. The "bucket" style sabot design from the 3BM-26 was carried over and modified, which meant that large bore-riding fins were still necessary. This was a problematic source of drag, resulting in the "Vant" losing more velocity per unit of distance compared to its counterparts.

The DU rod is made from an alloy based on Uranium-238. Although rather short, "Vant" could still be considered a decent round for the mid-80's as the relatively high thickness of the rod grants it a higher penetration efficiency for its length. The long rod penetrator is somewhat longer than the 105mm M774 and M833 and it was also substantially thicker, but it was outclassed by the American 120mm M829 (1985).

Muzzle velocity: 1700 m/s

Mass of the sabot: 2.07 kg Mass of the projectile: 4.85kg
Mass of the stabilizer fins: 0.435 kg

Total projectile length: 480mm
Penetrator rod length: 380mm
Maximum diameter of the projectile rod: 34mm
Average diameter of the projectile rod: 30mm
Penetrator L/D ratio: ~13:1

Penetration at 2.0 km:430mm RHA at 0° (From plaque) 

Penetration at 2.0 km:  250mm RHA at 60°
According to Mikhail Rastopshin in an article

Penetration at 2.0 km:400mm RHA at 0° 

According to Mikhail Rastopshin in another article

Penetration into spaced targets: 7-layer array at an angled of 60 degrees (630mm LOS) could be defeated at 3200 m. 7-layer array at an angle of 30 degrees (620mm LOS) could be defeated 3200 m. 3-layer spaced array at an angle of 65 degrees (1830mm LOS) could be defeated at 5000 m.

"Vant" is approximately comparable to the American M829, which began production in 1984 and entered service in the same year as "Vant" (1985) to equip the freshly inducted M1A1 Abrams tank. M829 is only 30 m/s slower than Vant at 1670 m/s, but M829 loses less velocity over distance due to its small, subcaliber low-drag fins and had a longer 540mm-long monobloc DU penetrator capable of penetrating approximately 275mm RHA at 60 degrees at 2 km. In terms of penetration performance, "Vant" is closer to DM23 than the M829.

3VBM-17 (Mango)

3BM-44 (3BM-42 Projectile)

Developed in parallel with "Vant", "Mango" has a more complex construction using jacketed tungsten penetrators instead of a depleted uranium rod. The 3BM-42 projectile has a two-part tungsten alloy penetrator, but technically it is a three-part penetrator, as the rod supplemented by a 112mm tungsten alloy segment at the tip with a diameter of 22mm - greater than the diameter of the main penetrator. The penetrator is encased in a thin steel jacket which holds the two long rod penetrators together.

The weakness of jacketed long rod penetrators is its reduced penetration power against homogeneous steel armour compared to a monobloc penetrator, as detailed in "Numerical Analysis and Modelling of Jacketed Rod Penetration". In general, decreasing the thickness of the steel jacket relative to the diameter of the rod to a ratio of 0.1 results in the smallest degradation of penetration against steel armour, and may actually increase the residual length of the penetrator emerging from behind the armour plate, albeit at a lower velocity compared to a monobloc rod. However, reduced penetration against homogenoeus steel targets was hardly an issue by the mid-80's given that the new generation of NATO main battle tanks were using spaced NERA armour.

The effectiveness of a high elongation jacketed long rod penetrator on two variations of ceramic armour targets and a type of spaced NERA armour was investigated in the study "Ballistic Performance Of Monobloc And Jacketed Medium-Caliber Penetrators Against Composite Armor And Spaced Targets" by H. Kaufmann et al. For the spaced NERA armour target, the first layer is a NERA panel made from a sandwich of 7mm Armox plates with a 3mm center layer of rubber. Behind this is an 80mm medium hardness steel plate and the final layer is a 10-20mm hard steel plate. The entire array is sloped at 60 degrees. The mass efficiency of the armour against the jacketed rod was 0.8 and 1.1 for the two test shots. As stated in the paper, the jacketed rod was unaffected by the NERA panel and did not break apart.

"The spaced armor targets show relatively low stopping performance against the jacketed rod. As x-ray pictures reveal, the tungsten cores pass the sandwich without breaking which explains the low protection effect."

The relatively high effectiveness of the jacketed rod on a spaced armour target is reinforced by other studies and probably explains the long ranges at which "Mango" was able to penetrate the multi-layered spaced targets. Monobloc tungsten penetrators were fired three times at the same spaced target but the details were not divulged. The jacketed rods also demonstrated superior performance against one type of ceramic composite armour and performed at least as well as the monobloc rod against another type. This hints that "Mango" is probably effective at defeating the spaced NERA armour known to be foundation of the special armour used in the M1 Abrams and Leopard 2 series of main battle tanks.

With that in mind, it is important to note that the use of a jacket was still primarily a method to maintain the integrity of the rod during acceleration in the gun barrel and during flight, which was commonly done for early long rod heavy alloy projectiles. In the case of "Mango", the steel jacket is thickest around the middle of the rod where the threads connect the projectile to the sabot. According to "Numerical Analysis and Modelling of Jacketed Rod Penetration", the common use of steel jackets on early long rod penetrators was due to the poor mechanical properties of the heavy metal alloys at the time. The most serious issue was the shearing of the threads that held the long rod penetrator to the sabot during acceleration inside the gun barrel when firing.

Since one of the long rod penetrators in "Mango" is shorter than the other, it is unclear why the shell is not longer than it is, as it should not be difficult to have two long rods instead of one long rod and one short rod. From various studies on the behaviour of long rod tungsten alloy penetrators on spaced armour and thin oblique plates, it is very likely that the short tungsten alloy segment at the tip of the projectile will prevent the rest of the rod from breaking up after perforating a spaced armour plate at high obliquity, or at least control the damage in such a way that the rest of the rod will penetrate any further armour plating with greater efficiency. The use of a separate tungsten alloy segment at the tip of the projectile was definitely a deliberate design solution meant to counter complex spaced armour, as there would be no limitations against producing a simple flat tip or a frustum for the shorter half of the two-part penetrator.

The segment at the tip is only partially jacketed, and is therefore largely separate from the rest of the projectile, so the damage sustained by the segment will be mostly isolated from the rest of the projectile. This will protect the integrity of the jacket, and thus preserve the performance of the projectile against a spaced or composite armour array behind the initial front armour plate. As the armour of both the Abrams and Leopard 2 are understood to rely on an array of NERA and steel armour plates, the effectiveness of "Mango" could be quite high despite the less technically advanced nature of the shell. The shorter half of the two-part penetrator is at the front, and may possibly give "Mango" the ability to counter dynamic protection such as Kontakt-5, besides generally improving its effectiveness against NERA.

German military expert, author and lecturer Rolf Hilmes has said that the German 120mm DM53 is specially constructed to deal with advanced composite armour and dynamic (reactive) armour. Its construction, he says, consists of a three-part tungsten alloy penetrator. If having multiple segments is a valid method of overcoming complex composite and reactive armour, then "Mango" may be more advanced than it appears at first glance. Of course, it should be stated that "Mango" is not nearly as advanced as DM53 in a variety of ways, but it is rather interesting to note that the design of "Mango" has more in common with the latest modern anti-tank ammunition than it has with other long rod penetrators developed during the 80's.

Photo from

As mentioned before, it is known that jacketed penetrators perform substantially better against spaced NERA armour and may perform better against certain types of ceramic laminate armour. The main weakness of jacketed long rod penetrators is its reduced penetration power against homogeneous steel armour, as detailed in "Numerical Analysis and Modelling of Jacketed Rod Penetration". The paper reveals that decreasing the thickness of the steel jacket relative to the diameter of the (depleted uranium) rod to a ratio of 0.1 results in the smallest degradation of penetration against steel armour, but the steel jacket for 3BM-42 is rather thick - much thicker than on the 3BM-32. The total diameter of the projectile at the middle is 36mm, but the tungsten alloy core has a consistent 18mm diameter throughout its entire length, meaning that the jacket is 9mm thick or 0.5 diameters. This hints at two possibilities: the jacket was intentionally designed for increased performance on composite armour at the expense of performance on monolithic steel armour, or they were simply not able to create an alloy that was strong enough to survive being propelled down a gun barrel. Quite frankly, the second possibility seems rather unlikely since the main issue with early tungsten alloys was the shearing of the threads joining the penetrator to the sabot, not the rod itself disintegrating. In fact, it is practically impossible for the rod itself to break apart during the acceleration stage. 

The jacket can clearly differentiated from the tungsten rod in the photo below, taken from Andrei Tarasenko's website, btvt.narod. 

Having such a thick jacket, the cross sectional density of the projectile is much, much lower than a monobloc tungsten alloy rod, but this is compensated by the high overall mass of the projectile of 4.85 kg, which is identical to 3BM-32 and compares favourably to the 4.6 kg of the 120mm DM33, which also had a slightly lower muzzle velocity of 1650 m/s.

The study "Comparisons of Unitary and Jacketed Rod Penetration into Semi-Infinite and Oblique Plate Targets at System Equivalent Velocities" by J. Stubberfield et al provides further evidence to show that a jacketed penetrator such as the 3BM-42 has enhanced performance against spaced armour. It was observed that the monobloc tungsten rod penetrated 12% more than the jacketed rod on the homogeneous RHA block, but the jacketed rod penetrated more than the monobloc rod by 17% on the spaced armour. According to the radiographs taken of the rods as they exited the spaced plates, the steel jacket on the jacketed rod was intact after the perforation and the jacketed rod itself appeared to be less fragmented compared the monobloc rod from the smaller quantity of debris. The spaced armour setup in the study was rather simple, consisting of only two oblique plates placed at an angle of 65 degrees, but even so, it is rather unlikely that a monobloc rod would perform better on an array with more spaced plates.

The 3BM-42 projectile is generally similar to the 3BM-32 in external layout due to the use of a similar "bucket" type sabot, but the projectile is significantly lengthier. The sabot is made out of a light V-96Ts1 aluminium alloy, helping to decrease parasitic mass and thus increase firing efficiency. The driving bands are made from plastic and help reduce barrel wear, although it is still quite high due to the high pressure of the shell. The long-rod tungsten alloy penetrators are encased by a thin sheath made from S-7 impact-resistant tool steel. It is known that jacketed or sheathed long rod penetrators have superior performance on composite armour arrays, because the sheath protects the rod from external perturbations and keeps it intact as the projectile passes through the array.

As you can see clearly in the photo above, "Mango" still has bore riding fins. The copper-coloured nubs on the apex of the fins you see above are copper ball bearings that contact the barrel bore and keep the projectile centered as it is propelled. Larger fins create more drag, leading to a lower velocity downrange.

The dimensions of 3BM-42 come from the table below. Original source unknown, and a mistake in the density of the tungsten alloy is noted. Alloy VNZh-90 has a density of 17.6 g/cc, not 17.0 g/cc.

Muzzle velocity: 1715m/s

Mass of Sabot: 2.2 kg
Mass of Projectile: 4.85 kg

Length of projectile: 574mm
Diameter of projectile (maximum): 31mm

Length of two-part core: 420mm
Length of tungsten alloy armour piercing cap: 112mm
Diameter of core: 18mm
Diameter of tungsten alloy armour piercing cap: 22mm

Penetrator L/D ratio: 20:1

Chamber pressure with
Zh40/Zh52:  443.8 mPa
Zh63: ?

EFC rating: 5

Penetration at 2.0 km:
520mm at 0°
230mm at 60°
These figures come from page 587 of the "Textbook of Means of Defeat and Ammunition" 2008 (Учебник Средства Поражения И Боеприпасы) published by Bauman Moscow State Technical University.

Penetration at 2.0 km:
210mm at 60° 
Mentioned by Mikhail Rastopshin in an article.

Penetration into spaced targets:
  • 7-layer array at an angled of 60 degrees (630mm LOS) could be defeated at 3300 m.
  • 7-layer array at an angle of 30 degrees (620mm LOS) could be defeated 3800 m.
  • 3-layer spaced array at an angle of 65 degrees (1830mm LOS) could be defeated at 2700 m.

Without knowing more specific details regarding these armour arrays, we cannot know how correctly they represent NATO armour at that time. It is very likely that the armour of the M1 Abrams will be defeated effortlessly by "Mango" at combat distances, since we now know the layout of the armour and some details of the steel plates it uses. It is also very likely that the thicker armour of the M1A1 can be still defeated by 3BM-42 at combat distances. Based on the limited information currently available, the Leopard 2A0-A4 is probably more resilient to 3BM-42 but still vulnerable.



9M119 "Svir"

The 9M119 "Svir" is a guided missile with a single 4.2 kg shaped charge warhead designated 9N142. The missile achieves excellent armour penetration for its caliber despite the limited overall length of the missile mainly because of the placement of the warhead at the rear of the missile, thus creating a large amount of standoff distance without the need for a special standoff probe. Because of this, a relatively high penetration of 700mm RHA was obtained from the 125mm caliber warhead as opposed to the typical 550mm RHA of penetration achieved by the 3BK-18M round. Another factor in the improved penetration is due to the increased caliber of the shaped charge cone as compared to a normal 125mm HEAT shell, which in turn was only possible because the missile is soft-launched out of the gun barrel and does not require a thick casing to survive the extremely high acceleration forces experienced by high velocity HEAT rounds. However, the missile spins at a very low rate which presumably reduces the performance of the shaped charge by a small amount. This could be a source of some loss in efficiency compared to a normal fin-stabilized HEAT round, although it is worth mentioning that fin-stabilized rounds also rotate at a very low rate during flight.

The missile is soft-launched by a 9Kh949 reduced load ejection charge, giving the missile some momentum before the rocket motor kicks into action. The piston plug is designed to properly seat the missile in the chamber. The total weight of the 9Kh949 charge is 7.1 kg.

The missile itself has an efficient layout with the rocket motor placed in the middle, the warhead at the very rear, and the control surfaces and mechanism at the front along with the fuse at the tip with the laser receiver and stabilizing fins at the rear. The stabilizing fins and laser receiver unit are covered by a protective cup that breaks away after the missile is ejected from the gun barrel. The cup protects the laser receiver unit from damage when the missile is rammed into the gun chamber (the chain rammer moves at a speed of 2 m/s) and also when the missile is launched out of the gun barrel. It also serves to contain the spring-loaded stabilizing fins until the missile has passed the muzzle of the gun barrel, whereupon the opening of the stabilizing fins breaks apart the protective cup. The cup also contains three electrical sockets which interface with three corresponding electrical contact pins located at the center of the 9Kh949 ejection charge, which can be seen in the drawing on the left, above. This is a datalink that transmits several firing procedure subroutines to the on-board guidance system of the missile immediately before it is launched.

The large distance between the fuse at the tip of the missile and the warhead gives the warhead a good standoff distance without the need for a special standoff probe. The layout enables the 125mm missile to have a comparable flight range as the 127mm ITOW missile and superior armour penetration performance, but in a much more compact package. With 700mm of penetration, "Refleks" is a much more serious weapon with a much better chance of defeating the new generation (at the time) of NATO tanks like the Leopard 2 and M1 Abrams, albeit from the side. The chances of defeating such tanks from the front with this missile are rather slim.

The missile uses a solid fuel motor, with four nozzles arranged radially. Flight stabilization is maintained via five pop-out tail fins with curved and angled surfaces to impart a slow spin onto the missile, while steering is accomplished by the two canard fins at the front. These are operated pneumatically, so the more corrections the gunner makes while the missile is mid flight, the less responsive the missile will be over time, though the gunner will have to be tracking a very difficult high-speed target like a maneuvering helicopter for this to become noticeable.

Guidance is accomplished by the integrated 9S517 modulated laser beam unit on the 1K13-49 sighting complex. The system has a maximum range of 4,000 meters, although there may be some difficulties in finding a target at such distances due to the limited magnification of the sight. The guidance commands are sent by modulated lasers divided into four sectors along the horizontal and vertical axis. This is illustrated in the drawing below.

This type of guidance is very difficult to jam. There are various guidance modes programmed into the 1K13-49 sighting complex, but the missile is always controlled in a SACLOS guidance regime. In the normal operating mode, the gunner must lase the target first to determine the range. Then, the gun is automatically elevated to a predetermined angle and the missile is fired. The guidance system directs the missile to fly at an altitude of 2-4 meters above the line of sight of the 1K13-49 sight, which is done partly to ensure that the missile does not come into contact with terrain features or other obstructions that may cause a premature detonation. At a distance of 600 to 800 meters from the target, the missile is lowered to the same level as the line of sight and guided directly toward the target, leaving it no time to react. This mode risks the chance that the target detects that it has been lased, but prevents it from detecting that it is being painted by a laser which would indicate that it is being targeted by a laser-guided bomb or missile. As such, this system prevents the target from realizing that it is being targeted by a laser-guided anti-tank missile. Also, it is possible to lase some other structure close to the target and not the target itself in order to prevent it from detecting any laser signature at all.

The alternate guidance mode is the direct fire mode which is akin to a simple SACLOS guidance regime. In this mode, the missile is launched out of the elevated barrel and immediately descends to the same level as the line of sight of the 1K13-49 sight. The missile is guided in a level trajectory in a straight line toward the target by the laser emitter which is slaved directly to the optical channel of the gunner's viewfinder. This mode is used when engaging helicopters and also when engaging targets that appear suddenly at short range, as this reduces the reaction time by 3 seconds because the gunner does not need to lase the target. This mode requires the modulated laser emitter to be aimed directly at the target until the missile arrives, thus allowing the target around 11 seconds to react if it is 4 kilometers away. There is also an emergency mode that is used when the laser rangefinder fails. This mode is essentially the same as the direct fire mode.

Missile Diameter: 125mm
Missile Length: 695mm
Wingspan (Stabilizer Fins): 250mm

Firing Distance: 100 - 4000 m
Flight Time to Maximum Distance: 11.7 s
Cruising Speed During Flight: 340 m/s

Penetration: 700mm RHA

Minimum Hit Probability: 0.8
Hit Probability On Tank-Type Target Cruising Sideways At 30 km/h:
100 m to 4000 m =  >0.9

The 9M119  missile was the longest type of ammunition available to the T-72 before the end of the Cold War. This fact is illustrated by the photo below (the missile shown in the photo is a 9M119M "Invar" with identical dimensions).

3UBK20 "Invar"


The 9M119M "Invar" is a follow-up to the "Svir" 9M119 missile. The rocket motor was improved, giving the "Invar" missile an additional kilometer of flight range over its predecessor and granting the tank a bigger standoff advantage over most potential targets. Also, the missile is now supersonic at sea level. Supersonic missiles are incredibly loud, of course, but this is not relevant during combat since the missile travels faster than its own sound signature so the target will not be able to hear the missile approaching until it has already impacted. The missile also features a new 9N142M tandem warhead. Externally, it is identical to its predecessor. "Invar" entered service in 1989 and is currently in service in the Russian ground forces.

Missile Diameter: 125mm
Missile Length: 695mm
Wingspan (Stabilizer Fins): 250mm

Firing Distance: 100 - 5000 m
Flight Time to Maximum Distance: 17.6 seconds
Cruising Speed During Flight: 350 m/s

Mass of missile: 17.2 kg

Primary Charge Caliber: 125mm
Secondary Charge Caliber: 46mm

Armour Penetration:
700mm RHA (Main charge only)


It appears that the primary charge of the missile is the same as the 9N142 warhead, and the only difference between the 9N142 and 9N142N is the addition of the precursor charge. The precursor warhead defeats ERA by fully perforating the ERA panels without initiating the explosive charge contained within, creating a relatively large channel for the shaped charge jet of the primary warhead to pass through unmolested. The primary warhead is timed to detonate 300 microseconds after the precursor.

The only issue is that the sights of the fire control system on Soviet tanks of that period did not have a high optical magnification, so seeing and targeting a tank-type target may prove difficult, especially in non-optimal weather conditions. The visibility issue of Soviet-era tanks was mostly solved by the replacement of the 1K13 sight with the Sosna-U thermal imaging sight in the T-72B3 modernization. However, the poor digital magnification capabilities of the sight makes it impossible to identify targets at long range so it will not be possible to distinguish a distant tank from other vehicles.


The PKTM is mounted as a coaxial machine gun. It is fed with 250-round boxes with seven boxes of additional ammunition, five of which are stowed on the cover of the autoloader near the commander's feet, while the other two boxes are stowed in the external stowage bin at the back of the turret. The practice of using standard large capacity ammunition boxes for the coaxial machine gun dates back to the invention of the triple stack 63-round pan magazine for mounted DP machine guns in 1928 and its subsequent use in DT machine guns in Soviet tanks to facilitate longer bursts of suppressive fire. When the DT and DTM was replaced by the SGMT, it was fed with 250-round boxes that were also standard for wheeled and tripod-mounted SG-43 machine guns. This did not change when the Red Army transitioned to the PK series, as the 250-round boxes for the PKT were also standard for tripod-mounted PK machine guns and for PKS machine guns on a variety of mounts. Proprietary high capacity boxes did not exist for the machine guns of any Soviet armoured vehicle until the advent of the BMP-1 which had a special 2,000-round container for its PKTM coaxial machine gun, but being based on the T-64, the T-72 inherited the traditional ammunition feed system of its parent design. Due to the use of individual boxes instead of a single large container, the commander must periodically reload the machine gun.

The barrel of the PK and PKM is rated for 200 shots fired continuously and 400 shots fired in bursts, meaning that one or two 200-round boxes can be depleted before a barrel swap is mandatory in order to avoid premature wear of the barrel and prevent a large drop in accuracy from overheating. The barrel of the PKTM weighs 3.23 kg which is over 0.8 kg heavier than the barrel of a standard PKM (2.4 kg). This increased the heat limit to 500 shots fired in bursts. Because tank gunners do not face problems with dust obscuration and recoil control when using the coaxial PKTM, the combat rate of fire tends to be higher than the rate expected from an infantry machine gunner, but just like infantry machine gunners, tank gunners are still trained to fire in controlled bursts. Combined with the pauses in firing caused by the occasional reloads, the barrel of the coaxial machine gun practically never needs to be swapped during combat.

The PKTM is mainly distinguished from the earlier PKT by the smooth barrel as opposed to the fluted barrel of the PKT. Internally, the PKTM and the PKT differ in the same way that the basic PK and PKM models differ. When the PKM replaced the PK on the production lines in 1969, the production of the original PKT also halted, so practically all T-72 tanks were equipped with the newer PKTM. 7BZ-3 API (armour-piercing incendiary) rounds with the B-32 bullet and 7T2 API-T (armour-piercing incendiary tracer) rounds with the T-46 bullet are linked in a 4:1 ratio. The machine gun has a cyclic rate of fire of 700 to 800 rounds per minute. The coaxial machine gun can be fired either by depressing the trigger button on the gunner's handgrips, or by pressing the emergecy manual trigger button located on the trigger unit installed at the back the receiver of the machine gun.

It is fired by the gunner using his "Cheburashka".

Notice the cable leading away from the PKT to the left

The machine gun is mounted to the right of the main gun, and protrudes from a pill-shaped port which provides vertical space for gun elevation. A conical flash hider is installed on the muzzle of the machine gun. It is unclear why the original long birdcage-style flash hider of the PK was not used, but it is possible that the sideways deflection of the muzzle blast caused by the birdcage-style flash hider were damaging the fabric gun mantlet cover.

The coaxial machine gun is only a limited solution to the problems posed by dispersed enemy infantry, especially if hard cover is available. In practice, the coaxial is only useful in very specific situations, and desirable only when HE-Frag shells are not suitable due to concerns of collateral damage or ammunition wastage.



The T-72 is equipped with the ZU-72 anti-aircraft installation with an NSVT heavy machine gun. The machine gun is primarily intended for the anti-aircraft role, though it may be used to shoot at ground targets as well. The ZU-72 has a range of elevation of -5° to +75°. The cantilever mounting of the machine gun is balanced by a pair of springs affixed at the center of gravity of the machine gun. This makes it much less physically taxing on the commander to control the elevation of the machine gun, which is done by working a flywheel on the right of the machine gun mount as shown in the photo below. It has been reported to the author that the elevation mechanism is extremely smooth and light, although rotating the cupola takes a bit more effort.

Due to the lack of a motorized traverse mechanism, the entire cupola must be rotated manually by the commander to aim the machine gun in the horizontal axis. Firing is done by squeezing a trigger paddle on the left handle. The commander has to stand on his seat in order to reach the machine gun. The elevation flywheel has a braking button to hold the machine gun at a fixed elevation while shooting to ensure better accuracy and to ensure that the machine gun does not jump when it is fired. If the brake is not activated while shooting, the machine gun may experience overwhelming muzzle rise due to the cantilever mounting of the machine gun.

As mentioned before in the "Commander's Station" section of this article, the machine gun is mounted on a race ring that can spin independently from the rest of the cupola, ostensibly allowing him to fire the machine gun with a modicum of frontal protection from the hatch. However, this is impractical in real world conditions.

The photo above is an excellent example of this feature being demonstrated. In this position, the commander cannot reach the trigger on the other side of the machine gun mount. This is only possible if the hatch is on the right side of the machine gun like on the T-72 in the photo below. However, this prevents the commander from reaching the elevation flywheel so he cannot adjust the machine gun to correct his aim, so firing with the hatch in front of him is generally impractical.

Unfortunately, the IR spotlight prevents the machine gun from being aimed when it is traversed to directly in front of the cupola, although it is possible to traverse the machine gun 360 degrees by elevating it to its maximum elevation to clear it from the IR spotlight. In the travelling position, the race ring for the machine gun mount is locked facing rearward by a spring loaded plunger, marked (18) in the diagram below. The inner cupola - which carries the commander's optics and hatch - runs on a smaller race ring along the intermediate band.

When not in use, the machine gun is kept in the travel position, meaning that the inner cupola rotates without bringing the machine gun along, making it lighter and easier to spin around when the commander is surveying his surroundings. When the plunger locking the intermediate band to the fixed base is released, the machine gun is allowed to traverse freely along the race ring between it and the fixed base. The inner cupola may be locked to the intermediate band or left free. In the former case, the cupola rotates with the machine gun, so spinning the machine gun to face the front would spin the cupola to face the rear. This is the normal combat procedure, because it gives the commander complete access to the machine gun and allows him to reload it more easily. In the latter case, the position of the machine gun relative to the cupola can be changed as the commander wishes. It is possible for him to open fire on either side of the turret (at strafing aircraft, for instance) while keeping the cupola facing where bullets are expected to come from.

The machine gun is complemented with a K10-T collimator sight to facilitate aiming at aerial targets. It is tinted to reduce glare when aiming in the direction of the sun and the illuminated reticle enables the sight to be used in low light conditions.

Using the collimator isn't compulsory. If it is damaged or unsuitable, the iron sights on the machine gun may still be readily relied upon. Using the iron sights is preferable when using the machine gun on ground targets, as the collimator is offset from the bore axis of the machine gun and does not permit rapid range adjustments. 

By equipping the machine gun with a separate collimator sight together with the foldable iron sights, the operator is given the option to use whichever aiming device he chooses for the occasion.

View through the sight
The collimator projects a clear, crisp aiming reticle

The NSVT has a rate of fire of between 700-800 rounds per minute - somewhat faster than the M2HB, M85 or the earlier DShKM installed on the T-54 and T-62. The higher rate of fire improves the chances of hitting a fast-moving aerial target and the substitution of the prominent muzzle brake of the DShKM for a conical flash suppressor undoubtedly improved the commander's shooting experience, not to mention improving his vision when firing the machine gun in low light conditions and reducing the likelihood of being detected by aircraft. The nominal maximum effective range of the NSVT is approximately 800 meters against aerial targets, but this is circumstantial. Obviously, the probability of hitting a hovering helicopter would be much higher than hitting a moving fixed-wing aircraft.

As a rule, anti-aircraft machine guns can cause some amount of damage to low flying aircraft but are more or less useless for shooting down aircraft. Although it isn't difficult penetrating some of the more obvious weak areas such as the plexiglass windscreen on a helicopter, the chances of actually hitting a fast moving target are rather slim. On the contrary, the role of an AAMG is to be a deterrent: its objective is to deter the enemy pilot into pulling back from an attack, or perhaps even make him miss his shot. Serious anti-aircraft work is to be carried out only by the SHORADS (Short Range Air Defence Sytems) accompanying the T-72. In real world conditions, the NSVT on the T-72 has probably never been used against aircraft at all but has been occasionally used to target infantry, but even so, such occasions are rare as the commander is always aware of the danger posed by snipers as he must be exposed to use the machine gun. 

The machine gun is fed from a 60-round box secured to the ZU-72 mount on the right of the NSVT. A standardized mix of B-32 API and BZT API-T ammunition held in 10-round belt segments is carried in each box. Three additional boxes of ammunition are stored in metal bins in the turret and two more are strapped to the side of the turret just next to the commander's cupola. These two boxes are the easiest to access from outside the hatch as the commander can reach down and place a box straight on the machine gun mount if he is facing forward. Including the box of ready ammunition mounted on the ZU-72 anti-aircraft installation itself, the total ammunition load carried by the T-72 is 300 rounds.

At a range of 500 meters, B-32 armour-piercing rounds have a penetration of 16mm at 0 degrees and 10mm at 30 degrees. This is enough to pierce the side armour of any armoured personnel carrier from practical engagement distances and eviscerate unarmoured utility vehicles including trucks and jeeps. The higher penetration of 12.7mm rounds compared to the coaxial 7.62mm machine gun also makes the NSVT useful for shooting at troops hidden behind the cover of concrete walls or other types of obstructions.

The commander has to pull a large charging lever to cycle the gun (pictured below). This was a unique design feature of the ZU-72 mount to accommodate the rack-and-pinion charging mechanism of the NSV machine gun. Spent casings are ejected forward where they will roll down the sloped turret roof and off the tank. A canvas spent belt segment catcher is secured to the left side of the ZU-72.

According to a Czech study concerning the effect of typical machine gun bullets on the aluminium skin of common aircraft and the modeling of these effects, the thin fuselage skin of helicopters like the Mi-8 and other Czech aircraft presents no challenge to 7.62x54mm and 12.7x108mm B-32 bullets.

According to the table below taken from the study, the velocity limit for 12.7mm B-32 API bullets for the skin of an Mi-8 is just 40-58 m/s. The corresponding distance for this speed is the maximum firing distance of 6 kilometers, so technically, the NSVT is capable of piercing the skin of a utility helicopter from 6 km away. The thickest parts are the flanges at the two ends of the fuselage, measuring 3.7-4.7mm thick including the skin and flange itself, but this only represents a fraction of a percent of the surface area of the helicopter. Due to the very low velocity limits and large entry holes, a B-32 bullet is guaranteed to penetrate at any distance up to the maximum of 6 km and the incendiary content in the tip of the bullet will be expended inside the aircraft fuselage where it will do the most damage. Even so, the small incendiary content makes the B-32 bullet a second-rate option against aircraft. To sum up, the NSVT is easily capable of penetrating the thin skin of utility helicopters and common fixed-wing aircraft at distances far beyond the abilities of the commander to reliably engage them.

The MDZ bullet would be a much better option against aircraft, but it is difficult to ascertain if it was ever supplied to tanks. As it stands, using the NSVT on helicopters with a typical mix of B-32 and BZT rounds is unlikely to amount to much unless a vital component were hit, such as the engine, control systems or the pilot.

The lack of a remote aiming and firing system for the machine gun like on the T-64 has been said to be one of the greatest drawbacks of the T-72, and for good reason. Former Russian tank crews who served in Chechnya mentioned that it was suicidal to man the machine gun when in combat, so despite its high power and high rate of fire, the NSVT is not suitable for suppressing manpower in urban environments. Tank crews in Syria have also never been observed to use the machine gun during urban combat, for the same reasons. It is interesting to note that Leonid Kartsev defended this design decision in his memoirs, stating that it helped to have an unobstructed field of view when firing at aircraft, but this is a rather weak argument since the commander of a T-64A could fire his NSVT from outside of his hatch as well if he desired. It was no coincidence that the ZU-72 anti-aircraft machine gun mount was replaced with a remote control system in the T-90 almost identical to the remote control installation of the T-64A.

Due to length restrictions, this article has been divided into two parts. Part two is available here.


  1. Excellent article. It makes me wonder what the Russians have in the T-14.

    1. Please notify me if there are any errors in the article. I've been very busy for the past few months, so I haven't really had the time to proofread the article before I posted it. I am constantly updating, but I don't think I've gotten rid of half of the mistakes in there.

    2. I have checked everything so far but have not seen anything wrong yet.

  2. I just did a brief writeup on the T-14 you may find of interest:

  3. Hello Tiles,

    I am asking something that has been on my mind for a while but constantly forget to ask. On the section regarding the T-72B's armor protection on the first image of the destroyed T-72 where do I exactly look to see the spaced armor array?

    Secondly on the section regarding how Soviet Bulging Armor on the two images comparing Forward and Backward moving plates seems mixed up. I think your description is good but it seems to conflict with the labeled images. Are the images wrong or are your descriptions wrong?

    BTW I hope your search for someone to take over the blog goes well. :)

    1. Hi!

      Well, let's just say that the "ramp" that the burned-out chassis is on isn't really a ramp. Notice the thing at the guy holding the RPG's feet - that's the front hull armour ripped right off from an ammo explosion. Notice the idler wheel socket.

      No, that's how the research papers I've read describe them. Backwards means moving in a direction opposing the cumulative jet, meaning accelerating towards it. Forwards means accelerating in the same direction as the cumulative jet. It is viewed from the cumulative jet's perspective.

      Thanks. But even that's not going that well. I've got one guy who seems willing, but I always forget to chat him up. I've got an even bigger workload now, and this blog project is very, very low on my priority list. However, having seen the overwhelmingly positive feedback from readers, I try to update my existing articles as often as I can, and maybe add a few paragraphs to the ones that are still on-the-way. I've got 18 article drafts, and most of them are half-finished. Maybe in September I'll binge upload 5 or 6 of them :)

    2. Ah. It all makes sense now. My bad. :(

      Your work is excellent and I wish you well. We will be waiting patiently for more.

  4. This comment has been removed by the author.

  5. Did Algeria modernize its T-72s to the T-72M1M standard?

    1. As far as I know, they did not. Algeria did receive a number of Relikt kits (some 44 going off of memory), some of which have been fitted to in-service T-72M1's.

      Hope this answers your question.

  6. This is a fantastic article. You should write a book!

  7. This comment has been removed by a blog administrator.

  8. Good article but I couldn't help feeling that there was some bias towards the t72 in this article. Where are the pictures of the burn out t72s, the section on the weaknesses versus the strengths of the vehicle etc.? I feel like the problematic nature of the ammunition placement deserves more coverage here

    1. Richard Turner,
      I totally desagree. I am a military historian and former profesional soldier and I have served on the Leopard-2. This article about the T-72 is by far the most objective, solid and truth worthy I ever had the pleasure to read. In the military world we need much more propaganda-free and objective articles about military equipment. The T-72 for exmple has always been a big victim of western propaganda that on purpose lied about relevant facts about this tank. Here in the west we always tend to subestimate the russian equipment and in the wrong situation it can get you killed.

      By the way a Leopard-2 or a Abrams also stores ammunition in the chasis, if this ammunition gets hit it easily can kill the entire tripulation and even blow off the turret as well, but I agree that this is easier to happen in a T-72 compared to a tank that has stored the ammo in the same place instead of all over the compartment.

    2. This article is not objective by far but thanks for your comment. Ill write a response to tiles later there are certain things he said that I both agree and disagree with

  9. Check those photos you used on have to have a clear idea about which is A and which is M/M1....

    1. Well, in the context in which those photos were used, it is irrelevant if they were M1s or As...

  10. A fascinating account, thank you! A question though: the frontal fuel cells acting as side armor for the driver. I would have thought diesel fuel would be touched off by a penetrating HEAT jet?

  11. All i can say is that thank you for this marvelous detailed article regarding T-72!

  12. This comment has been removed by the author.

  13. "Equipped with Kontakt-5, a late model T-72B, most likely the obr. 1989 model, has successfully resisted a TOW-2 missile hit to the upper glacis plate."

    Any idea if the contemporary 1985-1989 M1A1 and Leopard2A4 can withstand the tow 2?

    1. They lack any type of ERA but their turret cheeks are rated at no less than 1000mm RHA equivalent to HEAT. OTOH TOW 2 is said to penetrate up to 900mm RHA so the clear answer would only be produced by live-fire tests. Also TOW 2A's precursor charge will offer an advantage.

  14. Excellent article, but you have forget to mention the 3BM46 Svinets APFSDS round in the ammo section.

  15. This comment has been removed by the author.

  16. "However, this is counteracted by the very poor sound quality from the headphones installed in the helmet. It is more pleasant for the commander to shout commands to the other crew members rather than use the intercom."

    I have to disagree with this. I have experience of intercomm systems of BMP-1/2 and CV9030FIN (which uses finnish system with sovier/russian tanker headgear) and system sound quality is fine. Of course it removes personal features of voice, but transmits command itself clearly. On the other hand sounds absurd that you could communicate by shouting. There is no change of driver hearing your shouts from turret and even gunner you have to kick first to get his attention. This as commander of previously mentioned vehicles. But seems to be well written article otherwise though i don't know much T-72 itself.

    1. I must agree that the intercom fulfills its purpose quite adequately, but I'm sure that it isn't pleasant. The speakers in the headphones are of the simple "moving iron" type using a fixed electromagnet and a steel disc as the diaphragm. There is some damping, but the audio is very low fidelity.

      As I understand it, communicating with the driver without the intercom while the tank is moving is impossible for the commander, but communicating with the gunner is not difficult according to a Russian ex-tanker I interviewed. When the tank is just idling the sound level is a bit lower and it is possible to speak normally. Hatches must be fully closed in all cases, of course. However, I am sure that there are some nuances that I may not fully understand, so I will really appreciate it if you shared your experiences with me. My email is available in the Contacts page.

  17. Just wonderful, such huge article, but almost each paragraph dedicated to description of tank`s systems contains errors or wrong conclusions.

    1. Well, I'm sure that the article could be improved. You could help by pointing out errors.

  18. Hi, great article. I think there is a typo. Is second BK-25 "M" or "B"? Or are they both named like that?

    1. Typo. I'm thinking of removing that part of the article since it's basically just guesses.

    2. OK, thanks for the response and article :-)

  19. Hi Iron Drapes,

    check this out:
    "T-80UK and T-90 MBTs are equipped with AINET system that allows to electronically fuse HE-FRAG rounds to explode at predetermined moment of flight. In order to use the system the gunner must lase the target before loading the round into the breech. The round is passed by the auto-loader through an automatic fuse setter, which sets the fuse to explode at the correct distance; the fused round is then loaded into the gun and is ready to be fired. This system allows to efficiently use HEF rounds against hovering helicopters as well as infantry and light armor in entrenched positions, out to 4 km and more. The effective fragmentation radius and range consistency improve three-fold, while ammunition expenditure for a typical mission decreases two-fold. All HE-FRAG rounds are compatible with this system, provided a new electronic detonator is used instead of the standard V-429E."

    Do you if any T-72 variant has ever become the AINET system?


    Kinds regards

    1. Ups, I forgot to insert the word "know"....

      So here comes my question again:

      Do you know if any T-72 variant has ever become the AINET system?

    2. AINET is vaporware. It was developed in 1988 and was planned to be installed in the (then) prospective tank designed that turned out to be the T-90, but it was never actually fielded. The laser rangefinder technology back then was not precise enough to reliably detonate shells above point targets. There isn't a T-90 currently in service that actually has it installed, let alone a T-72.

    3. That´s weird, there is a lot of sources that say that...

      Although in the promotion video of the T-90MS of Uralwagonsawod there is no mention of AINET system or similar, they just mention the range of HEF ammo.

      I also haven´t found a video of a T-90 firing that kind of ammo. So I guess you are right. Thanks

      On the other hand technically it shouldn´t be such a big problem. The Carl Gustav M2 was a anti-tank system of 1964, and it could already fire HE-ammo with a specific detonation point.

    4. The Carl Gustav M-2 itself has no fire control system that allows air bursting ammunition to be fired accurately. For ammunition like the FFV 441B, airbursting is achieved by using a mechanically timed fuze that had to be manually set before it was loaded and fired. Estimating the distance to the target had to be done the old fashioned way.

    5. Yes, I know that. I have fired it and it was no big deal employing it correctly.
      That´s way I meant that installing such a system into a tank that is a few decades newer shouldn´t be a problem at all.

    6. From a technical standpoint, it shouldn't have been difficult to achieve with modern technology, but the Soviet electronics industry wasn't really all that "modern". It would be trivial to implement such a system using today's technology, but then there are other obstacles. You know what the Russian military is like, and what the country is like as a whole.

  20. Excellent article, as ussual!

    Just one note: I believe SOSNA-U installed in T72B3s uses the newer Catherine XP model, instead of the previous FC one. Russians ( and Belorussians) have been mounting/manoufacturing this model since 2012.

    XP offers some advantage over the previous model, such as a slighty highter image resolution (768 x 576, vs 754 x 576 in the FC) and supposedly longer detection/recognition ranges, this acording to Thales.

    I suppose with that in mind, perhaps the final perfomance of the SOSNA-U system might be somewhat improved?


    1. I have also read that the production of the Catherine-XP began in 2012, but there is no real proof that it was implemented in SOSNA-U for the T-72B3. Only the Hawkeye panoramic sight is known to have it.

      It's certainly possible that newly produced SOSNA-U sights could have the Catherine-XP, but as a rule of thumb, it is much more profitable to be a pessimist. You are less likely to be disappointed that way :)

  21. Hi Iron Drapes,

    I was reading through your articles again and I have a question, because I couldn´t find it.

    The main difference between a TPN-3M and the TPN-3MK is the 2°gen image intensifier, but
    what is the difference between the TPN-3 and the TPN-3M commanders sight?

    Do you know the introduction year of the TPN-3M (1973?) and TPN-3MK?


    1. TKN-3 has no image intensifier (only active IR illumination), TKN-3M has the 1st Gen image intensifier and TKN-3MK has a 2nd Gen image intensifier.

      Unfortunately I do not know the introduction dates. Sorry.

    2. Iron Drapes.

      Oh I thought all had a image intensifier...No that makes now much more sense. So thanks a lot.

      Don´t worry about the introduction dates I can more or less figure it out with the tank they were used, thanks to your work.

      Kinds regards

  22. in the end, which APFSDS can the older T-72B3 use?

    1. so the old T72B3 has the 2A46M cannon and svinets while the T-72B3 mod 2016 has the 2A46M-5 and the svinets 1-2? Also will all the non-upgraded T-72BS be upgraded to the Mod 2016 level?

    2. The older T-72B3 has 2A46M-5, but the T-72B3 obr. 2016 is specifically reported to be configured to fire Svinets 1 and 2 in a memorandum from UVZ to the Russian MoD. I highly doubt that every T-72B in Russia will be upgraded to a T-72B3.

  23. can you send me optics data for t-55 and t-72 tanks .

    thank you