Variable Quality Panther Glacis

[rexford]

Comments raised on other web sites brought up issue of variable quality Panther glacis. Were all Panthers cursed with sub-standard glacis?

The conclusion is that applying a 0.85 modifier to every Panther may be extreme, and random penalties might be closer to the truth.

At Isigny, France during August '44, 1 of 3 Panther glacis performed like good quality, 2 of 3 cracked on 17 pounder APCBC hits.

At Kubinka, Russia during September '44, 2 Panthers had glacis armor which performed well against 88L71 and 100mm, and 1 Panther glacis allowed 122mm penetration in a manner which suggests poor armor performance.

While these tests only involved six Panthers, which represents a small percentage of the entire production, the results suggest that wargames might treat each Panther in a random fashion. A random selection of armor quality could be chosen for each Panther prior to scenario start and the result could be hidden from player knowledge.

The range of quality multipliers could vary from 0.85 to 1.00.

It may be somewhat extreme to penalize every Panther for a problem that occurred on 50% to 67% of vehicles.

We would also note again that the Panther nose (front lower hull) did not crack at Isigny, and resisted penetration like good quality armor in all cases. Applying a 0.85 quality factor to all Panther armor may exaggerate the actual situation. The Panther mantlet held up well in tests against 76mm APCBC, and side armor did not appear to be sub-par against 75mm APCBC.

We suggest that the 0.85 modifier be restricted to the Panther glacis, and that random penalties be applied. This would seem to improve the historical accuracy of the game.

Potapov site discussions do indicate that Tiger II armor may have been deficient on all areas, and U.S. armor prior to 10/43 seems to have been generally deficient. But Panther glacis clearly has some random variations that would be easy to model in a computer wargame.

[Ari Maenpaa]

I agree and support your excellent suggestion whole-heartedly. But maybe it's too late to see this in CM1

Possibly a simpler routine which randomizes the whole tank's armor quality.

Ari

[mensch]

I agree.. but maybe a 0.86 would be better?

[engy]

Way OT...

Ok, call me an idiot (every village needs one), but I must have missed something somewhere...

The Background: It seemed like all of a sudden one day these posts full of extremely technical information started popping up all over the forum. I've searched all of Rexford's posts, and to the best of my ability, I never even saw an opening "Hello, I'm uber-grognard Rexford and I'm armed with data...". The very first post just opened with more physics than my high school curriculum contained.

So, the questions:

1. Who are you, Rexford?

2. Where in the world is all this data coming from, anyway?

3. Often in your posts you use the plural form "we". Who is "we"?

Sorry for being extremely nosey, but this has intrigued me for a couple of weeks now.

If curiosity killed the cat, I just used up about 8 of his lives...

engy

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

"He who makes war without many mistakes has not made war very long."

Napoleon Bonaparte

[Guest Andrew Hedges]

A certain amount of randomness in armor quality is appealing, but I am a little concerned that it might affect the playability of the game if it were implemented too strongly. In a lot of scenarios, and a lot of QBs, there is only one Panther, and balance could be thrown off if in one game the Panther had a glacis quality of 67% and in another game it had a glacis quality of 95%. This would not be a problem if there were several platoons of Panthers in the game, however, as the armor quality should average out.

Possibly there are ways of adding randomness and addressing these problems. Perhaps scenario designers could set the armor value themselves, or turn off randomness. Perhaps there could be some randomness, but not complete randomness (i.e., instead of ranging from 67% to 96% (or whatever the range would be), Panther armor would range from 80% to 90%. Maybe there would be a way in competition to either lock out the randomness or to have the same randomness apply to both sides (i.e., if you play both sides in a scenario, the armor value remains the same for both sides).

If it is supported by the data, I do really like the idea of having different armor quality for different areas of the tank. The Panther's glacis, even at 85% quality, is still pretty darn good. However, the 60mm side armor (or whatever the actual value is) is weak to begin with, but having it at 85% quality makes it even weaker. If that is historical, that's appropriate, but if the evidence shows that it was higher quality armor, this would be very important to correct.

At least for cm2.

[Lacky]

I agree and disagree!

A problem would arise when players in-game check the armor quality figure. I wouldn't mind variable armor qualities if the number remain hidden, so player's wouldn't know for certain if they had a good panther or a lemon.

[PeterNZer]

Rexford is the uber-grog

I love his posts! Don't understand them, but appreciate the effort and bar-raising standard of his research and argument!

But Rexford, you must remember to add, at the end of each of your posts "BTS! please do or fix sumfink!" That's the real QED of all grog arguments.

PeterNZ

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

"Patriotism is the virtue of the viscious" - Oscar Wilde

"Don't F*CK with Johnny Cash!" - Chupacabra

[rexford]

Allied firing tests suggest that only the Panther glacis had deficient resistance, the rest of the tank was okay. The Panther cast mantlet came from different processes than glacis, and since Tiger cast mantlet is generally held to be good stuff Panther mantlet should also be.

Panther cast mantlet performed well in Allied tests and held U.S. 76 to 200 yard penetration range in Normandy, based on Faint Praise reports.

I work with Robert Livingston, and we're putting a booklet together on armor penetration that will present consistent penetration and armor data, everything to same basis.

We have made CM suggestions along the way, maybe they got lost in the verbal forests.

CM figure for resistance deficiency of Panther glacis (85%) is medium flaws, many tanks had no flaws, some had high severity flaws. We roll dice in secret.

[Guest Andrew Hedges]

I think that there are two conclusions you can draw from the tests Rexford referenced in his post. One would be Rexford's conclusion that some Panther's did not have weaker armor. Another conclusion, though, might be that all Panther's had bad armor, but the armor was not uniformly bad. That is, parts of the Panther armor in the test were 85% good, and other parts were 95% good (or whatever), and the test results are explained by precisely where the test shells hit the glacis. (Note that this is different from hitting a weak spot, which is entirely different).

If other data supports -- or at least does not contradict this evidence -- it might be modeled by including an appropriate randomness factor on glacis hits. That is, 50% (or whatever) of shells hitting the glacis strike 85% armor, and 50% strike good armor.

This would have the advantage of not causing balancing problems, as everyone would be using the same Panthers.

[Ari Maenpaa]

Originally posted by rexford:

I work with Robert Livingston, and we're putting a booklet together on armor penetration that will present consistent penetration and armor data, everything to same basis.

I think that months ago John Waters mentioned about this book in the '88mm lacking punch'-thread. At that time, I believe, some of us began to wait for it to puzzle out the conflicting opinions about 88L71's penetration at 30 degs. And dozens of similar dilemmas too

So what is the current state of the booklet?

Ari

[Jeff Heidman]

I would think that MUCH more important to the issue of the randomness of the Panthers glacis is the issue of the rest of the Panthers armor not being poor quality.

If the rexfords are correct (he he) then that is a much more serious issue than the glacis itself. That would increase the survivability of the Panther immensely.

Jeff Heidman

[rexford]

Panther glacis armor should be modeled with some sort of variable quality factor.

Isigny, France

August 1944

1 of 3 Panther glacis is good quality, other 2 crack after 17 pounder APCBC hits defeated at close range.

Kubinka, Russia

Late 1943

Panther glacis attacked by 88L71 and 100mm Russian gun resists penetration like good quality armor. 88L71 penetration limited to 650m, 100mm to 1450m.

122mm easily and reliably penetrates glacis at 2500m, which suggests less than 100% resistance.

Above two tests suggest that half of Panthers have good armor, half poor. Six Panthers out of thousands is not a statistically valid sample to draw wide conclusions. But some Panthers definitely were close to good quality, so a game that penalizes all Panthers is not really being fair.

Potapov site states that Panther armor deteriorated quickly starting late 1944, due to German loss of alloy mines in Russia. Tiger II's captured by Russian has consistently bad armor on all areas, although side is worse. Panthers met in France during 1944 probably would include alot of good armor vehicles.

To repeat another problem with CM, only the glacis was consistently flawed in allied and Russian tests, we have never seen any evidence that the rest of the Panther had problems. The Panther nose and mantlet perform as good quality in every report and test we have ever seen.

Our group invented the curves for flaw effect as a function of T/D, and our work probably is the basis for armor quality in CM. When we apply flaws in our games, only the Panther glacis has flaws on a regular basis, and the quality varies.

Panther cast mantlet armor was not flawed because it was on the Panther, if Panther mantlets came from the same factory as Tiger cast mantlets one would have to explain why one is bad, one is good.

Panther glacis armor shows a high percentage of flaws in our study, the rest of the Panther seems okay and that is why we feel that the 0.85 multiplier should only apply to the mantlet, and then on a random basis.

A computer dice roll for each hit would be fair, although the entire glacis would probably be bad if manufacturing resulted in problems. Not letting anyone know if a tank had bad armor on the glacis would be ideal.

Panther glacis armor is listed as 85mm thick in many Russian, British and American documents, and penetration ranges at Kubinka suggest 85mm Panther glacis. 80mm may have been the design thickness but 85mm may have been the actual in many cases.

Sherman armor prior to 10/43, on the other hand, was consistently poor with a small percentage of decent armor.

[Dittohead]

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

The box is locked, its Robot Fighting time.

[This message has been edited by Dittohead (edited 01-26-2001).]

[PzKpfw 1]

Lorrin the problem with the Soviet Tiger B report is we have no access to the actual data only what was provided in the article, as far as I know their is no access to the 2 reports cited in the article.

So the question remains on how far you can take the armor quality comments as being written in stone.

Their have also been questions raised on various boards past & present concerning the authenticity of the 122mm penetration of the Panther Ausf D glacis @ 2500m as their are operational cases where the Panther glacis was not defeated at 1000 - 1500m by the D-25T also.

All food for thought as well.

Regards, John Waters

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

"We've got the finest tanks in the world. We just love to see the

German Royal Tiger come up on the field".

Lt.Gen. George S. Patton, Jr. February 1945.

[Simon Fox]

Unlike a lot of rexford's posts this thread is full of conjecture and giant leaps of statistical inference.

What is "good quality" and how is it defined with respect to ability to shrug off a single round?

As pointed out by Andrew, could there be regional variation across the plate?

What about poor welding of joints (as remarked upon by allied inspection of some German tanks) and consequent impact of the infamous 'free edge effect'?

Why define .85 as your lower limit? Perhaps that is the median quality?

Why define 1.00 as your upper limit?

Metallurgical test results please

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

"As has been said, we only listen to bootlickers, and Simon is one of the best out there!"

[rexford]

John,

We have actual American test data for 122mm against U.S. armor plate, penetration ranges in Potapov have been compared to our data and they are reasonably consistent in every case.

Jentz book has Soviet test results for a long line of guns and ammo against Tiger II. These results have been compared to our data predictions and good quality armor is suggested.

So Tiger II armor may be like Panther glacis, quality varies from tank to tank. But Panther glacis is only part of tank that exhibits poor quality, entire Tiger II appears to be poor quality alot of time.

When our booklet comes out you can calculate the estimated penetration ranges for Soviet ammo and you'll see that it matches up well with stories and tests we quote.

We were able to estimate penetration for other Soviet guns from U.S. test of 122.

We don't accept every Russian penetration range report or quote as godsend, we compare to our data and other material and see if it is reasonable. In most cases it is good stuff.

[rexford]

122 reliably penetrated Panther glacis in tests, 88L71 and 100mm penetrated to a certain range.

Soviets doing penetration tests were smart, and like British they probably ruled out hits within a given distance of edges and weak points like MG port. Give the tester some credit.

Good quality resists like it should based on slope effects and penetration data, bad quality cracks after defeating hits or is penetrated easier than it should be.

Metallurgical results are impossible to obtain for tested tanks, but we can still use test results to draw conclusions based on how armor reacts. Real world analysis.

In real world data is incomplete and results may be all that is available. If one Panther glacis cracks wide open at 200 yards, and another resists hits without any failure or major damage at same range, quality is inconsistent.

0.85 is comparison of penetration resistance with medium flaws to what would be expected of good quality armor. U.S. did tests on 1940 armor plate, and correlated results with sonar grams of some sort that showed size and extent of flaws like laminations and inclusion.

We analyzed penetration resistance vs. flaw size and extent and have curves that relate penetration resistance as % of good quality armor. Good quality armor had no flaws in tests, was fully ductile and resisted in a consistent way.

For armor wargaming purposes, it is not necessary to define metallurgical qualities of armor plate, it has low, medium or high severity flaws at a given % of cases. We are not going to hold up our booklet looking for data that is impossible to obtain.

Alot of assumptions but it presents the results of alot of mathematical analysis and is the best that is available. CM is based on our work, with some variations.

If someone wants additional details on how we determined our results or put together a conclusion, ask. If you question someting I say ask for supporting data or an explanation.

We have spent close to 30 years on the armor penetration booklet, can explain everything we say in detail or in summary fashion, and are open to new data and changes. Initial statements are usually general and leave out alot of details, if you want more ask.

There are many engineering areas where it is impossible to mathematically predict what is going to happen due to too many factors. Soil mechanics can be like this.

So general results are correlated with a few factors and equations or procedures are proposed that catch the main drift of what is likely to occur. But people using the math realize that what actually occurs may vary.

Panther glacis armor and Tiger II's don't resist hits in a consistent fashion, and we have tried to model this using 1940 American tests on flaws. We had to use something and the results seem realistic, that is all one can look for.

[rexford]

We're currently looking into some data that may significantly reduce PzKpfw IVH frontal armor resistance against allied APCBC projectiles, so we're not looking for German advantages. Just trying to get fair depictions of armor that wasn't bad all the time (Panther glacis), or wasn't bad too often (rest of Panther).

If we wanted to skew the game how about "shatter gap", where 76mm APCBC breaks up against Tiger 100mm near-vertical plate from 300m thru 1000m. CM doesn't have this and should, in our opinion.

Our rules have shatter gap, and U.S. 76 has no teeth against thick armor until HVAP.

In Faint Praise reports, 76mm U.S. couldn't do squat against front of Tiger beyond 50 yards.

We have a long list of things that CM would have to do to match our rules, but don't expect the world to change to match our vision. Wargame design and research is an art and what suits Picasso would irritate Renoir, and vice versa.

Panzer IV web site from Will Phelps presents German penetration data based on U.S. tests that exceeds CM figures by enough to give Germans added advantages. It even exceeds our figures.

Is that pro-German uber-tank enough?

[Sgt. Beavis]

Hi everyone.

I have owned and played CM for about a month, and this particular issue has raised some thought in my mind as well.

Saying that the Panthers or Shermans in the game will always have 85% quality armor in a particular battle is too...linear, and not very realistic in my opinion. Some Panthers will have armor that is 100% effective, and others will have less than 85% effectiveness. The same holds true for the Shermans, or StuGs, JPIVs, all the rest that potentially have less than 100% effective armor.

Someone mentioned that the player of a particular side should not know if a tank's armor is of a certain quality, and I agree. This prevents one from charging out a Panther with 100% first, and leaving the 85% vehicles behind. That is, if BTS decides to change the armor quality among a series of vehicles in a battle. I would strongly encourage them to do this, and apply a random modifier to a base armor quality.

I think that in doing this it would be more likely for a small variation to take place in armor quality. For example, if BTS keeps the 85% standard concerning Panthers, it would be more likely that the effectiveness is 5-10% either way, rather than 15% or more (more on the low end, obviously; one cant have armor that is 110% of what a perfect Panther would be).

Applying this random factor to all other vehicles would result in more realism, even those that historically did not have major problems with armor defects. Obviously, the random factor would affect these "perfect" vehicles to a lesser degree.

As a final point, Rexford mentioned that according to his data, it seems that the Panther glacis was the only problem spot for most vehicles, and the rest of the vehicle was of good durability. I'm sure this situation existed with other vehicles on all sides in WW2. More research should be undertaken... and the problem areas of various vehicles would be subjected to the random armor quality factor.

Anyway, that's my long first post.

[Guest machineman]

Originally posted by Simon Fox:

What about poor welding of joints (as remarked upon by allied inspection of some German tanks) and consequent impact of the infamous 'free edge effect'?

George Forty quotes a 'comprehensive British firing trial against an Ausf G in late 1944' as saying:

"The use of interlocked joints provided major stability even when the main welds were fractured"

"The design of the vehicle is such that its structural stability is considerable, the effective use of interlocking joints being chiefly responsible."

[PzKpfw 1]

Lorrin:

The recently released material concerning the two Tiger B that the Soviets live fire tested on was an abreviated form. Soviet LF testing procedure is still a fuzzy subject. Their are 2 reports that the extracts concerning the Tiger B metulargy & LF results were based on, neither is available as far as I know publicly.

To make a determination that all Tiger B armor was low quality from reading the article IMHO is wrong, & I'm not saying you are wrong or your methods are in question

just pointing out from your examples listed in the post, that the actual data would be a good starting point before a final determination.

Regards, John Waters

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

"We've got the finest tanks in the world. We just love to see the

German Royal Tiger come up on the field".

Lt.Gen. George S. Patton, Jr. February 1945.

[rexford]

I believe my last post pointed out that Soviet penetration ranges in Jentz suggest that the Tiger II that was tested had good quality armor.

So I changed a previous statement in an earliet post to say that Tiger II's may have been like Panther glacis, some good, some bad.

I agree with the recent posts, make a random computer dice roll where quality ranges from good to high severity flaws. We do this.

Most Shermans were bad, maybe 70% to 50% of Panthers if one is willing to use a six Panther firing test sample as the universe. In the absence of anything else, we use it because there is nothing else.

Or one can model bad armor using the bell-shaped curve that CM uses so much. Alot of tanks near 1.00 quality, alot below 0.90 quality, some below 0.85 quality.

Every Panther was not bad, and the rest of the Panther armor does not appear to have been bad often enough to warrant a quality decrease.

Shermans are another matter. Some were good quality, but most were bad. American engineers talking to Russian engineers said they had bad armor, which may suggest that it was bad front, side and rear.

[Jeff Duquette]

Not sure if this will add to the discussion or not. US ARMY test APO230 (July 44) conducted before US ARMY Test APO655 in Isigny (Aug 44). It sounds as if only one Panther was in the test sample.

REPORT OF PROCEEDINGS OF BOARD OF OFFICERS

Organization: Headquarters, First U.S. Army.

Place: APO 230, U.S. Army.

Proceddings of a board of officers which convened at Headquarters, First U.S. Army, pursuant to Special Order No.196, Headquarters, First U.S. Army, 19 July 1944, a copy of which is attached as Exhibit A(1).

The board met pursuant to the foregoing order at Headquarters, First U.S. Army, APO 230, at 1400 on 12 July 1944 and on subsequent dates to conduct the firing tests. The fuinal meeting was held on 30 July 1944.

Present: All Members.

Purpose: To conduct tests to determine the effectiveness of tank and anti-tank weapons in First U.S. Army, against the German Mk V "Panther" and Mk VI "Tiger" tanks.

1a. Firing was conducted on terrain permitting 1500 yards maximum range with zero angle of site. All guns and types of ammunition, suitable for anti-tank purposes, available to First U.S. Army were defeated on targets whose armour plate was slightly burned. Upon determination of critical ranges , all penetrations were proven against the armor plate of a German Mk V "Panther" Tank with armor undamaged and in excellent condition. All firing was conducted normal(2) to the target. No firing was conducted against the German Mk VI "Tiger" Tank as there were none available.

1b. The following normal types of tank and anti-tank weapons and ammunition were tested;

WEAPON AMMUNITION

Launcher, Rocket, AT, 2.36" Rocket, AT, 2.36", M6A1

Launcher, Grenade, M8 Grenade, AT, M9A1

37mm Gun, M6, Mounted on Light Tank, M5A1 APC M51

40mm Gun, M1, AA AP M58

57mm Gun, M1 APC M86

Sabot (3)

75mm Gun, M3, mounted on Medium Tank, M4 APC M61

HEAT M66 (Special)

3-inch Gun, M5, mounted on Motor Carriage, M10 APC M62, w/BDF M66A1

AP M79

90mm Gun, M1A1, AA AP M77

105mm Howitzer, M4, mounted on Medium Tank, M4 HEAT M67

1c. The board assumed that the effect of hollow charge ammunition is not dependent on terminal velocity but the effect does vary with the angle at which the projectile strikes. Hits approaching 90º angle of impact give better penetration.

2. Record of firing with Photographs. See Exhibit B (1).

Findings: The board having carefully considered the evidence before it, finds that:

1) Launcher, Rocket, AT, 2.36"

Rocket, AT, 2.36", M6A1 will penetrate the side of the turret and the side and rear armor plate of the 'Panther' Tank at 100 yards. On the bassis of the assumption in paragraph 1c it follows that as the range increases, thereby reducing the angle of impact(4) against the side of the turret and side armor plate, the possibility of penetration will materially decrease.

2) Launcher, Grenade, M8

Grenade, AT, M9A1, will penetrate the side of the turret and the side and rear plate of the 'Panther' Tank at 60 yards. On the basis of the assumption in paragraph 1c it follows that as the range increases, thereby reducing the angle of impact(4) against the side of the turret and side armor plate, the possibility of penetration will materially decrease.

3) 37mm Gun, M6, Mounted on Light Tank, M5A1

APC, M51 will penetrate the sides and rear of the 'Panther' Tank at 600 yards.

4) 40mm Gun, M1, AA

AP, M58 will penetrate the sides and rear of the 'Panther' Tank at 600 yards.

5) 57mm Gun, M1

a) APC, M86 will penetrate the sides and rear of the 'Panther' Tank at 1500 yards.

Sabot fails to penetrate front glacis slope plate and gun shield at 200 yards. Due to difficulty experienced in obtaining hits no conclusion as to the effectiveness of this ammunition was reached.

6) 75mm Gun, M3, mounted on Medium Tank, M4

a) APC M61 will penetrate the sides and rear of the 'Panther' Tank up to 1500 yards. APC M61 at 200 yards will not penetrate the front armor of the 'Panther' Tank.

HEAT M66 (Special) will not penetrate the front glacis slope plate at 500 yards (see assumption made in paragraph 1c).

7) 3-inch Gun, M5, mounted on Motor Carriage, M10

a) APC M62, w/BDF M66A1 will not penetrate front glacis slope plate at 200 yards. Will penetrate gun mantlet at 200 yards and penetrate sides and rear of the 'Panther' Tank up to 1500 yards.

AP M79 will not penetrate the front slope plate or the mantlet at 200 yards. It holds no advantage over APC M62 ammunition w/BDF M66A1.

8) 90mm Gun, M1A1, AA

AP M77 will penetrate front glacis slope plate up to 600 yards, the gun mantlet up to 1,000 yards and the turret up to 1,500 yards.

9) 105mm Howitzer, M4, mounted on Medium Tank, M4

HEAT M67 will penetrate front glacis slope plate and gun mantlet at 500 yards (see assumption made in paragraph 1c).

In addition to testing the normal types of tank and anti-tank weapons and ammunition, additional types were tested with the following results:

A) 75mm Gun, M3, mounted on Medium Tank, M4

WP M64 - Three rounds were fired at 500 yards for the purpose of obtaining an incendiary or blinding effect. The results were unsatisfactory.

75mm Gun, M3, mounted on Medium Tank, M4

HE M48 w/fz T105 - Three rounds were fired at the front glacis slope plate at 500 yards to determine its armor penetrative characteristics. The rounds failed to penetrate, ricochetting from the plate and bursting in the air.

C) 90mm Gun, M1A1, AA

HE M71, w/fz M48 - One round was fired at 1500 yards as a ranging shot. No perceptible effect was obtained beyond cracking the welds between the glacis and nose plate and the glacis and side plate for a distance of approximately 12 inches. The corner of the glacis slope plate appeared to have a slight depression as a result of this round.

Recommendations: In view of the above findings the board recommends:

1. That steps be taken by the Ordnance Department to develop armor piercing ammunition of materially higher muzzle velocity and penetrative characteristics for the 3-inch, 76mm, and 90mm Guns, accepting, if necessary, a gun tube life as low as 200 rounds.

2. That upon availability of 90mm APC M82 ammunition in this theater, tests be conducted to determine the effectiveness of this ammunition against the 'Panther' Tank.

3. That consideration be given by the Ordnance Department to the development of a liquid filled incendiary shell capable of igniting the target adjacent to the point of impact, for the 75mm Gun and the 105mm and 155mm Howitzers.

The board adjourned at 1630 on 30 July 1944.

Peter C. Hains III, Colonel, Cav (Armd), President.

Charles E. Hart, Colonel, F.A., Member.

Mason D. Salisbury, Captain, F.A. (Armd), Recorder.

[This message has been edited by Jeff Duquette (edited 01-26-2001).]