Panther Glacis Resistance

[rexford]

The following provides a full description of the research that went into the finding which suggests that a good number of Panther glacis contained flaws and had reduced penetration resistance. References to the book, WW II BALLISTICS: Armor and Gunnery, by L. Bird and R. Livingston occur throughout the post when page numbers are stated:

During our research about 10 years ago, we obtained a copy of the Isigny test report for August 1944 from the U.S. National Archives, where two of three Panther glacis cracked after a few hits from 76mm and 17 pounder hits that failed to penetrate (short range hits). This suggests brittle armor with flaws.

A report was found on the Russian Battlefield Internet site where three Panthers were fired upon during September 1944. The armor performed like high quality plate against 100mm and 88L71 ammunition, but appears to have lost significant resistance against the 122mm hit.

We have U.S. tests of Russian 122mm APBC against homogeneous armor, and the Kubinka result occurs at a greater range than would be expected against good quality ductile armor. This suggests armor flaws.

Through the years we have come across many pictures and reports regarding brittle Panther glacis armor, big cracks after AP, APCBC, APBC and HE hits, extensive damage not consistent with good ductile armor, etc. Valera Potapov has published descriptions of 122mm hits on Panther glacis complete with penetration ranges and unusual results that suggest brittle behavior.

Through the NTIS we obtained report AD-A954 952, Metallurgical Examination of a 3-1/4" Thick Armor Plate from a German PzKw V (Panther) Tank, January 1945. The annotation in our book on this report is that it concerns "a sample from captured Panther glacis, likely captured in France. Flaws noted". This annotated reference is listed in section c., of the bibliography, which is entitled German, Czech Armor. Page 129.

We have also discussed the issue with others who have examined German firing tests and German analysis of their armor during the war. At Kummersdorf during August 1942, German firing tests against 80mm through 110mm plate suggest that the armor had decreased resistance compared to expectations. German authorities suggested that the plate be retested for acceptance. A researcher we discussed the issue with speculated that a good percentage of Panthers had deficient glacis armor.

Within the next month we should be receiving detailed German firing test data against Panther armor over a several year period that includes all three Panther models. This data will be analyzed and the results posted on the Internet and in possible errata to our book.

As noted by Nathan Okun, reports usually fail to include all the information that is really needed to consider the factors that might impact the results, if one is aiming for 100% assurance in conclusions. We view penetration and armor resistance research in the following manner:

On many issues 100% assurance is not possible due to the variations in data and information available to each individual researcher, so one offers a conclusion that may not be accepted by all. The Isigny and Kubinka tests run contrary to the theory that German armor was usually ductile and good quality throughout the war, and the U.S. metallurgical report presents further material to throw into the stew. But there is certainly material to suggest that flaws were not as prevalent as we suggest, which is recognized.

But at this point neither side can really be proven with 100% certainty, so both "flaw" theories are neither correct or incorrect, neither right or wrong, since there is no "absolute truth" on the issue (which occurs with many WW II issues).

Notes:

1. The probability of exactly duplicating the Isigny results (2-of-3 three glacis are flawed and brittle) follows, based on assumed flaw rates:

1-in-3 chance when 50% of glacis are assumed to be flawed

1-in-37 chance if 10% flaw rate is assumed

1-in-140 chance if 5% flaw rate is used in analysis

2.

We have U.S. firing tests against flawed and unflawed armor which suggest reductions in resistance based on the size of the flaw and the projectile diameter compared to plate thickness. Panther glacis with medium flaws should lose about 7.5% of the resistance against 75mm and 76mm hits, which means that the Sherman 75mm APCBC and 17 pounder APCBC should not be able to consistently penetrate a medium flawed glacis at point blank. 85mm APBC might at very close range, but we have not done the calculations.

3.

It should be noted that since we used curves of best fit through data, it should be understood that estimates from our book will be subject to variations (we show the variations in slope effect between different U.S. projectiles in the book as a function of diameter and T/D ratio even though we use one equation for 57mm-90mm APCBC).

4.

We have compared our estimates with firing test data/battlefield reports in many cases (we analyzed all of the firing test data we could find), and the two are reasonably close. We're either close to the actual with our data or we're lucky, but many people say it's better to be lucky.

5.

The Germans tested 80mm armor using 50mm uncapped rounds, which is discussed in Appendix 11 of our book. 50mm hits on an 80mm thick plate may result in flaws not being exhibited through a decrease in penetration resistance, while 75mm and 152mm hits may cause flaws to impact the resistance.

It is also possible that 50mm tests for complete penetration damage type (ductile or brittle) may hit areas with weak or no flaws and miss isolated flaw areas.

[rexford]

If the statistical analysis is based on a given flaw percentage exactly duplicating the three pieces of data we have (2-of-3 flawed at Isigny, 1-of-3 flawed at Kubinka, 1-of-1 flawed in U.S. metallurgy test, chosen from a random sample), the following are the probabilities for duplication of the three data pieces:

5% of Panther glacis are flawed, 1-in-21,000

10% flawed, 1-in-10,000

30% flawed, 1-in-40

50% flawed, 1-in-14

70% flawed, 1-in-17

80% flawed, 1-in-34

While the samples are small, the statistics suggest that the most likely percentage of flawed Panther glacis armor is around 50%.

[Spook]

Impressive first post, Lorrin. Well laid out in its logical progression for attempting to assess the probability of flawed Panther glacis armor plate. Very good citing of references, too; you even invoked Nathan Okun. (The NWS guys who have been reworking Fighting Steel are quite mindful of Okun's analysis techniques & tools.)

But your very last statement in your second post is starting to overreach a bit, and the reason for this is ironically mentioned in that same statement: the small sample population. Considering the timeframe of manufacture for the Panther tank series (1943-45), and the numbers produced (app. 5,000 of all types), the test samples are far too small, IMO, to establish a meaningful statistical breakout on defective glacis population probability.

I recommend that your review & updates consider also the following, along with the impending reference material that you're expecting:

1) The focus seems to be that of testing the end product. Beyond penetration tests, were metal surface hardness tests also attempted on the plates? Perhaps the NTIS report will shed a bit on this?

2) Were the vehicles identified by their specific type (Vd, Va, Vg, etc.) or by vehicle production numbers? This could help determine when the test vehicles were produced.

3) If production timeframes can be established, can it further be established the circumstances to steel production at that same time? Were key alloy ingredients limited in availability? Were primary steelmakers operating in adverse conditions due to allied bombing raids, again at what times if this was so? Or, were the plate-producing steelmakers capable of making proper ductile plate, but were forced to make shortcuts to meet demand, such as improper heat treatment techniques?

Addressing these issues, and considering variations to plate flaw probability based on specific manufacturing times in the 1943-45 period, will help in presenting your case. Good luck.

[JasonC]

From those numbers, I'd say anything between 1/4 and 3/4 would be believable, and your 1/2 is a reasonable conclusion. Lower than 1/4 flawed plates, the chances against the observed test results seem too high to be reasonable. That leaves a wide range of uncertainty (+/- 1/4), but as you note pretending one can tell more exactly about many of these things would be misleading, with such small samples. A useful anaylsis, BTW.

[Spook]

Hmmmm. I had presumed that the "pro-German" or "pro-Panther" CM lobby would've batter honed in on this rather informative topic.

[Jeff Duquette]

<blockquote>quote:</font><hr>Lorrin Said: On many issues 100% assurance is not possible due to the variations in data and information available to each individual researcher, so one offers a conclusion that may not be accepted by all.<hr></blockquote>

It would be interesting to know if any of the Isigny Panthers had suffered hits on their glacis plates prior to being captured and subjected to controlled testing. Metal fatigue induced from non-penetrating hits suffered during actual engagements would presumably skew field test results reported in the infamous Isigny tests.

There is no mention of the condition of the target Panthers within the version of the report I have access to. There is not even an indication as to whether any of the Isigny Panthers had been brew-ups prior to being drug to the Isigny "bone-yard" for testing. Of interest is a sentence on the first page of the Isigny report indicating that the results of the test were urgently required. It is possible that this urgency resulted in some essential background information being inadvertently left out of the final document.

It has been awhile since I read the Kubinka Report. I don't recall if the condition of the Panthers or Tiger II prior to commencing test firing was noted or not. Had the test Panthers burned prior to being captured\recovered by the Red Army?

[rexford]

Thanks for good responses.

The possibility of burned out Panthers at Isigny was brought up on AFV News site, and is a valid concern.

Isigny report indicates that two Panther glacis cracked after a few hits, so pre-test damage from hits would be minimal. I assume that testers checked glacis for cracks prior to test to assure that test results were valid for generally undamaged armor. The assumption that the testers checked for pre-test damage is one of those issues that is not stated and lots of people feel uncomfortable trusting that the test team considered all sorts of things.

Our view is that if reports are thrown out due to incomplete statement of test factors and pre-test precautions one would be left with little or nothing. This is a point Nathan Okun has brought out several times.

During other U.S. firing tests, burned out tanks were used to set penetration ranges against captured Panther tanks, and then undamaged tanks were used for final tests. I believe that the final test results were consistent with the earlier tests against burned tanks.

Will look further into points raised.

[rexford]

Following info was posted on AFV News and adds some points regarding Panther glacis resistance and ductility on Eastern Front:

From Valera Potapov post several years ago on IS-2 ammunition performance:

"However, in autumn of 1944 the problem of the poor AP performance of its shots disappeared. Suddenly the performance of the D-25T gun of the IS-2 against the German tanks improved dramatically. The reports from the front described cases where the BR-471 122mm projectile fired from 2500m ricocheted off the front armor of a Panther leaving it in huge holes and cracks". This is a round that previously could fully penetrate up to 600-700 meters and now it is doing substantial damages in more than one event (cases).

Sounds like Isigny results, only worse. 17 pounder APCBC results in cracks after ricochets at close range, 122mm does quite a bit more at 2500m (where penetration/armor ratio would be alot less than 17 pounder at 200 yards).

The first encounters between IS-2 and Panthers (late fall, winter, early spring?), and tests during January 1944 (dead of winter), resulted in poor performance of BR-471 ammo. Then, during autumn 1944, things appear to change drastically.

Your point is well taken, weather and temperature could impact conclusions drawn. The above information suggests that the observed decrease in Panther glacis ductility and resistance was not temperature dependent."

As an aside, the first IS-2 tanks appear to have suffered from exceedingly brittle armor which is given as one reason for high loss rates. Tests with 76.2mm ammo resulted in excessive fragmentation and splintering inside the tank. Problems with armor ductility are not limited to Panther glacis, and Sherman cast armor prior to October 1943 was known for brittle, crystalline structure at Brinell hardnesses over 240. Rolled armor on Sherman could also have significant loss in penetration resistance due to flaws and brittle structure.

[Ari Maenpaa]

<blockquote>quote:</font><hr>Originally posted by Spook:

Hmmmm. I had presumed that the "pro-German" or "pro-Panther" CM lobby would've batter honed in on this rather informative topic.<hr></blockquote>

Well, probably I would be considered as a pro-Panther lobbyist when it comes to how CMBO currently models the Panther. The slow turret speed and 85% armor quality seem both to be quite conservative views of the Panther's real potential. At some spot we also discussed the differences in turning radius between different tanks. CMBO doesn’t model the advantage of neutral steering which would apply to the Panther (and on some others too). Not to mention the total lack of superior optics

As many have already noted, Rexford's data is "the data" of the subject. There's not much point or reason to argue against it (If you meant that) without first-grade evidence, although small sample gives freedom to make differing conclusions. Anyway the presented analysis sounds right to me. It will be very interesting to see how the German firing test data alters the current conclusions.

Correct me if I'm wrong, but basically Rexford stated that ONLY about half of the Panthers had a flawed glacis plate (They are all flawed in CMBO). And even an averagely flawed glacis (which is what CMBO tries to model, I presume) should be 5 - 10 % tougher than currently in CMBO. This applies to the Western front where the allied AT-gun calibers didn't exceed 90mm. IIRC the resistance of an armor plate gets dynamically weaker when the AT-shot’s caliber (size) increases. So the glacis resists very differently against 76mm shot than 122mm shot. There was also the point raised earlier that the Germans tended to compensate for decreasing armor quality by making the plates thicker.

Ari

[Spook]

<blockquote>quote:</font><hr>Originally posted by rexford:

Your point is well taken, weather and temperature could impact conclusions drawn. The above information suggests that the observed decrease in Panther glacis ductility and resistance was not temperature dependent."

<hr></blockquote>

If this comment was in response to my earlier "heat treatment" issue, I didn't quite mean that so much as to what the ambient conditions were at the time of the field tests. (Although that too could be noteworthy.)

Rather, what I meant by heat treatment was by specific manufacturing processes done to the plate after rolling it. The heat treatment is done to certain steels as to change its characteristics, either in strength, ductility, hardness/machinability, or perhaps for other factors. By example, a steel component can be heated (to app. 800 deg. C), then cooled by a controlled time-temperature cycle as to form either pearlite, bainite, or martensite, each with unique qualities in hardness and ductility. The trick again is that the time-temperature cycle of cooling has to be properly managed in cooling "set points." How well in late-war could these manufacturing cycles be applied consistently?

Or as an added possibility, the Germans also might had utilized "hard-rolling" of plate as to increase hardness mainly on the plate surface?

Establishing the manufactured nature of the armor plate is thus desireable, and allowing for possible variations during the war's progression.

[rexford]

This thread is intended to draw out all opinions on Panther glacis quality, especially those that differ from mine. So all doors are wide open and previous, as well as future input, is appreciated.

Post on Matrix Games forum indicated that Germans may have switched from high quality, "pure" E armor to other types as demands of war placed more and more constaints on tank armor production. We have some German armor plate reports where the 80mm stuff was without flaws in all cases, this is 1942 production and might apply to Tiger.

Matrix Games thread on Panther glacis is at:

http://www.matrixgames.com/cgi-bin/ultimatebb.cgi?ubb=forum&f=26&SUBMIT=Go

One of the questions that came up on another site is whether brittle glacis armor on Panther would resist U.S. 76mm hits in ductile fashion with scoops and no cracks, but show brittle nature against heavier more penetrating rounds.

At Isigny France, U.S. 76mm bounces off all 3 Panther glacis plates in seemingly ductile fashion, but 17 pounder is deflected and leaves a crack in two glacis plates at close range.

We speculate that the crack would have appeared after the shot bounced away, and may have been due to energy release causing the brittle material to pull apart. Not knowing much about metal, input on this would be appreciated. If the crack appeared as the round dug in, it would seem that the projectile would have stuck in the plate or even got competely through (excuse my English here, but I lived in Brooklyn for awhile).

The Russians noted that 122mm AP hits were deflected and gaping holes and cracks appeared in the Panther glacis in reported cases. They do not state what percentage of 2500m or long range hits resulted in dramatic results (although they probably would only see the tanks that fell apart and were abandoned on the field).

This is being brought up because a few people have stated that many Panther glacis plates show evidence of rounds being deflected in a ductile fashion, nice clean looking scoops, without cracks or ragged edges.

The Matrix Games forum posts also speculate that any flaw tendencies in Panther glacis may have shown up after a certain date during 1944 due to some changes that may have occurred.

The British appear to have had much experience fighting Tigers and Panthers, and they assessed the armor quality as good (with an infrequent bad plate here and there). So it doesn't sound like 1943 and early 1944 Panthers were any different from other panzers. The Russians also note similar findings.

Just trying to broaden the base of the analysis on the Panther glacis issue.

Thanks.

[Brian]

I wonder if this could be related to the problems that the Royal Navy encountered in the deck plates for the carrier HMS Illustrious? Because of a pre-war shortage of armour plate, the builders ordered a batch from Czech sources (obviously before the occupation/dismemberment of that nation) and used them in the carrier's deck plates.

During action in the Mediterrean, in 1941 the carrier was hit by several bombs from Stukas and the deck plates failed quite badly, resulting in the carrier having to return to the UK for extensive repairs. The subsequent investigation, if memory serves me correctly, found that the plate was far more brittle than it should have been. I wonder if the manufacturing process utilised for that plate could well have continued in use into the plate utilised for the Panthers?

That might explain the problems with it cracking when it really shouldn't have. If one was able to discover the registration numbers for the vehicles used in the test and then tracked back to find where the plate utilised in their manufacture had come from, it might be possible to prove it one way or the 'tother.

[Spook]

Exactly my point, Brian. If the source of the Panther's armor plates and the manufacturing processes utilized could be determined, then this would help establish the nature & probability of this armor plate to have "flaws."

Sources like the ASME (Amer. Society of Mechanical Engineers) have extensive sets of "test standards" published earlier that could also be of use in providing non-destructive evaluation (NDE) tests of specific metals, and of impact resistance tests.

[rexford]

Robert Livingston posted the following on the Yahoo! Tankers forum regarding the ability of the Panther glacis to resist hits in a ductile fashion while being vulnerable to internal crack propagation when hit really hard:

----------------------------------------

"Guilford11 raises an interesting point.

Metallurgical analysis in the US during WWII showed a sample of Panther glacis to be flawed in the center. Crystal structure was "bainite" in the center layer, which would be classified as an internal flaw. It is possible that a severe

hit on this kind of flawed plate would start cracking, which would be stopped by the outer layer of tougher steel, until hit another severe

blow which would cause further migration of the crack through the entire thickness of the plate.

This is a common mode of failure of steel railroad rails; one of the methods railroads use for prevention of rail failure is to periodically image the interior structure of the rails from a slow-moving test vehicle. Those interior cracks grow slowly under the repeated stress cycles of supporting heavy wheels.

Obviously the Germans didn't examine the Panther glacis this way after it successfully defeated hits in combat, nor did the Allies do radiographic studies of Panther armor which had been hit but not penetrated. But I do suspect

that interior flaws in armor would result in cracking after a series of hard hits, even though the armor would appear sound and unharmed on the

surface after the first projectile defeats."

-------------------------------------------

The Isigny firing test report notes that 2 of 3 Panther glacis cracked after a few hits while the third glacis withstood over 20 rounds without cracking. 17 pounder hits at close range cracked the glacis, 76mm APCBC and HVAP did not.

[JunoReactor]

Someone please explain to me why larger caliber rounds would crack the plate whereas the smaller ones would not?

Could it be simply because energy is dissapated over less material?

[rexford]

Larger rounds take advantage of brittle characteristics more than lighter ones. 17 pounder APCBC has a harder nose and strikes at a higher velocity than U.S. 76mm APCBC, so the impact is greater.

It's kind of like the difference between throwing small marbles at high velocity at a thick glass wall, and then throwing a brick at a much slower velocity but about the same energy. The glass may deflect the marbles with minimal damage but crack (or shatter) against the brick.

Big things take advantage of problem areas better for some reason.

[rexford]

The following is Robert Livingston's response on the Yahoo! Tankers site to several questions that I raised:

My answers to Rexford's questions are inserted below:

Rexford179@c... wrote:

Thanks for the response.

Would the following be true, based on your review of the analyzed Panther:

1. a low impact hit, such as U.S. 76mm, could be resisted in a ductile fashion since the surface layer was unflawed and the center area would not crack.

YES, but see item 5 below.

2. after a number, maybe one, of sharp impacts a crack could start in the plate interior which would lower the resistance to crack propagation and penetration

YES

3. deflected hits that start interior cracking could still appear to be ductile looking on the surface due to unflawed armor

YES, the more ductile outer layers could deform or scoop in more ductile fashion.

4. hits from large rounds, such as 122mm and 152mm, might be able to start a crack and take advantage of the fissure during the same hit (one hit with lowered penetration resistance, and an ensuing penetration that would not occur against fully ductile armor)

YES

5. It is very good to obtain a metallurgical explanation of how flaws contribute to crack potential and propagation. One important item is that the Panther glacis was not entirely flawed, and contained a good, ductile surface layer.

BOTH front and back of the Panther sample glacis were better quality than the center. BHN was 269 at 0-28mm depth, then 262 at 28-53mm, then 269 at 53-83mm. The slight decrease in hardness in the center was accompanied by a drop in Charpy notch toughness from 14.5 ft-lbs in the

"good" zones to 11.75 ft-lbs in the brittle zones (at -40degF). Fracture texture was observed to be crystalline throughout the thickness.

6. What would be the implications and probability of fully flawed Panther glacis armor, and have you examined any pictures or penetration descriptions that might suggest flaws extending to the surface. Also, could you suggest why the Panther glacis might contain flaws based on heat treatment?

To quote the Panther glacis report: "inferior toughness was attributed to a combination of incomplete transformation to martensite upon quenching and temper embrittlement". The report says that the plate was quenched incompletely, or too slowly, which would leave the center of the plate not sufficiently quenched (transformed to martensite), although the outer sections would be completely quenched to martensite.

It is further suggested that the armor may have been tempered in the temperature zone which leads to embrittlement (400-1000degF). This would degrade the quality of the steel in both "good" and "bad" sections, leading to overall poor quality due to brittleness. Thus, the Panther glacis was of adequate hardness, but inadequate toughness, throughout the cross-section. A zone at the center of the section was even more brittle than the outer zones. Roughly speaking, one could call the entire plate "flawed," with an "especially flawed" layer at the

center.

The US rejected one of its own cast hull fronts after failure in a ballistic test, where 90mm APCBC and AP hits resulted in severe back-spalling. The following Charpy notch toughness was determined: 3.6 to 25.0 ft-lbs at the center, and 12.4 to 29.9 near the surface. At the location of backspalling, the lower values (3.6-12.4) were obtained.

The US example above illustrates the actual ballistic results of brittle spots in armor (backspalling) and the wide variability which could be encountered in a single piece of armor. When both Panther and US armor samples were re-heat treated as an experiment, toughness improved

markedly. This shows that the BHN was adequate, as was the chemical composition. The problem was in the heat treatment procedure.

Source reports for both armor samples described above are contained in the bibliography to Bird and Livingston's _World War II Ballistics: Armor & Gunnery.

[offtaskagain]

So this problem would apply to the KT as well right? Though I doubt the flaws in the center would matter too much since it was twice as thick.

[TSword]

The problem the German "Pro-Lobby" has is that ballistics are plain wrong.

A 76mm APCBC in M4A3 76 which penetrates the lower Panther glacis always with first hit while at a side angle of 30 ° is absurd (I can send you the testsetup if you like..). No shattergap is in,

Tiger Mantlet still is wrong in CM (Also Rexford delivered proof of it).

The Tiger, StuG's, Panther had the "Scherenfernrohr" as standard equipment which vastly improved their longrange spotting cababilty, in CM not there. Btw StuG's were the most feared type tanks by the russians, in CM they are cheap and die very fast.

As rexford states there's no proof of the 15 % decrease in armorquality to date only some evidence that there were poor plates. It's also very likely that some production batches of T-34 had severe flawes in their cast-armor as statements exist of the german side which indicate poor armor quality in 1943 T-34's for instance (I doubt that somebody already has investigated on that topic, of course the Panthers in CM2 will still have 85 % or lower rated armor..)

Btw Rexford states that M4's also had flawed armor plates to what extent ?

CM clearly boosts allied tanks and undermines the german ones.

[Andrew H.]

<blockquote>quote:</font><hr>Originally posted by TSword:

The Tiger, StuG's, Panther had the "Scherenfernrohr" as standard equipment which vastly improved their longrange spotting cababilty, in CM not there. <hr></blockquote>

Do you have any evidence that Tigers and Panthers were issued scissors-type rangefinders as standard equipment?

[TSword]

Anderw Hedges, look at some Turret outfits of Panther, Tiger and StuG and you will see that there were special provisions for them in the turrets/tanks. (Jentz's books are great for that for instance or some in the museums, like the Tiger refurbished in Bovington..)

[Andrew H.]

Yeah, I've read Jentz, and other things, and AFAIK, none of them discussed the Scherenfernrohr being standard equipment. On the other hand, they are pretty specific about things like Turmsehroehre and Turmzielfernroehre: not only their existence and designation, but also other specific technical details.

There are occasional pictures of men in tanks having the occasional scissors-type rangefinder (esp. StuG crews, which makes sense because they were part of the artillery branch), but neither photographic evidence nor AARs suggest that the scissors-type rangefinders were standard equipment.

You may want to take a look at this thread for an earlier discussion that included this issue.