Amizaur

Great article, unfortunately the video is said to be not available in my country to see. This part sheds some light on a subject of how a tank crew could act, and what came trough average (?) crewman's mind, when their tank was hit/penetrated: *** "I was firing the 75mm gun at a German tank next to a barn. I had fired one armor-piercing round when all of a sudden we received a direct hit to the turret. The shell hit the cupola ring and a flash of fire hit my periscope. The shell blew the tank commander's hatch open, took part of his head off, and then proceeded to blow off the anti-aircraft gun and mount. Bill Hey was killed instantly and he fell down on my back, covering me with blood. By the time I could get Bill off of my back, the assistant driver had bailed out and the loader had crawled through the turret and out the assistant driver's hatch. Instead of checking to see where the gun tube was located, all I could think about was getting out of the tank, since when they hit you once they generally keep hitting you until the tank catches fire." *** The tank was almost undamaged, only the cupola was penetrated and unfortunately, commander killed in a gore way - no one seemed to be hurt, but all the crew (even those who didn't see it and were not covered with blood) wanted to abandon the tank... They didn't try to fight on, even not think about it, didn't try to find where the shot came from, they just wanted to save their lives. Because they knew the next hit will arrive in few seconds, just a time needed for enemy tank gun to be reloaded, and most probably it would be more accurate and deadly... And that it will penetrate their armor without any problem... Sometimes high-motivated crewman, like fanatic SS crewman may want to continue fight in it's thick-armoured Tiger tank, after first penetration, believing the tank can maybe withstand the next one, but even such crewman would probably not want to stay and fight inside a penetrated PzIV... He knows he will die if he stays in, because next shots would penetrate too and kill him... Same would be true for Shermans and usually for Panthers too... Yes, sometimes even a highly motivated Sherman crew after not-so-deadly penetration would try to reverse (or even fight!) instead of bailing out... but, it should happen really _SOMETIMES_... Not usually... This is not what I see in the game...

Believe me, you WOULD NOT want to be close to 30 grams of RDX going off, even less so if those 30g were closed in a steel case, and you absolutely would not want to be in a closed space on few m^3 in which such a charge goes off. 30g of HE may be not very impressive in an open, watched from far, but in closed space, or if it was used to propell steel fragments, such "small" charge is deadly.

Calibers like 37mm, 45mm, 47mm, 50mm needed burst charges to be more effective. Basing solely on kinetic effects they were able to knock out enemy tank (kill or injure it's crew) but often it took multiple penetrations. With good and working burst charge shells like German 50mm were said to be MUCH more effective after penetration and capable of one-shot knock-outs, especially if the penetration was barely achieved, with such low energy that it had small chance to be effective. The caliber, muzzle velocity and energy of AP shells were raising quickly and soon shells like high-velocity 75/76mm were effective ENOUGH with only kinetic energy, and guns like 88L71, 17pd and probably 75L70 as well were penetrating with such velocity and energy that burster would not add much to the equation. On the other hand, IF equipped with working bursters those rounds were even more deadly and effective, detonation of the burster was increasing number of shrapnel, chances for fuel or ammo ignition, was creating deadly overpressure peak that alone could kill in confined space, and filled the inside of the penetrated crew compartment with thick smoke from detonation products... Those additional effects increased casualties, but - even if burster detonation didn't add much ENERGY to the total equation (the kinetic energy was far greater), then I guess that it greatly increased chances for making the crew of penetrated tank willing to abandon it (because of all this explosion, shock, burns, wounds, perforated eardrums, smoke, maybe fire ect). Sometimes I find amazing that anyone was able to survive detonation of medium size burster inside a buttoned tank, in a cramped space filled with mechanisms, human bodies, lots of ammo cartridges and sometimes even fuel tanks (T-34), with not much free space and the crewmembers relatively close to each other. If typical burster of medium-caliber AT round had 30-60g of HE, then it's like a defensive fragmentation hand grenade detonating INSIDE of the tank - with smaller number of fragments, but similar HE effect. I guess that at least some of the survivors had enough luck that burster of the penetrating shell for some reason didn't detonate, so they only experienced kinetic effects.

IMO behind-the-armor effects are too weak in the game. To achieve good chance of causing enemy crew casualties and firt-penetration knock-out you need very powerfull/energetic gun like 17pd or 88L71. Anything less (like 76mm, 75L48) achieves only medicore performance even against thin, side armor (thin enough, to ensure clean penetration with little loss of shell velocity, but thick enough - like 40-60mm - to produce good amounts of spall and shrapnel). Both kinetic effects and burst-charge effects (if they are modelled) seem undermodelled for me. It's a feeling based on in-game observations and knowledge of numerous tanker memoirs, statistics, and physics. Some testing I've done showed number of casualties in Shermans almost as high as historically documented, BUT I used solely 88L56 to achieve penetrations, and only against Shermans. My later game experience showed that less powerfull guns achieve much lower casualty rate (extreme example being those 47mm guns), and that German tanks like Panthers/Tigers seem to be more resistant (have lower chance of casuality happening if penetrated). So - overall - the casuality rate is too low IMO and the formulas used should be revised. Especially in case of side penetrations.

Hmmm... It would be interesting to check: Static Sherman vs moving Sherman - and note the spotting times Static PzIV vs moving PzIV - as above Moving Sherman vs moving PzIV Static Sherman vs static PzIV. The we would know what is the effect of movement vs static, based on two examples, and what is the effect of tank type (advantage of Sherman vs PzIV). Maybe the Sherman has JUST a big spotting advantage and nothing more. But maybe it's only related to moving/static as you observed in your first test.

Of course that if there are some nice bushes available that you can hide your AT gun in, then there is no need for additional camouflage. Even without additional camouflage, the gun will be hard enough to spot. Wit camuflage, it would be even harder. But if the crew had to somehow set up their gun in area where there is no good natural cover, they would make their best to camuflage the gun SOMEHOW. This is where the camuflaged gun model is needed most - where there is no good natural cover.

WW2 AP didn't loose much of it's penetration because of layered armor. The biggest effect would be detonating shell's burster charge, but still the remaining part of the 76mm or 17pd shell in most cases should penetrate inner 40mm of armor. edit: well, especially the 17pd, which didn't have burster at all so it wouldn't lose almost any penetration after passing a skirt.

And there would not be "penetration" text displayed ? Even if the underlaying side armor was penetrated ? Uh... That could be fixed... To check if the side armor was really penetrated, we should make the test c3k described, and see if there are any crew casulties happening after skirt hits. Systems degradation maybe could come from external hits, but for crew being killed there would have to be a penetration or heavy spalling.... We could also set a Jagdpanther to fire at side of a Panther and see if any shots get's completly trough that would be a definite answer

The AI units COULD coordinate their movements, but this would require writing additional code . The Hunt command COULD work, if there was HUNT ARMOR option, so they would not stop firing on infantry contacts, the "engage" logic could also be improved - for example the tank could stop to engage ONLY if certain conditions are met - leghth of the contact, it's size ect. It could stop to engage important contacts that are visible at least for xxx seconds and not for a split of second. Everything can be fine-tuned. The curent rules of emgagement ect. are very basic. My point is - it COULD be done, but of course not with the currect code and current set of orders and their simple rules. It would have to be improved and enchanced. And it would require additional work. But please don't say it just can't be done .

Then how they managed in real-world to make any coordinated tank attack in WW2, shooting from short stops during an assault ? I would think, it should be easier to coordinate AI units in computer game, than to do it in real life

If I can join with similar questioon - I would like to ask BF what is the chance of immobilisation if an AP shell hits the track or suspension ? It should not be 100% or even close to that. Most of suspension/wheels hits would be ineffective, only produce minor damage. Some types of tracks, especially very heavy and wide tracks of PzV and PzVI also could take AP penetration and be only damaged, weakened, but do not loose mobility. I hope that chance for immoblisation depends on shell size and mass of the target tank (heavier the tank, heavier, wider and more robust it's tracks), for 76mm against for heavy Tiger/Panther tracks I would expect it to be about 50% chance for immobilization, the rest 50% of hits resulting in heavy track damage but no mobility kill. For suspension hits, against Panther/Tiger running gear, I wouldn't expect more than 20% chance for immobilisation from single hit, while 80% doing relatively little damage... I had a Tiger immobilised after a sinle bazooka round hit it's wheels. I hope it doesn't happen too often.... Not easy to test. One would need few hundreds/thousands of shots to get about a hundred track/suspension hits to get some idea of what are the probabilities. The issue with Panther side skirst is an interesting one. I've seen ineffective AP hits on skirts myself.

The slow-down would depend on amount of armor that was penetrated. If a Tiger I round, with penetration potential of about 150mm RHA from close range hits a 40mm plate of armour, it would not slow down very mucb after penetrating it. The thin plate, falling critically, would not take much energy from it. Let's assume the shell was going about 750m/s when it hit the first armor, with penetration potential about 140mm of RHA. That would be about 300m range. Penetrating an overmatched thin plate would require less than "40mm" of it's potential - several close-spaced thin plates are not offereing the same resistance as single thick plate of the same overall thickness. Let's assume assume the shell "lost" about 30mm of penetration potential after that. So right after penetration, it has about 110mm of penetration potential left. What velocity would that be ? 110mm of penetration a Tiger I shell has at about 1500m. It retains there about 640m/s. So the slow-down was from 750 to 640m/s after penetrating 30-40mm of armor. Now the burster detonates, taking away 1/3 to 1/2 of shell mass. How would this affect penetration ? I guess it would be severly degraded, maybe even halved (the shell cross-section density was severly degraded), so we got about 60-80mm of penetration potential remaining, maybe more, maybe less, depends if the shell is tumbling. Now the shell has to get out of the tank, it won't have much left after that. But still can penetrate a halftruck parked right after the tank. If the shell was undamaged, I mean fuse didn't work, or the armor penetrated was too thin to trigger it - like 10-15mm - there is another story, full 110mm of penetration left or more (in case of 10mm of armor, less than 10mm would be "taken away" from shell penetration potential, full 140mm still left, and the slow-down would be minimal), and still 70-80mm of penetration left after leaving the tank trough the second armor - so it can still easily kill another one trough another 40mm of side armor. As for trajectory change - all depends on hit angles. If the hit was at angle, the shell would change it's trajectory a little (against thin armor) or more (against medium armor). If the hit was perpendicular, from 90deg side angle - there would be almost no change in trajectory, the forces are symmetrical.

Even if the fuse was triggered and the burst charge detonated, the shell would not turn into cloud of splinters. Just it's rear part would be fragmented, the whole massive front part - probably more than half of shell mass - would just continue on it's path, retaining still tremendous penetration potential, at least for several meters. After more than several meters or after penetrating second obstacle (getting out of the tank), it would most porbably start tumbling and it could hit next targets flying side-on or even backwards. But STILL retaining some penetration potential against light targets like halftrucks. It's still several kilos of hard stel flying several hundredfs meters per second. The game overmodels the shell's after-penetration potential, probably the shell after penetrating loses only the amount of energy it spend on penetrating first obstacle and retains the rest. Instead, it should get additional penalty and have it's remaining penetration/energy decreased by some smart formula that would account for loss of mass after burster detonation and for loss of stability (tumbling) after first obstacle, depending on the obstacle's armor rating and randomised a bit. Game should also model additional energy loses if the shell hits engine, transmission, gun cradle and other things like that, while penetrating a vehicle. Tanks are not ampty boxes of steel. Not sure if the penetration of the second armor plate - getting from inside out - is currently modelled. One could check it, trying to penetrate trough, from a rear, a vehicle with very thin rear armor and very thick front armor . Multiple penetrations ARE possible and realistic. The shell, even after detonation, doesn't desintegrate completly. Multiple penetrations in CMBN are just too optimistic, but not entirely wrong.

I wish there were contour lines (switchable on/off) programmed into 2.0 or maybe 3.0...

Thanks for the file. I've run it and got many weapon hits. Indeed the Tiger I seem to have more than one "hitpoint" for the barrel. Not only at the end of it (the "muzzle brake") : but also on the thick part of the tube close to the mantlet: I did my tests tested initially with Tiger I, but with the barrel pointing at shooting tanks. Then changed the target tanks to Panther and Tiger 2 at the time I checked rotated barrels, and then I didn't notice those additional hitpoints. Probably every tank has little different number and positioning of barrel "hitpoints" or "hit areas". With your setuop I got some hits at the "muzzle brake", every shell that was little too high and would normally go just over the roof of the turret, had good chance of hitting the "muzzle brake" if it was flying over center of the Tiger turret: As you can see above, the shell after disabling the Tiger's gun, continued on it's path just above Tiger's turret roof and hit the ground far behing the target. But most "wepon hits" were at the thick part of the Tiger I barrel, that is close to the mantlet: So, at least against Tiger I, weapon disabling hits are also possible in that place. I suppose that Panther don't have that hitpoint, as it's barrel don't have the thick part. Not sure about KT. Anyway, gun disabling hits are "weapon" hits. There may be only difference in placement and number of "weapon hit areas" for every tank.

How then would you explain two facts: 1. the animation of weapon hits is ussually positioned at the end of the barrel 2. the weapon hits almost stop happening when the barrel is rotated to the left or to the right. If it was about the mantlet hits, there would be almost no difference. and 3. A weapon hit almost always means gun damage. On hundreds of hits I've seen maybe two weapon hits that didn't cause gun damage. And I'm even not 100% sure those about two. "Weapon mount" hits, which are probably mantlet hits, do not cause gun damage. I never seen gun damage caused by turret or weapon mount hits. I think few hundred shells is statistically significant. I've done tests very similar to those of Vanir Ausf B, on somewhat smaller scale I suppose. But after few hundred hits and experiments with gun positioning, the results seemed quite obvious. I was away for over a week so I didn't read that one. Will check. Do you have maybe a save with this setup, test #5 ? Anyway, I was quite sure about what I saw. I didn't test much with gun looking upwards, but in one test when Tiger 2 was positioned a little upwards with it's gun pointing into the sky, there was NO gun hits in about 50 turret hits. Same when the gun was rotated 20deg to the left (that test included several tanks positioned this way). No gun hits. If I can, I'll try to replicate your test #5 and see. edit: Test 5 Cromwell VII firing at Tiger I Late full hull down on reverse slope (2 meter berm with Tiger on reverse slope, as opposed to further back from the berm on level ground in other tests) range 500m 648 hits * 364 -- 56% -- Weapon Mount (mantlet) 362 no damage 1 partial penetration 1 penetration * 255 -- 39% -- Weapon (gun hit/disabled) * 27 -- 4% -- Front Turret 22 no damage 5 spalls * 2 Right Front Turret (rounded side visible) 2 no damage Comments: Placing the tank angled vertically increases the exposed area of the gun barrel and decreases the exposed area of the rest of the turret relative to the shooter so the direction of migration in hit location makes sense, although 39% of hits disabling the main gun does seem a bit excessive even in this situation (but I admittedly can't prove that ) I would like to see that setup - how high the gun end was positioned. The result is strange - from my experience, and multiple tests with barrels positioned to the side or raised, the side area of the barrel is not simulated at all, the only detectable hitpoint is at it's end (call it muzzle brake). Maybe - somehow in this test - the Tiger's raised muzzle brake was still positioned accidentally in some dense part of the Crommwell gun spread pattern... No ideas really, what caused that, untill I see that setup and test it myself.

Not really. The "weapon mount" hits don't cause gun damage. At least I didn't see that. The "weapon" hits do. And the "weapon" hits SEEMS to happen when a shell hits the "muzzle brake" at the end of the barrel. I mean just the end of the barrel. This is where the hit animation is displayed, and also this is why the highest percentage of gun damaging "weapon" hits is against hull-down tanks which are aiming at the shooter (the muzzle brake is dorectly in the way of incoming shells, in the center of the spread pattern, close to aim point), and there are no "weapon" hits when the barrel is pointing elsewhere, even 10-20deg to the side or up (the muzzle brake is away from the path of incoming shells).

From 1000m it would be much smaller angle and even longer distance possible. So, the "weird tank gun trajectories" are actually correct .

I believe the values from ballistic tables, but we may also try to estimate it this way . I would take 3s as the shell is slowing down considerably at such range. But let's continue with 2.5s. right Not really. The drop lasting 2.5s would be 31m (s=1/2*a*t^2 = 31.25). As you noticed, the shell is falling only half of this time - 1.25s. But height of a 1.25s drop is not half of 2.5s drop - the time is squared in the formula. 1/2*1.25^2*10 = only 7.8m. Your resulting angle would have to be halved, and your resulting length of shell travel would have to be be doubled Half of that - 7.8mm. On average. But the trajectory is close to a parabola so the trajectory angle would be higher at the start and at the end, than the "average" on ascending and descending part of trajectory. So it's actually bit closer to your initial doubled results . The angle you would get from "on average" drop of 7.8m/1000m would be arctg (7.8/1000) = 0.45deg. The real angle on the end of trajectory - because of parabolic shape - would be around 3 times larger. That would be half of that, so 68mm. 143mm now. 417mm/68mm = roughly 6 tank lenghts. So 55m. But it would be less as the trajectory is not linear but parabolical, and the descent angle is higher than that. From the tables, descent angle of 85mm shell with initial velocity of 792m/s and retaining 608m/s at 2000m (BR-365) - so close enough to estimate the US 76mm shell. (I don't have any ballistic tables of German or US tank ammo, sadly). From those tables, the angle of descend of this shell when shooting at 2000m is 22mils or 1deg 19min. With this angle, and the clearing of 560-76mm = 484mm, the shell could theoretically travel further (assuming for simplicity that the trajectory is now straight) 1/sin(1,31deg)*484mm = 21m. It's about 12m behind the rear of the Panther. Does it sound right now ?

Why not ? I see that people usually tend to overestimate the descend angles of tank rounds. They are really not that high as you may think. At a range of 1000m the descend angle of typical tank AT round would be in order of 7-14 mils and at 2000m it's 15-33 mils (that's 0.4-0.8deg at 1000m and 0.86-1.9deg at 2000m. Data taken from ballistic tables of Russian D-10 100mm gun (900m/s muzzle velocity - the smaller angle) and ZIS-5 76mm gun (680m/s muzzle velocity - higher angle). So - it's really small angle, it's almost horizontal. With such trajectory, almost horizontal with very gentle descend - a shell that didn't hit the front lower hull can easily fly below the tank and hit the ground tens of meters behind it. With a descenting trajectory of -1deg a shell that flew under a tank at height of 50cm would fly further 57m before it meets the dirt - assuming absolutely flat ground (0.5m * 1/sin(1deg) = 57,3m). With such low angle it's also obvious that any accidental hits of top armor against a tank target would be ineffective on a flat ground (really any hit at 2deg impact angle would glance even from very thin top armor). Penetration of top armor is only possible if firing from well above the target, or the target is on downhill slope.

Now this one is REALLY oversized and much too heavy . And it's seems to maneuver much worse than KT. Even the Carius didn't like them and called them a mistake

Thanks, didn't see that before. Hmm the film looks like it's playing with natural speed, the people are moving naturally. We see the KT running - well of course that it could run 30 or maybe even 40km/h on a flat ground, if it has time to accelerate. We see in maneuvering (moving in a circle) on a soft but flat ground and it's quite maneuverable considering it's mass. But unfortunately we don't hear it's engine so we don't realise that it's working really really hard on high rpm to achieve that. It looks so "easy" only if you don't hear the engine... Now, take a look and hear how it really sounds: Now it doesn't looks so "easy" when it moves... Well, I don't say it was "barely able to move" - it was suprisingly maneuverable for it's size and mass - but it wasn't so quick and easy and fast turning like the MBT-like thing we have in CMBN . And the crews probaly preferred to not strain the (already overburden and made with defects from poor materials) drive train (engine, transmission, suspension) too much, because using it to it's max performance would grealty increase chances for a breakdown. On the other hand, we also see it's much smoother riding tank than those lighter tanks and vehicles - thank's to it's advanced suspension and great mass. On the other hand, the suspension (and the tank as a whole) was so complex and sophisticated (read - expensive, time consuming to build and maintain, not too reliable...) What suprised me was the speed of turning in place. But again, it was flat soft ground. Anyway, it was not a good idea to turn this way in reality, we see that there is a risk of blocking the trakcks/wheels with dirt/mud and/or throwing a track. A turning while moving was much preferred. Don't know if there was any technical reason for Tiger 2 being a better maneuvering tank than a Tiger I, but T2 had the same engine with much greater mass (56tons vs 68 tons) so it had to have worse performance in accelerating, rough or heavy terrain. The ground pressure was also very similar, higher than for the Panther. Unfortunately, we don't see it maneuvering in rough terrain or moving uphill - where it's large mass would be a greater problem.

Didn't ever see a fim from 50s of a running Tiger I/II, do you have any links ? Well, many of the films from 40s and 50s are were shot at a different framerate that is used today, and the are often playing too fast. One would have to take it into consideration and try to estimate if the video is playing at correct fps and everything is natural, or maybe it's played too fast and everything is slightly speeded up. If the engine of the Tiger - a restorated and fully functional (now) engine - is working hard on high rpm to move the tank, then I know the engine is just too weak for the mass of the tank to make possible better performance. The tiger was said to be maneuverable for a heavy tank, considering it's mass. But it was not fast tank, and it's engine had to work hard to push it trough maneuvers, soft ground or uphill. On the other hand, I seen (on a video) a really dynamic demonstration of a Panther tank . Well, same engine, but only 2/3 of the mass. No wonder. http://www.youtube.com/watch?v=26cSk60Aj3Y&feature=related I think it's a good driver and it shows well what a Panther could achieve

Womble - can you accomplish what you descrived now (without that change) ?

I second that. This change can be made relatively easily (using existing commands and interface - the changes in vehicle code should be minor) and it would improve the control over a unit - while not complicating anything. So IMO it's worth doing.