Tirpitz plate results...

[George Elder]

After WWII, some of the armor was salvaged from the Tirpitz, and the British tested it with 14" shells -- with some lackluster results. As I reviewed the study, I noted that some of the test plates had been cut into small sections, and at first I thought that the cutting process itself might have impaired the resistance of the metal (e.g., heat effects). However, it now comes to mind that the reason for some of the results has everything to do with the size of the plate selected. Here is a bit from Paul Lakowski, who does a lot of research in this area:

"When determining the resistance of steel plate several additional factors should be included . These are ‘lateral confinement’ and the ‘T/d effect’. T/d refers to the ratio of the thickness of the armored plate to that of the attacking projectile, while lateral confinement refers to the ratio of the diameter of the attacking projectile to the width of the armored plate. Tests done on armor material will always yield different results if either the T/d or the Lateral confinement ratios are too low. For modern APFSDS the width of the plate must be more than 30 times the diameter of the attacking rod , for all results to be stable and transferable to another case for comparison. Along the main turret walls of a real tank target, this effect is marginal, but near the mantle the effect reduces the armored resistance to 0.85-0.9. Further, test on ceramic steel targets show the effect is much more dramatic . For WW-II AP type shot @ ~ 500-1000m/s the effect of free edge and effect of previously penetrated plate is different and test conditions suggest that the maximum area of effect for the free edge is only ~4 projectile diameters before the plate is considered to be confined. "

The bottom line is that reading results is simply not enough. The methodology used is critical, and thus we need to study the subject in great deapth to make any sense of the literature. Moreover, we results seem odd, there is usually a very good reason for this. More on this later.

George

[Bill Jurens]

The effects of shells striking near the edges of armor plates were well understood at that time, as many many test plates had been fired at and the results carefully observed. (as noted, the effects are really quite small with that type of armor/projectile combination we are talking about. The proving grounds conducting these tests were staffed by experienced professionals. In my opinion, there is, therefore, little likelihood that their test results would have been distorted by a simple blunder such as firing too close to the edge of the plate(s).

Bill Jurens.

[George Elder]

... in many respects. I certainly think that the terminal ballistics people were aware of the edge effects early on, and thus the rules used to semi-standardize plate trials we see at the turn of the century. Yet even these rules were problematic in that the corner quadrant shots often were subject to edge effects unless they were very carefully placed. Moreover, the samples prepared for the tests in question rather violated the typical trial "rules" with regard to size. As for other contitions that were accepted at the time, but are now known to be suspect, the policy of using a single plate for multiple trials is noteworthy -- except in those cases where the plate is indeed very large relative to shell size and the hits are spaced wide enough apart. Even here, the policy of using a single plate for multiple trials is problematic. One can easily understand why this could be the case with cemented/FH armor, and especialy armor that was prone to delamination effects (e.g., Class A plate). Here we have the prospect of post initial-impact crack propogation and delamination effects that could most definitely influence subsequent trials. These are fairly well known effects. And although you may classify edge effects as small -- I note that most of the modern ballistics experts I have communed with do not hold this view -- nor subscribe to the notion that that such effects should be dismissed. Moreover, how does one quantify "small"?

George

[Bill Jurens]

George, you seem to assume that the people doing the tests were a bunch of feeble-minded morons who couldn't (after uncounted thousands of tests) figure out how to adjust for edge and adjacency effects. Proof shooting wasn't just done "for fun", you know, and very often a great deal of money rode upon whether a given plate batch was rejected or accepted. The manufacturers of the plates were always on site, and were very careful to ensure that the location and spacing of various shots was such that any rejections of there plates were due to real quality control issues and not due to inadequate or inappropriate spacing. They may not have been present for the Tirpitz tests, but there is no reason to assume that in this particular case the proving ground(s) suddenly forgot how to measure penetration probabilities properly. They knew what they were doing.

Unfortunately, the penetration of heavy face hardened plates by large caliber projectiles at (what would now be considered to be) moderate velocites, is more or less a 'closed book' theory-wise, with no serious testing having been done after c. 1958. It's also sad but true that more modern testing and theory -- to which you often allude -- can add little or nothing to our knowledge of these phenomena; the geometry, materials, scales and velocities of the newer weapon/target combinations are so dramatically different from those of the old that almost nothing of value regarding WWII vintage armor etc. can be gleaned from them. If anything, the inevitable loss of individuals with actual hands-on testing experience with heavy plates and projectiles means that we actually now know LESS about the subject than we did in the 40s, not more.

Bill Jurens.

[George Elder]

Hi Bill:

As in... "George, you seem to assume that the people doing the tests were a bunch of feeble-minded morons who couldn't (after uncounted thousands of tests) figure out how to adjust for edge and adjacency effects." Now where on earth did you get the idea that I assumed the folks who conducted the tests were innane? If the truth be told, I have great respect for the people who did these tests, but they were often short funded, had limited material, and were pressed for rime, as we see in many experiements that could have been done in much more detail (e.g., the British spaced array armor tests, Japanese plate tests. etc.).
Furthermore, what proof do you have that the experimenters in question adjusted for the edge effect? I have the study in question, and I don't see any adjustments. Perhaps you have a more complete document, and can thus present the adjustment ctriteria. I would like that very much indeed.
As for modern research, I would not say it has nothing new to offer with regard to failure modes that can be applied to FH armor -- as in ongoing brittel materials research. Indeed, I think modern ballistics scholars have ample material that could inform many views on all aspects of the ballistics family of studies, but it internal, external or terminal. For example, they can even model how yaw effects influence projectile trajectory deviations in external ballistics problems. Now that might be useful to a lot of researchers. But to each his/her own.

George

[José M. Rico]

George, do you have those Tirpitz plate results at hand? What results were actually obtained from those tests? Just curious.

[George Elder]

Neil sent me a copy. It may have been scanned for TechSpec, but I will check to see. This may take me a day or two because there are hundreds of files to sort through. If it hasn't been scanned, I will do so. Anything for the archive.

George

[George Elder]

Hi Jose:

As I look through these files, I am finding 1000s of pages of material that has not been scanned, some of which is quite valuable. Heck, I even have reports concerning welding problems with STS aboard the USS Mass. when she was being built. It strikes me that I might never get this material into digital form, and thus I ponder the possibility of us sponsoring a more massive on-line archive -- something in the 7-10 gig range. Then I can distiribute this material to intertested parties on the condition that they can have the origionals only if they scan them for the archive. What do you think of this idea?

George

[José M. Rico]

George: I asked you about the results obtained from those Tirpitz plate tests because I was wondering what values made you think the results were poor. Perhaps you expected those plates to offer greater resistance to 14" shells and that made you think the tests were not properly carried out? What was the actual resulting data of those 14" inch shells striking Tirpitz's armor plates? Penetration results, type (KC, Wh…) and thickness of particular plates, striking angles, you know... those kind of things. No need to post the full report here if it is too long, a few examples will be enough.

As for hosting scanned material your idea sounds fine, but I don't know if 7-10 gig is viable at all. I will contact you via e-mail in any case.

José

[George Elder]

... in the report, and it is not very long. Neil found the document, and it describes how a salvaged 320 mm KC plates from the Tirpitz was secured for test purposes. As for my expectations concerning a given test -- I hardly think I had any expectations. The results are the results, and one has to ask why a given result was the case. Indeed, ballistics trials are fairly strait forward, and the real problem resides in seeking to understand why they come out as they do. In some case, the reasons are found in detailed examinations of the test specimens, but not much was done in this regard in the study here. Certainly, not nearly as much as was the case with the British examination of Italian armor, which was as exhaustive as it was enlightening. Here we find the Italians may have produced the finest cemented armor of anyone. I hope this information is of some small use. But I strongly advise reading the report, which should be available in day or two.

George

[George Elder]

To their credit, our friends at the Hood site maintain an archive, and these are some of the data that I have cut & pasted for purposes of discussion only. These trials were done with 15" shells, but the odd procedure of using "Small plates roughly 4'-6" x 5'-6"" is described. A similar series of trials may have been done with 14" shells, but I cannot recall nor find the document. The thing to note here is that a 54" target width is less that 4 x the diameter of the attacking shell, and we will see below that little mention is made of any edge effects or of using any computations to correct for them. This is problematic in light of what we now know about edge effects. How this would influence the general findings is a matter for debate. I am quite content with the notion that there isn't a great qualitative difference between German and British armor, although British armor may have a small edge.

George, with a special thanks to Frank



Official Records
ADM 213/378: Tests of German Armour
Updated 15-Sep-2004
This document contains a transcript of Admiralty record ADM 213/378. This document is a record of post-war (World War II) testing of German armour. Additional comments have been made by our website staff where clarification is necessary (items in [brackets]). We'd like to thank Peter Klein of Germany for his invaluable assistance in transcribing the file for us. The original file is held at the Public Records Office (PRO) at Kew, London. It is reproduced here by kind permission of the PRO. Crown Copyright reserved.

Note: Not all sections of this transcription are complete. We ultimately plan on posting the entire document.




- 1 -

TESTS OF GERMAN ARMOUR

ACSIL/ADM/47/76C

D.N.C. 5/JULY 1946.

Armour Technical Committee.

Meeting 15th August 1946.

Item 4. German Armour.

(a) Reports on visits to Germany.

With the fall of Germany opportunity has occurred for visits to the German Steel Works and for examination of German methods in armour manufacture to be investigated. There have been various reports circulated to the Committee of this visits of Mr. D.E.J. Offord of D.N.C. and Mr. Walker of Messrs. Firth-Brown Ltd., to Czechoslovakia and germany, and those of Lieut. Commander B.R. Queneau U.S.N. to Dortmund and Essen. In regard to ship armour the reports stress the importance attached by Krupp, Essen, to low sulphur and phosphorus steel for armour manufacture. The Baoic (Basic (sic!)) Open Hearth furnace appears to be the main source of steel supply and elaborate precautions are taken to ensure deoxidation and desulphurization of the steel. The employment of a chromo recovery process for the manufature of armour at Krupp, Essen, is mentioned by Lieut. Commander Queneau U.S.N. in his report, it is believed that this process only applies in the production of low alloy steels for Tank armour and not normally for ship armour. The analysia employed by the Germans for heavy ship armour is similar to our own. In regard to manufacture of the finished armour plate it would appear that the German practice is to roll direct from slab ingot without forging, to avoid as they say, lamination in the finished plate. Treatment of the armour does not differ markedly from that employed in this country. For cemented armour the German prefers a deep face and produces this by carburising for a longer time than we do, thereby ensuring deeper carbon penetration. The process produces higher percentages on the plate face which would seem to be undesirable in the light of our own experience.

It is suggested that the Committee might consider in the light of additional evidence which may be available, wether there are any factors in the fundamentals of steel making, brough to light by the German visits which are worthy of investigation for British armour manufacture.

(b) Armour on Tirpitz

Armour plating removed from the ex German Battleship TIRPITZ was shipped to this country in August 1945, and arrangements were made for full metallurgical and physical examinations to be made by each of the three armour firms. Separate portions of armour (a) uniform thickness 12 1/2 " (b) tapered thickness 12 1/2 " - 6 1/2 ", from side armour of the vessel, and (c) cast armour believed from control tower or range finder hood were examined and reports have been received from the three firms as follows: -

Beardmore Ltd., Report No.462 11/3/46

Firth-Brown Ltd., Report R.L.C.4415 2/1/46

English Steel Co., Ltd., Report G.82 8/2/46.

A summary of the analysis of plates and mechanical properties of the German armour is given on the attached sheet which includes for comparison plates of British manufacture. The main point of interest on this sheet is the low sulphur and phosphorus and the greater reduction of area obtaining on the German armour.

The most important point brought to light by the investigation was the greta depth of hard face on the German armour. To the 300 Brinell level the German plates have approximately 45 - 50 per cent depth of face. In British plates 25 per cent - 35 per cent is the more usual practice.

The castings examined were of a low carbon analysis as [continued on next page]



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

- 2 -

follows, .15 per cent C., 3 per cent Cr., 0.3 per cent Mo., 0.1 per cent Va, having a maximum stress of 40 -42 ton/in2 and Ized 80 - 100 ft. lbs.

They were free of unsoundness.

Firing trials against a number of the armour plates from TIRPITZ have taken place at Shoeburyness.

The 12 172 " plates were attacked by 15" A.P.C. Shell at 30° angle firmpact, in two groups.

(a) Small plates roughly 4'-6" x 5'-6", to determine resistance to perforation and gather information on the suitability of small armour plates for shell proof.

(b) Large plate No.27885 roughly 11' x 9'-6" for penetration and perforation limits

Group (a) was used also to give an indication of the likely starting velocity for (b) where limited room only allowed of a maximum of 3 - 4 rounds being fired.

A summary of the results obtained for Groups (a) and (b)is as follows:-

(a) Small plates approximate size 4'-6" x 5'-6".

Velocity. Result - Shell Plate damage.

1539 fs. Perforation - whole Broke into pieces.

1507 fs. Stop - broke up on plate. Broke into two pieces.

1501 fs. Penetration - frags. front&rear. Holed. Broke into three pieces.

1456 fs. Stop - broke up on plate. Smooth bulge.

1450 fs. Stop - broke up on plate. "

1348 fs. Stop - broke up on plate. "

From plate No.27885.

1508 fs. Peforation - whole. Broke into three pieces and disced.

(b) Plate No. 27885. Size 11" x 9'-6" .

Velocity. Corrected velocity Result - Shell . Plate damage.

1497 fs. 1507 Perforation - whole Disced and broke away to top edge.

1454 fs. 1462 Penetration. Frags. front and rear. Hole 13" x 11.8"

1371 fs. 1379 Stop - Broke up on plate. Smooth bulge 12" x 14"

D.N.C. was mainly interested in the performance of the large plate No.27885 and it would appear from the results obtained that

Limit of perforation = 1485 fs. ± 23 fs.

Limit of penetration = 1420 fs. ± 42 fs.

For comparison with these results we had the figures for two British 520 lbs. C, plates used for proof of supplies

Firth-Brown Ltd., No.5020 3 1/2" Face 1389 fs. (Penetration).

E.S.C. No.9894 3 1/4" Face 1377 fs. ± 17 fs. (Penetration).


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

- 3 -

In regard to penetration limit, therefore, there was an apparant superiority in favour of the German armour of the order of 30 - 40 fs. This may have been an under estimate of the difference since in assessing the penetration limit of the German plate the possible error was as much as ± 42 fs.

It was interesting to examine the results of trials against E.S.C. 9894 in relation to those obtained on the German plate in order to determine if possible the the order of difference with more accuracy.

E.S.C. 9894 1373 fs. 1394 fs. Damage Hole 9.2" x 10.8" Plug/disc 19" x 14" Lip 1.5"

TIRPITZ. 27885 1454 fs. 1462 fs. Damage Hole 13" x 11.8" Plug/disc 22" x 13" Lip 2".

i.c. the damage were comparable for a velocity difference of 68 fs. in favour of the German plate.

Whilst it was not a strictly fair, on the above basis of comparison, to assume that, in regard to penetration limit, the same difference would apply it seemed reasonable to state that the German Plate No.27885 showed a superiority over the British armour of the same thickness of about 50 fs under attack by 15" A.P.C. Mark XVIIB shell at 30°.

Little information was available regarding perforation limit of British armour. Trials have been undertaken against E.S.C. No.9894 to establish the limit of perforation. The following results were obtained.

E.S.C. 9894. Damage.

Velocity Corrected Velocity Shell. Front. Back.

1. 1439 fs. 1463 fs. B.U.O.P. Stop Smooth bulge.

2. 1504 fs. 1528 fs. B.U.O.P. All frags. in front. Hole 12.5"x14" Disc 42" x 42".

3. 1555 fs. 1582 fs. Perforation (broken). Hole 20.8"x31.2" Disc 72" x 50".

Note in round 3 shell was recovered just behind the plate.

The assessed limit of perforation of this E.S.C plate was 1582 fs. (shell broken) on the above results.

The assessed limit of penetration of the German plate 27885 was 1485 fs ± 23 fs. (shell whole).

The remarkable feature of this trial is that the deep face on the German plate did not break-up the shell at perforation velocity. The lesser face of the British plate did so and gave a result approximately 100 fs higher for perforation.

The result on E.S.C. 9894 where a stop was obtained at a velocity of 1463 fs (corrected for thickness), is at variance with Proof of Supply result where the limit was assessed as 1377 fs. This is due to the fact that, at the trial for proof of supplies on 30th June 1943, the first round at bottom of ingot plugged the plate at a velocity of approx. 1400 fs (corrected for thickness). The general quality of the plate is felt is more keeping with a limit assessed on the second trial namely 1485 fs ± 35 fs (corrected for thickness).


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

- 4 -

In conclusion it seems that if Plate No.27885 and No.9894 are generally representative of German and Britsih armour then,

(a) Penetration Limit.

British armour is probably superior to the deep faced German Plates by approximately 65 fs. (If the proof of supply result is taken the penetration limit of British armour is about 50 fs. below that of plate No. 27885).

(b) Perforation Limit.

British armour is superior to the deep faced German plates by approximately 100 fs, with added advatage that the British armour breaks up the shell at bare perforation velocities.

The trial presented by the trial results, therefore, is a credit to British armour manufacture.

(c) Armour plates removed from MEPPEN.

The following armour plates have been removed from Krupp Proving Tange at Meppen, Germany and are available at Shoerburyness.

Plate No. Thickness Size.
٪
No.33084 3 ¼ N.C. 19'-9 ½" x 10'-6 ½"
٪
No. 33085 3 ¼ N.C. 19'-9 ½" x 10'-6 ½"
٪
No. 32438 3 ¼ N.C. 19'-9 ½" x 10'-6 ½"

No. 37363 4 ¾ N.C. 23'-9" x 9'-10"

No.29785 4 ¾ N.C. 17'-0" x 10'-0"

No. 29691 4 ¾ N.C. 17'-10" x 11'-6"

No. 42716 6.4" N.C. 9'-11" x 8'-2 ½"
ж
No. 33228 6.4" C. 20'-10 ¾" x 7'-8"
ж
No. 32551 8.7" C. 18'-3" x 11'-6"
ж
No. 34507 8.7" C. 18'-8" x 11'-8 ½"
ж
No. 29332 8.7 " C. 16'-9" x 11'-5 ½"
ж
No. 29863 14.4" C. 19'-5" x 10'-0"
ж
No. 29873 14.4" C. 19'-6" x 9'-10"

No. 29906 14.4" C. 19'-6 ½" x 9'-11"
No. 23787 17 ¾" 16'-3" x 8'-6" (shaped and curved).

٪ These plates are partly used.

ж Copies of the mode of manufacture of these plates have been circulated to members of the A.T.C. Committee in February 1946.

Arrangements are being made for certain of the plates to be fired at under standard conditions of proof, at the earliest possible opportunity as follows:-

(a) No.33084 3 ¼ N.C. 19'-9 ½" x 10'-6 ½" 8" S.A.P.C.B.C. Mark IVB/60°/1380fs.

(b) No.37363 4 ¾ N.C. 23'-9" x 9'-10" 15" A.P.C. Mark XVIIR/65°/1240 fs.

(c) No. 29863 14.4" C. 19'-5" x 10' 15" A.P.C. Mark XVIIR/30°/1560 fs.

and representatives of the Armour Firms and the Ordnance Board will be invited.

Consideration will be given to further trials when the above are completed. Unfortunately most of the other thicknesses of plates shown above do not lend themselves for trials for direct comparison with our own armour and suggestions how to make best use of the plates would be welcome.


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

- 5 -


COMPARISON OF BRITISH & GERMAN (EX TIRPITZ) CEMENTED ARMOUR.

ANALYSIS 520 LBS THICKNESS.

[George Elder]

... so I took the opportunity to clean up my postings to keep them focused a bit more.

George, the lame and confused.

[Dave Saxton]

What are these homogenious plates reffered to? Do we know when these plates were made, or much about their history? Is there any data such as the chemical composition? Is there any indication how the trials on these plates went?

[marty1]

Great material Mr. Elder. Thanks for sharing.

I do tend to agree with Mr. Jurens’ comments on edge effects and lateral confinement. Lateral confinement is well understood and fairly simple to correct for when interpreting APDS or long rod hyper-velocity penetration test data. I am familiar with some of Paul Lakowski's internet postings on the subject, and believe his emphasis on this effect are being somewhat over-blown or over-dramatized. A 16D or 17D sample width is typically sufficient to yield a P/L that is within 95% to 97% of P/L resultant from test samples with an idealized 30D width. Normalizing ballistic test data to account for non-idealized confinement is pretty straight forward. Of course the real question is whether or not lateral confinement considerations for hyper-velocity impact and hydrodynamic penetration via projectile and target erosion has any real applicability to our understanding of plate penetration via conventional AP and ductile failure and/or plug failure. Penetration testing of this era was often conducted at far less than idealized confinement geometry. But again why would one care about lateral confinement effects for rod penetrators when considering conventional AP projectiles and impact velocities that are well below the hydrodynamic limit velocity?

Edge effects have also been well documented and understood even back in the "oldin days" of WWII.

[George Elder]

Hi Marty:

But if you read the study, and another one involving 14" shells against Tirpitz plates, you will not see any adjustments due to edge effects used in the data. Indeed, the 14" shell trials were so out of norm that Nathan dismissed them -- as sort of odd-ball effects. The reasons may be due to edge effects, the methods used to cut the plates into such small samples, the prior stresses the plates were exposed to, or perhaps other factors. But the data do not correlate well with prior testing, nor with GKdos 100, etc.. So what is one to think?

George