Jeff Duquette

More excellent work Slap. Thanks for sharing your findings and insights.

For the track throwing thing, high explosives are used by combat engineers as cutting charges for rail road tracks, bridge trusses, etc. So explosives can cut construction grade steel. But fuzing, direction of the blast wave and explosive tamping are all crucial to maximizing target effects or target cutting potential. Hardened nickel steel -- ala armor steel -- is a considerably tougher material to cut with explosives. Normally line shaped charges are used for this purpose. But track links and pins, road wheels, sprockets, etc are perhaps more vulnerable to the kinetic energy of a direct hit, or blast or splinters. Optics and antennas are also vulnerable to splinters. But again target effects of HE-shell fire vs. tanks are more typically suppression. Get the vehicles to button up. I don’t think the ToW game in its present form distinguishes between a buttoned up tank and an unbuttoned tank. Of course M-Kills (mobility-kill – be this track related damage or engine related damage) and F-kills (fire power kill – be this optical damage – gun tube damage etc) are always possibilities as are K-Kills (catastrophic kill). It's just that the statistics don't jive with these effects being particularly prevalent. Again, it is why modern artillery has resorted to munitions such as copperhead or ICM. It bumps artillery capability vs. AFVs from one of suppressive effect to one of target destruction. As with tactical air support, HE shell fire is at best only marginally effective at killing tanks. Does that mean tanks were not killed by TAC air or HE shell fire – no.

Slap: Sorry -- works got me swamped. But yes, let's kick this around a bit. I'll send an email today as a commo check. Jeff

Hi Slaphappy: Yes, I think so. I just had to do this several months back for one of the Steel Beasts designers. He was interested in similar information, but of course much more modern ordnance. Let me look it over and I'll forward along the results. Perhaps we should do this via a few emails? Best Regards Jeff

While this is far removed from the actual topic\subject, I suppose this might be interesting to some folks here. Regarding the Battle of the Bulge being a major killing ground for Allied Tactical Air Attacks, a similar operational study as that I described above for the Mortain counterattack was conducted throughout the vicinity of salient by a joint group of Americans & British in January of 1945. This was conducted under the auspices of No.2 ORS Group. They were of course very interested in proving the efficacy of modern tactical air support and his ability to shape ground combat results. The group examined 101 German Tanks, self-propelled guns, and other "light armored vehicles". Of this total, four panzers, one SP-gun and one Lt AFV were identified as definite air-to-ground related kills. One SP-Gun examined had both a air to ground rocket hit and an AP penetration. Assuming the second SP was killed by the air to ground rocket, that would put the air to ground tank kill percentage at slightly less than 7%. High Explosive Shells killed 3 Panthers, 1 Sp-Gun, and 4 Light AFVs. That's a total of eight HE shell kills; a little less than 8% of the sample attributable to HE shell fire. Abandonment + Demolition accounted for 39-armoured vehicles. The rest were killed by AP. See No.2 ORS 2nd TAF Joint Report No.1 "Air Attack on Enemy Amour in the Ardennes Salient, Operational Research in Northwest Europe." [ June 13, 2007, 08:19 PM: Message edited by: Jeff Duquette ]

So if I understand you properly, Jeff, you would suspect .2 or .3 calibers for most HE rounds or 100mm x .2-.3 = 20mm to 30mm penetration? But in the case of the heavy, higher than normal velocity OF-412 you would say .45 to .65 calibers or: 100mm x .45-.65 = 45mm to 65mm penetration? Or am I off by a decimal place? In any event, what it does make me wonder is how the RPG-40 (non-heat) grenade manages to penetrate 20mm of armor (at least by several sources) without having anywhere near the velocity of an HE cannon round and lower projectile mass as well? I'm trying to "normalize" the HE effects on armor in the game and am having a rough time doing so because of lack of available information sources. </font>

Does the game include Haft-Holladung (spelling?) magnetic antitank mines?

Looks great Oudy! The invisible portions of the Firefly barrel -- is this some sort of texture issue and the background color of the game’s encyclopedia screen? It looks like there is a camouflage scheme on the 17-pdr barrel. Do the same portions of the barrel disappear during actual game play, or is this only occurring in the game’s encyclopedia screen?

I agree. Rather than assigning the player an on map Hummel or two for a scenario, give the player a couple of indirect fire missions. I have no idea why the designers would include this vehicle in a game of this scale when other more appropriate AFVs could have been included in the unit mix. Like an 105mm Sturmhaubitze 42 or 105mm Assault gun Sherman, or 88mm Flak 18/36 gun, etc.

Yes, the small map size and minimum range of say a Hummel is partly an issue. Although I suppose you could in theory have a Hummel parked at the very edge of one of the bigger maps and be able to come up with the right charge and gun elevation solution to plop a shell onto the far half of a map. Like a howitzer firing charge one at max elevation, or a mortar battery. But to add to the map size restrictions, inherent round to round dispersion, danger close etc, indirect fire within a game of this tactical scale should be abstracted – as it already is. The problem with allowing a Hummel or the like to use indirect fire in a game of this scale is that before indirect fire can be employed the location of the gun or battery has to be established via surveying – both X & Y coordinates as well as the guns altitude (elevation). Back in those days there were no GPS systems or satellites and such. Artillery crews had to be psudo-surveyors as well as artillerists. An Aiming circle needs to be setup, aiming stakes need to be placed. The gun crew needs to very accurately establish what its actual location is in X, Y, and Z coordinates. This information needs to be relayed to the fire direction center (FDC). While not always the case, typically a forward observer (FO) needs direct observation on the target. His position also needs to be known by the FDC. The FO needs to guesstimate the position of the target and feed this information to the FDC. The FDC than figures out the location of the target in X, Y, Z coordinates relative to the position of the FO. The FDC than needs to determine an appropriate firing solution for the gun or battery or battalion or whatever level of unit is going to do the indirect fire mission. This will be a function of the guns positions, the FOs position and the Targets position. Than the FDC can than establish range, deflection and altitude for the gun and target line and determine propellent charge. Adjustments are made for a number of variables – propellant temperature, gun barrel ware, air temp, humidity, non-rigidity of trajectory for large contrast in altitude between the gun and target line, etc etc – even curvature of the earth if the range is great enough. Of course there will than follow a series of spotting rounds or adjustment shoots while the FO + FDC “walk” the fire onto the target. The details associated with different army’s could be a tad different than the above. But in general these are some of the reasons what players could not realistically employ indirect fire with on map guns. Players constantly want to move their Hummels or Wespes about. Each time they are moved, the aiming circle would need to be broken out, aiming stakes set, position of the gun reestablished, etc. I think the only intent of having these vehicles in the game is for use in direct shoots only. The games present indirect fire mission format is much more appropriate to this scale of game than allowing players to utilize indirect fire for Hummels or Wespes or whatever. I suppose one might be tempted to add FOs or FOOs to the game. The manner in which indirect fire is currently used in the game seems OK to me. It's not the best system, but it is a reasonable approach and a good compromise given some of the complexities of indirect fires. [ June 10, 2007, 08:21 AM: Message edited by: Jeff Duquette ]

Hi Oudy: What Nev403 said. Just to add a bit the US 12th Army Group began pressing the British to begin sending them 17-pdr Firefly’s a couple months into the campaign in Northwest Europe. According to the 12th Army Operations reports a handful did end up being shipped over to the Americans. As to your new very nice(!) vehicle edit -- The error isn’t on your part. The Challenger, Archer and Firefly should all be using the same penetration values (and if someone were to make a mod with the Achilles -- M10 TD with 17-pdr -- it should be using the same gun and penetration values as Challenger\Archer\Firefly). The problem is that the game is showing one set of penetration values for the Challenger and another for the Archer. Both should be the same. The Comet was equipped with the 77mm, and would have lower penetration values than the gun on the Archer\Challenger\Firefly\Achilles. Knowing what I know thus far about how the game seems to be modeling slope effects, I think the Archer’s penetration values represent a better guesstimate for your Firefly mod. However, the same penetration values for the Archer should also be utilized by the Challenger. ToW values for Archer’s 17-pdr APC @ 0-degress 100m = 172mm 500m = 162mm 1000m = 150mm 1500m = 136mm 2000m = 124mm NOTE: These are not actual 17-pdr penetration values @ 0-degrees, however if you are trying to maintain consistency within the games engine the above would be the values you would use. Best regards Jeff [ June 10, 2007, 08:31 AM: Message edited by: Jeff Duquette ]

Very good point. I have had the same issue with infantry in built-up areas. It's almost as if tanks can get around built areas easier than infantry. Moreover tanks can plow right through fences and walls -- conversely infantry has to circumvent these obstacles.

Nice Oudy! How come your not using the penetration figures listed for the Archer? Jeff

Oudy: Sprengrante muzzle velocity for the 50mmL42 was 450m/s. Weight of the spgr shell is listed as 1.85Kg. Bursting charge is pretty much the same as the L60 so effective fragmentation between the two guns spgr rounds is probably a wash (same). There are a couple errors in the games listed data that I spotted while plugging and chugging. There should be no difference in weight between the pzgr-39 APC projectiles for the 50mmL42 and 50mmL60 guns. They should both be 2.06Kg. My game shows the L60’s pzgr-39 weight as 2.08kg. I couldn’t find any information on rate of fire for either the L42 or L60. The L42 pzgr-39 cartridge is of course lighter and perhaps 5 or 6 inches shorter than the L60 pzgr-39 cartridge. I suppose in theory, this would allow a loader to load a bit quicker. Although loading rate is also a function of the ready rack arrangements, amount of room in the fighting compartment, etc. Firing rate is often more related to the gunner and his ability to lay the gun on its target than the speed of the loader. I think it would be reasonable to leave the rate of fire the same as that of the L60; or it is also probably reasonable to increase the rate of the L42 over that of the L60 by perhaps 1 or 2 rounds a minute. What is the game’s listed ROF for the L60? I also dug around a bit more in my bookshelves and these are the more typical penetration figures one comes across for pzgr-39 fired by both the L42 and L60. 50mmL42 firing pzgr-39 @ 0-degrees 100meters = 73mm of RHA 500meters = 59mm 1000meters = 45mm 1500meters = 34mm 50mmL42 firing pzgr-39 @ 30-degrees 100meters = 53mm of RHA 500meters = 43mm 1000meters = 32mm 1500meters = 24mm 50mmL60 firing pzgr-39 @ 0-degrees 100meters = 95mm of RHA 500meters = 78mm 1000meters = 57mm 1500meters = 35mm 50mmL60 firing pzgr-39 @ 30-degrees 100meters = 67mm of RHA 500meters = 57mm 1000meters = 44mm 1500meters = 34mm Sources can vary by a couple mm’s here and there. As you indicated above the ToW data for the 50mmL60 pzgr-39 are as follows: 50mmL60 firing pzgr-39 @ 0-degrees 100meters = 80mm of RHA 500meters = 68mm 1000meters = 54mm 1500meters = 42mm These are of course much lower than what one normally sees for the 50mmL60. However, it appears to me that ToW 50mmL60 figures are based upon typical perforation values one sees for the L60 @ 30-degrees and subsequently bumped up to the 0-degree values by using a uniform slope effect multiplier of about 1.2 for 30-degree obliquity. For example the figures I reported above for the L60 are from the Dattenblatte for the 50mm KwK-39 L60 firing pzgr-39. These are also consistent with the figures reported by Jentz in “Tank Battles in N. Africa”. Which probably implies Jentz got his numbers from the original Dattenblatte for the 50mm KwK-39 L60. For example Jentz & the Dattenblatte figures for 50mm KwK-39 L60 firing pzgr-39 at 30-degrees multiplied by 1.2 are: 50mm KwK-39 L60 firing pzgr-39 @ 0-degrees. 67mm x 1.2 = 80.4mm 57mm x 1.2 = 68.4mm 44mm x 1.2 = 52.8mm 34mm x 1.2 = 40.8mm These look reasonably close to the ToW values for the L60 -- close enough for government work and computer games. While the merit of using a uniform 1.2 slope effect multiplier for 30-degree is questionable or debatable, it is in fact consistent with at least one simplistic German wartime method for estimating penetration at 0-degree from 30-degree penetration data. To maintain consistency within the game I guess I would tend to run with the games slope multiplier rather than using a more rigorous approach to determining slope effects for the 50mm KwK L42. Moreover, I would simply take the figures for the L42 at 30-degrees and multiply them by 1.2 to obtain the 0-degree values. 50mm KwK L42 firing pzgr-39 @ 0-degrees via ToW slope effects method. 100meters: 53mm x 1.2 = 63.6mm 500meters: 43mm x 1.2 = 51.6mm 1000meters: 32mm x 1.2 = 38.4mm ============ I spotted an unrelated an error in the Tiger-1’s listed muzzle velocity for pzgr-39. The game lists Mv as 733m/s. Jentz lists Mv = 773m/s for pzgr-39. The Schusstafeln for KwK-36 lists Mv = 780m/s for pzgr-39. Whether this make any difference to game play or not would be a function of what the actual games files are using and how (if at all) it is using Mv. Best Regards Jeff [ June 09, 2007, 12:26 PM: Message edited by: Jeff Duquette ]

Hi Oudy: Real slope effects (SE) should vary as a function of t/d -- t/d being simply the ratio of plate thickness (t) to the projectile diameter (d). If we hold projectile diameter constant, slope effect will increase as plate thickness increases. A thicker plate inclined at 30-degrees will have a greater advantage from slope effects than a thinner plate inclined at 30-deg. Sounds pretty obvious in the sense that a thicker plate should resist more cause' it's thicker. But the added advantage of higher t/d oblique impacts is of course above and beyond the simple advantage of a thicker plate vs. a thinner plate. As to the caliber thing, I wouldn't worry too much about this level of detail given the caliber of guns in the game. You are of course technically correct in that there are scale effects involved with rigid projectile penetration of steel projectiles vs. steel armor. Larger caliber shot and shell will penetrate more efficiently at the same t/d than smaller caliber AP-shot or AP-shell. In other words, a 37mm APC shell tested against a t/d = 1 target will require more energy to perforate the target plate than a battleship's 15" APC shell fired against a t/d = 1 target. But as I say, I wouldn’t worry about the level of error this introduces as it is relatively small given the gun calibers in the game. There are larger sources of error that make this level of slop relatively insignificant. As to the range thing and variations you are seeing in your calculated t/d values, if you hold plate thickness constant and projectile diameter constant, the slope effect will remain constant as well. What you are seeing in the variation of SE with range is the plate thickness is varying -- both the 0-degree thickness value and the 30-degree thickness value are changing at each range interval of interest. The t/d is also dropping with increasing range. For example take your values at 100meters and 1000meters range: 100m 80/69 1000m 54/47 t/d for the 100meter range target is 80/50 = 1.6 t/d for the 1000meter range target is 54/50 = 1.08 We are holding obliquity constant at 30-degrees. We should therefore expect the thicker plate (the 100meter range target with a t/d=1.6) to have a greater advantage over the attacking projectile than a thinner target. In other words the slope effect for the thicker target should be higher than that of the thinner target (the thinner target being the 1000meter range target with t/d=1.08). Examining your calculations of SE based upon the 50mmL60 we in fact see that your slope effects do make physical sense. Your numbers were: t/d = 1.6, SE = 1.159 t/d = 1.08, SE = 1.149 1.159 > 1.149 What I would probably use for 30-degree obliquity SE for APC\APCBC would be as follows: SE = 0.0854Ln(t/d) + 1.2695 SE = Slope Effect for 30-degree Obliquity t/d = plate thickness divided by plate diameter Ln: is of course natural log For example: t= 50mm d= 50mm t/d = 1 SE = 1.2695 (or 1.27 is fine) These are average slope effects for APC\APCBC. They are not unique to specific foibles of a given projectile, but they are reasonably good for game purposes and should let you convert 30-degree penetration data to 0-degree data. Best results are bounded by t/d = 0.25 and t/d = 1.7. You can extrapolate a bit as long as you don’t stray too far east or west of the above t/d boundary conditions. Another boundary condition is that the function will not yield reasonable results for APBC or AP or subcaliber penetrators. So yes you should see a variation in SE with range – but not because of the range itself – if you get my meaning. You see the contrast only because plate thickness and t/d will vary with range. This is a function of how penetration data is typically presented in mass consumption reference materials -- ala Jentz, Chamberlain, Hunnicutt, etc. I think it is easier for folks to understand penetration data in this format, or maybe we have just grown accustomed to seeing penetration data in this very watered down format. Real ballistic testing data is invariably presented in terms of limit velocity or limit obliquity and failure criteria. There are many more advantages to discussing plate perforation in terms of limit velocity or limit obliquity, but that is perhaps something more appropriate for another discussion. I think your approach of using the 50mmL60 data to back out the L42 data is very logical and probably the best and most valid approach to the question. Much better than my initial approach of looking at 57mm M86. The problem as I see it may be rooted in how the game designers are dealing with slope effects. But I need to plug and chug through some additional examples to make sure the 57mm M86 example isn't simply some weird outlier. Best regards Jeff P.S. I’ll see if I have something laying about on ROF for the L42.

Thanks a bunch. Got your email. I'll give him try. Best Regards Jeff

Finn: What are you using to look at the games data files? Thnx Jeff

Perhaps this was the tactic you should have used in the first place rather than the extremely creative frontal assault. Out of curiosity, how many casualties do you personally think is a reasonable number when conducting a screaming stand-up and sprint frontal assault against an AT gun blasting away with HE? Zero, one, two, three, four, six, eight? Or are any one of these casualty rates a statistical possibility?</font>

Sorry – the above doesn’t help you with making the new tank. Here are 0-degree Obliquity perforation values for 50mm L42 firing APC: 100meters = 73mm of RHA 500meters = 59mm 1000meters = 45mm 1500meters = 34mm 2000meters = 26mm

Excellent work Slaphappy.

Never mind about my last idea. I started looking at the games slope effects model, and I’m not quite sure what to make of it yet. Below are comparisons of actual terminal effects data for the US Army’s 57mm M1 antitank gun firing M86 APCBC. I am comparing the actual testing data for the projectile with the ToWs penetration data. I have assumed that the games penetration data for the M86 represents penetration at 0-degree obliquity. As you can see from the first graph, the ToW M86 data is a bit skewed from what one normally sees for performance of this projectile. I pressed on with back-calculation of the ToW slopes effects, but the results are very odd. When determining slope effects for a projectile from proving ground data, one will tend to see some statistical slop, but the general trend for full caliber AP-shot and AP-shell will always be increasing slope effect with increasing t/d. They are directly proportional. In the second graph I show actual slope effects as determined from ballistic testing of the M86 projectile (actual test results shown in red). There is a bit of scatter in the data as one would expect, but the trend is one of increasing slope effect with increasing t/d. I have also included Lorrin Bird’s average APCBC slope effect – shown by the blue curves. These blue curves are of course generalized slope effects and represent Lorrin's averaging of a large number of APCBC projectiles. Again, the thing of importance to note here is the increase in slope effects as t/d increases. The ToW slope effects for this projectile are represented in yellow. These were calculated by utilizing the ToW listed data for the 57mm M86. I than utilized the projectiles actual performance at 20-degrees and 30-degrees to back-calculate the games slope effects. For example: ToW Data: 57mm M86 APCBC penetration @ 0-deg @ 100-meters = 94mm of RHA Actual 57mm M86 APCBC penetration data @ 20-deg @ 100-meters = 89mm of RHA Actual 57mm M86 APCBC penetration data @ 30-deg @ 100-meters = 79mm of RHA t/d = 94/57 ~ 1.65 Slope Effect @ 20-deg = 94/89 ~1.056 Slope Effect @ 30-deg = 94/79 ~1.190 Slope Effect Assuming Simple Cosine Relationship would be a constant at all t/d ratios and would be: @ 20-deg ~1.064 @ 30-deg ~1.155 The trend in the ToW slope effects for this projectile is of course completely opposite of what one should expect. Moreover, the yellow ToW slope effects curves represent a decreasing trend in slope effect as t/d increases. This is obviously rather contrary to how this type of projectile should be expected to perform. Best Regards Jeff Duquette [ June 07, 2007, 07:38 AM: Message edited by: Jeff Duquette ]

Oudy: Excellent idea. The game does need the earlier PzIII models. Regarding your slope effects question, it is complicated as slope effects are a function of obliquity, t/d ratio and projectile type. German wartime estimations also included a slope effects factor related to impact velocity. I have not looked at this two carefully in game terms, but as I understand this from earlier disscussions on the subject, the Anglo-Allied penetration figures represent 0-degree values. What you might initially try looking at is obtaining 30-degree penetration values for Allied guns firing the same form of projectile as your 50mmL42 KwK. I think the early 50mm KwK would be firing mostly APC, so look for Allied guns in this caliber range and same projectile type -- 2pdr, 6pdr, 75mm are possible candidates. Divide the 0-degree values provided in the game by the 30-degree values you have obtained from a little bookshelf research. This will give you an idea of what sort of slope effects multiplier the game designers may be employing. Moreover, lets say the game data for a specific gun has a penetration value of 50mm at 100-yards at 0-degrees. Pop open one of your tank books (or do a google) to see what the author gives for 30-deg penetration. Most references indicate penetration in terms of 30-degree obliquity. Lets say the reference indicates the gun does 40mm at 100-yards at 30-degrees. Your slope effect is than the game 0-deg penetration value divded by the 30-deg value....50mm/40mm = 1.25. In other the game designers are implying that the projectile type does 1.25 times more penetration at 0-degrees than it does at 30-degrees. [ June 06, 2007, 01:49 PM: Message edited by: Jeff Duquette ]

Of course penetration should be greater at 0-degrees than 30-degrees -- not less as you have implied with the above. I think what you are trying to say is given penetration at 30-degree obliquity what is penetration at 0-degrees. Assuming a simple cosine function for slope effect the answer is: Penetration @ 30-deg divided by COS(30) = Penetration @ 0-degrees. But this particular type of AP projectile is not well suited to prediction of penetration via a simple cosine function -- that is it's not well suited for any sort of plate inclination beyond about 8 or 10-degrees from the normal. Moreover, the actual performance vs. a cosine function diverges rapidly beyond about 8 or 10-degrees.