Allied tank armor, quality control

Discussions on the vehicles used by the Axis forces. Hosted by Christian Ankerstjerne
Post Reply
Avalancheon
Member
Posts: 373
Joined: 23 Apr 2017, 07:01
Location: Canada

Allied tank armor, quality control

#1

Post by Avalancheon » 02 Jan 2019, 22:40

On the internet, you will undoubtedly have heard lots of opinions about the quality of German tank armor: Some of it well founded, some of it not. On this forum, much ill-informed speculation from an blogger named TankArchives has been refuted. The short story is, German armor plate was actually of rather high quality, but their mills had persistent issues with quality control from 1944 onward. That is when they switched to a different method of heat treatment for their armored plates.

Do we know of any similar instances of manufacturing defects that afflicted British, Soviets, or American tanks? Isn't it true that the cast armor of the Sherman tanks sometimes suffered from shrinkage cracks?

Richard Anderson
Member
Posts: 6350
Joined: 01 Jan 2016, 22:21
Location: Bremerton, Washington

Re: Allied tank armor, quality control

#2

Post by Richard Anderson » 03 Jan 2019, 05:01

Avalancheon wrote:
02 Jan 2019, 22:40
On the internet, you will undoubtedly have heard lots of opinions about the quality of German tank armor: Some of it well founded, some of it not. On this forum, much ill-informed speculation from an blogger named TankArchives has been refuted. The short story is, German armor plate was actually of rather high quality, but their mills had persistent issues with quality control from 1944 onward. That is when they switched to a different method of heat treatment for their armored plates.
Indeed, it was likely problematic for all...suddenly ramping up production of a highly technical item like that tends to have problems. For the U.S., it was an increase from producing five tanks 1 January-31 August 1939, to 95 from 1 September 1939-30 June 1940, to 280 1 July 1940-31 December 1940, to 4,022 in 1941, to 23,895 in 1942, to 29,505 in the peak of 1943. It would be surprising if such expansion did not result in problems.
Do we know of any similar instances of manufacturing defects that afflicted British, Soviets, or American tanks? Isn't it true that the cast armor of the Sherman tanks sometimes suffered from shrinkage cracks?
"When steel manufacturers finally produced the turret casting for the Medium Tank M4-series and the hull casting for the Medium Tank M4A1, testing by the Ballistics Research Laboratory discovered problems with the quality of the armor, which suffered from pitting, soft spots and variable hardness. One turret, produced by Union Steel Casting Company on 21 May 1942, was subjected to cold weather tests 1-25 February 1943 and was found to be of “relatively poor quality, containing large amounts of fine macroscopic and microscopic shrinkage distributed throughout”, albeit the ballistic failure of the armor in the test was attributed to the sub-zero temperature of the test rather than the poor quality of the steel.

The welded-hull M4, M4A2, and M4A3 types also suffered problems with the quality standard of the welds and rolled armor plate produced by American steel manufacturers. As late as February 1945, a Watertown Arsenal Laboratory analysis of 96 two-inch-thick repair welded cast armor ballistic test plates from two different foundries found significant variances in hardness levels and decreased ballistics resistance in the welds. Then in June 1945, a Ballistics Research Laboratory analysis of four sample plates produced by the Great Lakes Steel Corporation and heat treated by the Standard Steel Spring Company found numerous defects. Back spalling from impacts in the ballistic tests was extensive and analysts found the culprit was extremely poor steel soundness.

The root cause for these problems was discussed in a postwar summary of the testing activities at Watertown Arsenal before and during the war. Prewar, testing was limited to rolled homogenous and face-hardened armor plate a maximum (except in one test) of 1” thick and mainly 1/2” to 3/8” thick. Metallurgical tests were limited to chemical analysis, hardness surveys (chiefly Brinnell), tensile strength tests, and macroscopic and microscopic examination, but excluded radiographic analysis. Ballistic testing was limited to resistance to penetration and then only with .30 caliber AP ammunition against 1/2” and thinner plates and with .50 caliber AP ammunition against thicker plates. No shock testing was done and all penetration tests were at “normal obliquity”, i.e. with the projectile impacting at a 90°angle to the armor surface. In November 1941, radiographic examination of the sub-surface structure of the steel began, then in early 1942 tensile and Jominy hardenability tests and in early 1943 fracture tests for steel soundness, fiber, and V-notch Charpy tests were added to the regimen at Watertown. It was not until the end of 1943 that the ballisticians began to understand the metallurgical characteristics required for optimum ballistic performance from rolled and cast armor plate. However, even as their understanding improved, the problem remained of how to achieve those characteristics, especially given the requirement for increasing armor thickness...

Testing also found the armor ballistic quality varied considerably; especially in cast components and that some design compromises, such as the cast and welded “hoods” on the front hull for the driver and bow driver created so-called “shot traps”, which also markedly reduced the resistance of the armor. The immediate solution again was the addition of welded-on armor applique plates added to the hull front to protect the vulnerable hoods and early-model direct vision slits." (from my draft of For Purpose of Service Test, sources on request, but reproduce without my permission and you are dead to me... :lol: )
Richard C. Anderson Jr.

American Thunder: U.S. Army Tank Design, Development, and Doctrine in World War II
Cracking Hitler's Atlantic Wall
Hitler's Last Gamble
Artillery Hell


User avatar
Don Juan
Member
Posts: 623
Joined: 23 Sep 2013, 11:12

Re: Allied tank armor, quality control

#3

Post by Don Juan » 03 Jan 2019, 15:42

With Allied tank production, there was a general policy from 1942 onwards to reduce the amount of additives in armour (Nickel, Vanadium etc.) due to the enormous expansion of armour output. This is why, for example, the standard British thick RHA plate starts the war as IT 80, and then as the war progresses a number of sub-variants are adopted, starting at IT 80A and ending with IT 80E. All these iterations signify reductions in the additive content, and modifications to the armour production process in order to maintain the quality of the armour without them.

Examples of sub-standard British plate include on the first batches of Cavaliers and Centaurs, this being due to the steelworks involved being new to armour plate production. I think this affected about 300 tanks. There was also a period when some of the 14mm plates of the Humber Mk.IV Armoured Car were found to be cracking in service - this was face hardened plate (CTA), so the obvious suspicion is that the plates had been made to be too brittle. It was also the case that production of the A30 Challenger was temporarily suspended in June 1944 due to sub-standard imported US plate.

I don't think there's any way that you can mass produce armour at the volumes witnessed in WW2 and not have severe quality problems. The only question really is whether the resultant sub-standard vehicles make it to the front or not.
"The demonstration, as a demonstration, was a failure. The sunshield would not fit the tank. Altogether it was rather typically Middle Easty."
- 7th Armoured Brigade War Diary, 30th August 1941

critical mass
Member
Posts: 740
Joined: 13 Jun 2017, 15:53
Location: central Europe

Re: Allied tank armor, quality control

#4

Post by critical mass » 03 Jan 2019, 19:44

Don J,

do You happen to have the exact I.T.80 specifications? Would You mind to share them with us?
I was reading through SUPP 22-58 and realized that the main critique of the british interrogators on german armor was the lower impact strength in some plates, which they associated to mistakes in temper heat treatment leading to unnecessarily reduced physical properties of german armor plate** and to defective armor procurement processes which allowed such "inferior" plate to be passed on to service.

However, when You study british MQ and FH armor shipped to the USAPG for comparative purposes, exactly the same critique is raised by the US testers on british MQ and FH armor. Particularely poor impact strength, insufficient through section hardening and temper britlleness was recorded in the the samples they received (the report is from early 1945 and deals with plate 70-120mm thick). I am wondering whether or not we witness here sort of a "testers bias", that is, that plates are usually "tuned" by their manufacturers towards best meeting a specific set of armor acceptance testing conditions. Cross country tests are thus likely to expose differences partially due to the different conditions not matching those of original intent.

"4. All of the plates possess satisfactory hardenability except the Firth Brown, Ltd., 120mm plate, which is greartly deficient- in hardenability.

5.All of the plates are susceptible to temper brittleness.

7. Inferior impact- toughness, In varying degrees, was revealed by the fracture end notched-bar impact tests of several of the plates. These plates
and the causes determined for their Inferior properties -are as follows:
E.S.C. & Co 70mm-, and 80mm plates: temper brittlenes; Firth Brown,Ltd., 120mm plate - inadaequate quench hardening; Beardmore, 70mm-plate - excessively large grain size -plus, perhaps, incomplete quench hardening.

8. The Firth Brown, Ltd. 120mm. plate and the ES.C. 70mm, plate are characterized by such poor impact toughness throughout most of the section
that these plates would be expected to exhibit poor resistance to severe ballistic shock. From the limited data available it appears that resistance to
penetation by matching and overmatching projectiles, especially at obliquities, might be Inferior to that of rolled plate similar in thickness and hardness but possessed of fibrous fracture."
The then practicesed temperature related charpy impact tests in combination with micro-photography of the surface of breakage is a good indicator for temper brittleness. Two of the three RHA plates tested (the 70mm and 120mm samples) showed predomintly crystalline structure the breakage and a lack of fibrous structure through most of the section. The other, 80mm thick RHA plate had a fibrous-crystalline mixture with roughly half of the section showing brittle crystalline and the other half showing fibre. All RHA plates tested showed evidence for temper brittleness to varying degrees.

These plates were relatively alloy rich (all of them contain relatively large amounts of Ni, Cr and Mo):
70mm plate: C:.31%; Mn: .60%; Si:.22%; Ni:3.27%; Cr:.71%; Mo:.52% -crystalline fracture 80%
80mm plate: C:.33%; Mn: .61%; Si:.18%; Ni:2.82%; Cr:.76%; Mo:.56% -crystalline fracture 50%
120mm plate: C:.33%; Mn: .63%; Si:.23%; Ni:.71%; Cr:1.24%; Mo:.46% -crystalline fracture 90%***

Richard A,

I am looking forward to read the whole article with interest.

hope it helps,


**not necessarely ballistic properties, too, as the example with 50mm Brummbär armor plates makes clear, where the re-heat treated plates were indeed offering better physical properties -as expected- but inferior ballistic results compared to the original plates -to the surprise of everyone, I´d guess

***the 120mm plate seems to have -for it´s section thickness- a too high %Mn (embrittling) and too low Ni (toughening agent). The larger %Cr works in combination with the Ni as hardening agent but the %Mo is not sufficient to change the onset of temper brittleness. It´s more difficult to make good quality (predominantly fibrous) thick plates due to the problem of cooling down the interior of the plate quickly enough to avoid the steel matrix to be exposed long enough for temper brittlenss to manifest.

User avatar
Don Juan
Member
Posts: 623
Joined: 23 Sep 2013, 11:12

Re: Allied tank armor, quality control

#5

Post by Don Juan » 04 Jan 2019, 22:57

critical mass wrote:
03 Jan 2019, 19:44
Don J,

do You happen to have the exact I.T.80 specifications? Would You mind to share them with us?
I was reading through SUPP 22-58 and realized that the main critique of the british interrogators on german armor was the lower impact strength in some plates, which they associated to mistakes in temper heat treatment leading to unnecessarily reduced physical properties of german armor plate** and to defective armor procurement processes which allowed such "inferior" plate to be passed on to service.

However, when You study british MQ and FH armor shipped to the USAPG for comparative purposes, exactly the same critique is raised by the US testers on british MQ and FH armor. Particularely poor impact strength, insufficient through section hardening and temper britlleness was recorded in the the samples they received (the report is from early 1945 and deals with plate 70-120mm thick). I am wondering whether or not we witness here sort of a "testers bias", that is, that plates are usually "tuned" by their manufacturers towards best meeting a specific set of armor acceptance testing conditions. Cross country tests are thus likely to expose differences partially due to the different conditions not matching those of original intent.
I haven't come across any exact specifications for any type of British armour, I'm afraid. As with shell velocity/distance curves, they seem to be oddly elusive.
"The demonstration, as a demonstration, was a failure. The sunshield would not fit the tank. Altogether it was rather typically Middle Easty."
- 7th Armoured Brigade War Diary, 30th August 1941

Peasant
Member
Posts: 798
Joined: 16 Oct 2018, 18:21
Location: Ukraine

Re: Allied tank armor, quality control

#6

Post by Peasant » 05 Jan 2019, 10:59

Don Juan wrote:
04 Jan 2019, 22:57
critical mass wrote:
03 Jan 2019, 19:44
snip
snip
From World War II Ballistics: Armor and Gunnery
Image
Not quite what you were asking for, but it's something.

critical mass
Member
Posts: 740
Joined: 13 Jun 2017, 15:53
Location: central Europe

Re: Allied tank armor, quality control

#7

Post by critical mass » 05 Jan 2019, 17:47

Don Juan wrote:
04 Jan 2019, 22:57

I haven't come across any exact specifications for any type of British armour, I'm afraid. As with shell velocity/distance curves, they seem to be oddly elusive.
I have some data on BHN levels, only. However, I am more interested in alloy composition for various section thicknesses and heat treatment regimes.
Anyway, thanks for looking, You too, Peasant, much appreciated.

Avalancheon
Member
Posts: 373
Joined: 23 Apr 2017, 07:01
Location: Canada

Re: Allied tank armor, quality control

#8

Post by Avalancheon » 10 Jan 2019, 01:26

I'm also aware that after they were invaded in June 1941, the Soviets had issues with the quality of their armored plates. The loss of their magnesium and nickel mines was part of the problem. And of course, so to was the relocation of their steel mills. It reminds me of the problems the Germans encountered from 1944 onward.

Further compounding the T-34 crews woes, their armor had been increased from 45mm to only 75mm, whereas 90mm was required to offer protection against the KWK 40. The low nickel content of Soviet plate armour meant it had a nasty habit of spalling when hit. This would send steel splinters flying around the inside of the tank, causing apalling injuries to the crew.
-T-34: The Red Army's Legendary Medium Tank, by Anthony Tucker-Jones.

"If a shell had gone through the turret of a British tank the commander and the gunner could have stayed alive, because there were virtually no splinters," Bryukhov observed "while in a T-34 the armour would spall a lot and the crew had few chances of survival." This was because the medium hardness armour of Britains Matilda and Valentine tanks had a very high nickel content (3.0-3.5 percent), whereas Soviet 45mm high hardness armour contained only 1.0-1.5 per cent nickel, resulting in much lower ductility.
-T-34 In Action, by Artem Drabkin.

critical mass
Member
Posts: 740
Joined: 13 Jun 2017, 15:53
Location: central Europe

Re: Allied tank armor, quality control

#9

Post by critical mass » 12 Jan 2019, 15:39

I am not convinced that the 8-S plates of the T34 lack in Nickel for the section thickness specified (45mm).
While Nickel indeed is an important toughening agent, one needs to keep in mind that it can also act as an embrittling agent, depending on the heat treatment and the presence of other, embrittling agents (the presence of nickel increases, for example, the tendency of occurrances of severe temper brittleness when in combination with chromium).

Alloys only give potential to improved mechanical properties of the steel (hardenability and toughneability) but it´s always the heat treatment, which performs the actual transformation into a discrete mechanical property and exploits -or messes up- the potential provided by any given alloy mix of the steel.

In case of the 8S plates, it´s not the Nickel content but rather the conscious choice of a too high hardness -through all the section- which turns the plate into a product which is less capable of dealing with shock over overmatched attack. Adding even 8% nickel wouldn´t have changed that.
In this context, the soviets made a choice, and it even allowed to forego any complex subsequent tempering treatements (which caused these plates to have higher toughness in this hardness region than US or british, tempered plates). However, it also had disadvantages under conditions where the projectile could not be broken up, or when this material, conceived for 40-50mm rolled, HHA plates was used for unintended application or different section thicknesses (f.e. cast high hardness armor for the turrets in thicknesses of 75-90mm).

Avalancheon
Member
Posts: 373
Joined: 23 Apr 2017, 07:01
Location: Canada

Re: Allied tank armor, quality control

#10

Post by Avalancheon » 04 Feb 2019, 14:46

I was browsing online this morning when I came across the WAL report on the T-34 and KV-1 tank. I had heard lots of snippets from this report before, but had never read the document in full. It gives a good overview about the nature of Soviet armor, and some problems with metallurgical quality. Most of this information will be familiar to you guys, but I still think its worthwhile to post some quotes from it.


Experimental Report No. WAL. 640/91.

2. With the exception of one component, namely, the bow casting from the Medium Tank T-34 which is primarily a structural element, the armor components were heat treated by quenching, probably in oil, followed by tempering. High temperature transformation products resulting from incomplete quench hardening were detected in some of the heat-treated armor sections.
3. The armor components of the Medium Tank T-34 were heat treated to very high hardness levels (429 - 495 Brinell) probably in an attempt to obtain maximum resistance to penetration even at the expense of structural stability under ballistic attack. The components of the Heavy Tank KV-1 were heat treated to hardnesses more nearly approaching American practice (285 - 321 Brinell).
4. The steel quality of the rolled armor sections varies free poor to excellent. Wide variations in production technique are indicated. Some rolled armor components were cross-rolled while others appear to have been straight-away rolled. The turret casting from the medium tank T-34 is of good quality, while that of the Heavy Tank KV-1 contains excessive amounts of hot tears in the joint examined. The bow casting is extremely unsound, containing excessive shrinkage.

The Ni-Cr-Mo 3-5/8 inch thick turret sidewall casting and the 1-1/4 inch thick Cr-Mo rolled steel plate from the Heavy Tank KV-1 are both somewhat similar in composition to currently produced domestic heavy armor. The Cr-Mo steel has a considerable excess of hardenability when applied to 1-1/4 inch thick plates.
The quality of the rolled steel armor components covers the entire range from poor to excellent, indicating wide variation in production technique. Several of the plates were incompletely quench hardened although possessing hardenability adequate to quench harden through the section thickness.

Peasant
Member
Posts: 798
Joined: 16 Oct 2018, 18:21
Location: Ukraine

Re: Allied tank armor, quality control

#11

Post by Peasant » 29 Mar 2019, 20:31

Have anyone posted this table yet?

Image

critical mass
Member
Posts: 740
Joined: 13 Jun 2017, 15:53
Location: central Europe

Re: Allied tank armor, quality control

#12

Post by critical mass » 30 Mar 2019, 14:22

Avalancheon wrote:
04 Feb 2019, 14:46
I was browsing online this morning when I came across the WAL report on the T-34 and KV-1 tank. I had heard lots of snippets from this report before, but had never read the document in full. It gives a good overview about the nature of Soviet armor, and some problems with metallurgical quality. Most of this information will be familiar to you guys, but I still think its worthwhile to post some quotes from it.


Experimental Report No. WAL. 640/91.

2. With the exception of one component, namely, the bow casting from the Medium Tank T-34 which is primarily a structural element, the armor components were heat treated by quenching, probably in oil, followed by tempering. High temperature transformation products resulting from incomplete quench hardening were detected in some of the heat-treated armor sections.
3. The armor components of the Medium Tank T-34 were heat treated to very high hardness levels (429 - 495 Brinell) probably in an attempt to obtain maximum resistance to penetration even at the expense of structural stability under ballistic attack. The components of the Heavy Tank KV-1 were heat treated to hardnesses more nearly approaching American practice (285 - 321 Brinell).
4. The steel quality of the rolled armor sections varies free poor to excellent. Wide variations in production technique are indicated. Some rolled armor components were cross-rolled while others appear to have been straight-away rolled. The turret casting from the medium tank T-34 is of good quality, while that of the Heavy Tank KV-1 contains excessive amounts of hot tears in the joint examined. The bow casting is extremely unsound, containing excessive shrinkage.

The Ni-Cr-Mo 3-5/8 inch thick turret sidewall casting and the 1-1/4 inch thick Cr-Mo rolled steel plate from the Heavy Tank KV-1 are both somewhat similar in composition to currently produced domestic heavy armor. The Cr-Mo steel has a considerable excess of hardenability when applied to 1-1/4 inch thick plates.
The quality of the rolled steel armor components covers the entire range from poor to excellent, indicating wide variation in production technique. Several of the plates were incompletely quench hardened although possessing hardenability adequate to quench harden through the section thickness.
One has to be very, very careful not to jump to conclusions from this report. The report was filed 1943, and I believe, compiled even earlier.
I´d agree that production quality (joining and welding of plates, soundness of cross rolling) was objectionable. But the steel quality itselfe was fairly good. 8S-grade steel was intended for rolled plate but the soviets incorrectly also used it for cast turrets of varying thicknesses, some of which exceeded to optimum thickness range, which made cast elements from 8S a general weakspot.

The US in particular was hampered in this 1943 period by still insufficient knowledge about what actually made a good quality plate. Very few people are aware that all conclusions in this report are biased due to the then present level of US understanding about tank armor metallurgy.

The soviet T34 RHA armor plate had 40 and 45mm section thickness. The largest section thickness specified for US tank armor in 1941/42 was still 1.5" (38mm). It´s necessray to understand that the US Army still lacked a sufficiently robust larger calibre A.P. projectile, which could perforate without change of shape. US production armor of larger section thicknesses were tested with 75mm target slugs primarely for shock effects but not ballistic resistence against intact penetration. Only later, when the 3" M79 bullet was added to the test procedure, this problem was alleviated for 1.5" to 3" thickness range but in thicker sections of armor plate (4" to 6"), metallurgic tests replaced ballistic tests until the early 1950´s when they finally managed to integrate the ww2, german A.P. experiences. Beforehand, all aviable, domestic AP broke up.

Back in 1942, when they recognized that the T34 armor was 450BHN, they didn´t test it ballistically because the US at that time didn´t had an undeformable AP to actually produce the required, severe test. It´s even worse, that all Aberdeen metallurgical analysis done before june 1944 were potentially flawed because they presumed, that the hardness needs to be low in order to guarantee optimum ballistic reistence. That was because in their 1941 series of test plates, 0.75" thick and 330BHN hard, records of inferior ballistic resistence and clearly crystalline fracture was produced. The USAPG at that time and due to the thin section thickness of the test plates, was convinced, that this inferiority was not caused by poor heat treatment and ascribed it -incorrectly- as a natural attribute of the high hardness. This premise, written in the summary of experimental work 1941, guided all further USAPG analysis well into 1944. As a consequence, it´s explictely stated in the WAL640/91 report that such high hardness armor is expected to be brittle. Only in mid 1944, research exposed that the ballistic inferiority was indeed caused by temper brittleness and lower temperature, Izod and Charpy V-notch impact strength tests in combination with metallographic analysis was added to the tests in order to expose more thoroughly the presence of temper brittleness.

Remember, the 8S-grade, 45mm plate samples were tested only metallurgically in 1942 (chamical analysis, metallographic analysis, hardness survey and through section hardenability). At that time, the metallurgists believed that all plates > 240-260BHN will become brittle and thus, were expected to have a lower ballistic resistence. Yet, surprisingly, they also managed to recognize, and even took the troubles to point out -to their surprise- that the homo-hard 45mm plates "compared favourably in % elongation and reduction of area than domestic armor in this hardness". In other words, rather than beeing expectedly brittle, it was found to be ductile. In fact, the plates were treated hard enough for the given alloy combination, that the embrttling temperature range could be bypassed and they could forego with any complex secondary tempering heat treatment. In my opinion, that´s a very good metallurgy, indeed, albeit in a very narrow range of section thicknesses and application (excellent against small calibre impact and inferior shot but against a matching or overmatching, discriminate, high quality A.P. attack virtually worthless).

Peasant
Member
Posts: 798
Joined: 16 Oct 2018, 18:21
Location: Ukraine

Re: Allied tank armor, quality control

#13

Post by Peasant » 30 Mar 2019, 19:39

On the topic, I have found some intresting data today:

Image

Image

Image

The most striking part is the sudden change in resistance between 2in./45° transmission cover and the 2in./56° glacis plate. They may've shot themselves in the foot when they switched to the 47° glacis in the later versions.

Yoozername
Member
Posts: 2615
Joined: 25 Apr 2006, 16:58
Location: Colorado

Re: Allied tank armor, quality control

#14

Post by Yoozername » 01 Apr 2019, 17:27

Nice. Can you post a DL-able version?

It does say estimate, and velocity is ~730 M/s.

Peasant
Member
Posts: 798
Joined: 16 Oct 2018, 18:21
Location: Ukraine

Re: Allied tank armor, quality control

#15

Post by Peasant » 02 Apr 2019, 11:13

Yoozername wrote:
01 Apr 2019, 17:27
Nice. Can you post a DL-able version?

It does say estimate, and velocity is ~730 M/s.
What do you mean? Can't you right click on the images and save them as usual?
Or if you mean, no I dont have the PDF file where these came from.

Yes, this is an "estimate" but obviously based on actual tests with real german projectiles.

Post Reply

Return to “The Ron Klages Panzer & other vehicles Section”