Metallurgical Study Of Enemy Ordnance

[Dave Saxton]

Hi Marty,

Wh wasn't exceptionally hard compared to most homogenious armors. It was desiginated harte only to show that it was harder than Krupp's other similar, but softer material Ww. Krupp also manafactured several different homogenious armor materials, that were quite different from either of their versions of Wh and Ww. It was also likely that the Wh in 1938, was slightly different from that in 1935, and so forth.

Krupp wasn't the only supplier of homogeniuos armor in Germany. There was also Rhurstahl, Stahlwerkes, Walzburg and so forth. Each seemed to have made different types of homogenious materials, some designated Wh. These materials ofton differed from each other, although the German Navy assigned it's own quality control people to each manafacture. The quality control guide lines specified a min mechanical property. For example a min of 18% elongation for Wh and 14% WsH. These min specs should not be taken as typical for the materials worked into the ships. Rhurstahl had a homogenious naval armour also designated Wh. Usually subfixed with an R. Wh-R was known to be much harder than Krupp's Wh.

Class B was a homogenious armor similar in composition to the material used in KC, but of course without the face hardening. Krupp discovered prior to the turn of the century that adding nickel and chromuim to steel improved toughness and strength. Nickel is the primary alloying agent used. In these materials, we find that it used about 3-4% nickel as the principle strengthing agent. Nickel is used with Chromium in about a 2 to 1 ratio. This is the classic Krupp armour formula. Sometimes Mo, and/or V was used as grain refining micro alloys, such as in HY. Each steel company world wide, had it's own variation on the classic theme. Carbon and Mn could differ from Vickers version compared to Krupps and so forth.

STS of the WWII era used lower carbon concentrations to improve ductilty for welding, although Carnegie, Midvale, and Bethlaham each had their own secret recipe. The USN during the war really only cared about min mechanical properties, and how each company meant these requirements was up to them. (Actually there were substitutes called NE30 and NE80 series steels used.) Traditionally, armour based on the oringinal Krupp armour formula contained about .3% C and .3% Mn. In STS, Mn is bolstered, and the C was usually lowered to slightly less than .2%. Actually STS is much like Stahl 52, but with relatively high nickel and chromium content. Stahl 52 is an HSLA material with low carbon content (.12-.20%) and about 1.4% Mn. St 52 is not a normal high tensile steel, because it is micro alloyed ( usually it's Cu and Al). Using higher amounts of Mn is good idea, if the steel can be relatively clean of S and P contamination. If the S levels are too high, then the Mn combines with the sulfer creating maganese sulfide gobular inclusions. These MSNM's reduce the ductilty of the steel in the transverse axis (but can actually increase it the logitudnal rolling axis), and create delaminations in the micro structure. They also cause a type of welding failure mode known as "laminar tearing" of fillet welds. LT's could have been a factor in some WWII weld failures.

In Wh, we find that the principle alloy's are molybedunum and chromuim. Not nickel. Nickel is used in rather small amounts, and Chromuim is incresed to about 3% and Mo is at about .5%. Both Mo and Cr are strong ferrite formers, so the micro structure can be what is now called duplex, with a certian type of ferrite called acicular ferrite involved. It's not a high nickel content armour like most others. This renders such statements that we ofton find in the secondary literature like: " Wotan was just another improved nickel chromuim alloy, like everybody else was using, being no better or worse", rather silly.

Modern RHA has been decribed as a type of 40 series chromemoly steel, but this rather misleading. I have seen the composition break downs for the RHA used on modern British tanks, and it quite clear that it not just a regular 4130 chome/moly steel, although it is chrom/moly alloy. Chromuim can be used to retain strength, when carbon is reduced. Krupp Wh used C at about .27% and the Mn was kept down at about .4%. Ww had a C content of about .22%. Modern British RHA is more like Wh, than STS or NCA, in that regard.

Modern USN HY materials ofton carry the same material numbers as Class B, although they seem to resemble the WWII Italian materials in many ways. Italian AOD used very high C content in comparison though. HY 130 uses less than .1% C. HY 100 is similar to Ww in C content, but it is still a high nickel content alloy.

[José M. Rico]

I wonder if we should ask Javier to move this thread to the Naval Technology portion of the forum?

Good idea! At this moment I am the only one who can move topics from one forum to another. I will do that asap.

[Dave Saxton]

I have obtained some internal company documents of some German industrial companies, through an Italian source. The nice thing is that they are already translated into English.

One document gives considerable insight into the German approach to metalurgy. It's a paper by two Thyssen AG metalurgists. Although it is based on how to best develop improved versions of St 52, it highlights several principles that were clearly applied to Wh. The Document is titled: "Review of High Strength Alloy Plate Metalurgy-Alloying, Normalizing, and Controled Rolling." The authors were Lutz Meyers and Harard de Boer.

"..In Germany the producers and the users have agreed that flat rolled products, which would considerably alter their properties as a result of normalizing are to be called "thermomechanically treated steels" or TM steels. TM treatement should be understood as a sort of very strictly controlled rolling..."
(I am going to quote something out sequance, because it further defines TM treatment. DS)

"..the essential points of TM are as follows:
-Defined tempature of the slabs and solution heat treatments.
-Defined deformations (rolling) at given tempatures
_Defined cooling rate during and after the rolling process

What does this treatment mean for the plate properties?
-Increased Strength
-Improved brittle fracture resistance- In creased directionality of properties
-Tendency to split up of fracture planes...
...TM treatement is limited to light and medium thickness. Very thick plates, greater than 15cm, require standard normalizing."


"...Mn, Si, Ni, Cu, Cr, and Mo which can be added to the steel in quanties between a few tenths of 1% to well in excess of 1% are the most convential elements...as they influance the mechanical properties, especially those of normalized steel, through their solid solution in the ferrite, the solid solution strengthing has frequently been dscribed the decisive factor..(Mn and Ni) have a positive but weak influance, But for normalized steel, and to a considerable extent for steels subjected to TM and particularly Q& T plates, the change in grain size and disloaction density play crucial roles, in addition to solid solution stregnthing..

...Grain refinement, preciptation hardening, and dislocation strengthing are the most important mechanisms for high strength steels...depending on the composition, rolling process, and methods of heat treatment, an alloying element can have considerable differing effects on the mechanical properties....only carbon need be mentioned here. Like nitrogen it can however, have effects differing greatly from those shown, if it is not in solid solution, but is bound to iron and other alloying elements like Mo and Cr. In the presence of carbides and nitrides, carbon and n are part of of the (overall)alloying concept. .....The varying effects which an alloying element has on the structure and the mechanical properties of steel, as it's composition and treatments, change in the case of micro alloying elements...here it will be sufficient to point out, that niobium can bring out a especially high dgree of grain refinement, and is preffered to V (although) V can have the effect of precipitation hardening, even at low austenizing tempatures. Titantium is characterized by the possibitiy of controlling sulfide shape....


.....with the introduction of Al deoxidization and micro alloying with St-52, more importance has been attached to normalizing...elements having a strong affinity to C, such as micro alloying elements (V, Cr, Mo, Ti, Cu, Nb..) can cause carbide preciptation too...These particles act as nulcei during the phase transformation, and they pin the grain boundries. The yield strength of the plate increases, and the ( notch brittle) transition tempatures decrease mainly owing to grain refinement...

continiued....

[Dave Saxton]

"....For the producer developing new steels, and for the customer, the fine grained Steel ST 52, having a min YTS of 36kg/mm2 is the starting point...Steel St 52 contains about .2% C. Although carbon has been known and feared as being harmful to welding in the form of pearlite, it is still a major factor for the strength of normalized steels.. Although Mn and even Ni improve the transition temp. a limit must be set to their use by the increase in hardnenablity (This may help explain the German move to CR/Mo materials from Ni materials. DS)...A certian compensation or supplementation of the classic alloying elements is possible through the use of micro alloying...consequently steels having a min YTS of about 45-52kg/mm2 can be produced with good weldabilty even though the welding process must be strictly controled....."

"......The concept of introducing better strengthing mechanisms instead of relying on pearlite can be logically continiued in normalized steel if the carbon % is lowered a few hundreds of a %, and Mo, in addition to Cr, and Mn, with a strong grain refiner, like Nb or V, is added to achieve transformation into low C bainite and acicular ferrite. YTS exceeding 45kg/mm2 can only be obtained after very careful temper anealing, since in normalized condition the yeild point is strongly depressed owing to the bainite constituents of the structure...

.....In addition to contruction and welding measures to control heat input, the avoidance of elongated inclusion stringers, especially manganese sulfides, has proved to be the most important step towards improvements in toughness in the through thickness direction.... figure 5 explains which metallurgical possibilties in this regard:
1)Desulferization.
2)Sulfide shape control
3)Desulferization with simultaneous sulfide shape control....."

"...SEM photos from fractures of through thickness tensile specimens of plates of of St52 (fig 5) show that sulfide controlled steel, with a satasfactory degree of oxide inclusion cleanness, is not suseptable to laminar tearing, under stress in the through thickness direction. ...."

[Dave Saxton]

"...The role of the micro alloying elements in TM treatments is much more versatile than it is for nomalizing...In the solid solution they bring about grain refinement, and dislocation strengthing by lowering the diffusion rate, and the austentite transformation tempature. In the form of preciptated particles, they pin the grain boundries in the austentite, and during phase transformation: in the ferrite the dislocations can be locked when the dispersion is extremely fine. This again manifests itself in grain refinement, but also in a strong percipation hardening..."

"...an increasingly severe treatment can be achieved by means of rolling measures, such as heavier rolling at lower tempatures, or by metallurgical measures, such as lower the tempatures of the phase tranformations, and of the beginning of re-crystalization. A high rate of cooling from the austentite, and an elongated cooling rate after, adequate alloying of the steel figure among crucial measures..."

"..The YTS of TM treated plate is considerably higher (than normalized), but the transverse value increases at a much high degree than the longitudinal value. (this indicates relative shear strength improvements. DS) The anisotrophy of the transverse YTS can amount up to 10% The influance of TM treatment on the form and position of the (sharpy) impact transition curve is very marked. In TM treated plate the brittle fracture begins at a much lower tempature....generally delaminations do not occurin the upper shelf of the ductile fracture range or in the brittle fracture(tempature) zone. The cause is a splitting up of the previous elongated austentite grain boundries due to percipitation. This more or less clearly transfered to the acicular ferrite structures. These former grain bounderies which are parallel to the plate surface would potentially tear more easily than the ferrite grains themselves."

"...Fig 11. shows the extent of the YTS increase due to decreasing transformation temps of fine grained steel. Here all the strengthing mechanisms important for the high strength steel are effective. The grain size decreases, and the dislocation density increases considerably, especially the transition to a non-polygonal ferrite structure... the schematic in the upper right corner of fig 11, suggests that a lowering of the transformation temp can be about through careful alloying. Mn and Mo are particularly important and help delay the austentite transformation, shifting it toward a lower tempature. The transformation products have increasingly non-polygonal acicular character and they can be regarded as bainite. These micro structures show a very high strength, with excellant brittle fracture resistance, as a result of the dislocation density, and very small effective grain size. Typical of the yeilding behaviour under tension load is the absence of a pronounced yeild point; the steels have a continious stress strain diagram..."

[Dave Saxton]

".. in addition to hardening by carbon nitrites in the micro alloyed steels on MnMo or CrMo basis, it is possible to make use of supplimentary percipitation hardening by copper..such steels may also contain increased Ni percentages... ( this probably explains the use of Ni in more than residual amounts, in the Cu containing Wotan, although it is considerably less than the Cr %. However, only trace amounts of Ni is used in most other German homogenious armours of the period. DS).... The copper precipitation only takes place after the under cooling of the solid solution in the tempature range of ferrite formation, so a tempering is needed to produce hardening (this would explain an increase in hardening through re-heat cycling. DS) .... Good deformablity is characteristic of treated steels. Using sulfide shape control, it is, nevertheless, possible to improve the ductility considerably, especially in the the cases of stresses in the transverse to the longitudinal axis of the plate..." (elongation in quer. DS)

[Dave Saxton]

In the Meyers and de Boer paper, we find numerious comments on the through thickness ductilty, and/or the yeilding strength or ductilty in the transverse axis. This pertains to shear strength. As a general rule, the shear strength of most steels is about 60% of the UTS. This usually works out that the shear strength is roughly the same as the YTS, in most cases. Simply increasing the UTS, will increase the absolute shear strength. However, we find, that in these specially rolled moybdenum alloys, the transverse strength and ductilty exceed the YTS by as much as 10%. This would mean that a material like Wh would have significantly higher absolute shear strength than obtained by the regular 115 ksi nickel armours. Something in the order of about 20ksi. The difference compared to older HTS type deck plating is enormous. Shear strength is crucial to the viabilty of spaced array protection schemes. If a two plate system is to have an effective thickness, equal to or exceeding, that of a single plate system, with the same sum thickness, shear strength is of critical importance. Some of the desirable charcteristics of a spaced array can be utilized by the designers more effectively when such a material is used. This may have been a major factor in the development of Wh.

[marty1]

Very informative Dave. Thanks for taking the time to present this material. A couple of stream of consciousness questions occurred to me while I was reading through your dissertation. Any insights you may have would be of interest.

1) The US -- in theory – produced RHA to varying hardness levels based upon plate thickness and the ballistic threat most relvent to a particular plate thickness. Was there a similar design approach on the part of the Germans with Wh or Ww armor? I guess what I’m driving at is what would typical hardness ranges for Wh and Ww have been.

2) Did the Germans employ a Brinell system for measuring hardness – or did they employ something of their own particular flavor? Same question for Charpy – were they routinely conducting notched impact testing on armor steel?

3) You mentioned that Wh was a low nickel content armor. Was this a function of strategic alloy shortages during the war? Did the pre-war Wh recipe change during the war such that nickel content was reduced or did the low nickel content have noting to do with any shortages of nickel in war time Germany?

On St-52 – I’ve seen a fair amount of modern long rod penetration test data in which St-37/52 was employed for the target blocks. Elongation for the experimental data I have examined is typically indicated as 22% (St52) to 25% (St37) -- with BHN being indicated as 180(St52) and 135(St37).

Are you familiar with a more modern armor steel material HzB,A? No elongation data was provided but hardness was in the range of BHN255.

[Ulrich Rudofsky]

2) Did the Germans employ a Brinell system for measuring hardness – or did they employ something of their own particular flavor? Same question for Charpy – were they routinely

The answer is yes. The rules for testing are spelled out in the naval administrative bulletin 147-27

See: Procurement Regulations for Armor Plating. Marine Dienstvorschrift 147-27.
http://www.kbismarck.com/archives/mdv147-27.html

[George Elder]

.... the Germans were well aware of scaling effects, and thus found it best to avoid using Wh over 150 mm in thickness. Appearently, they felt a point of deminishing return was reached when thicker strakes of Wh were employed. Moreover, manufacturing thick armor is difficult from a quality control perspective -- and perhaps even more so for a product such as Wh (?). One notes here that we seldom see deck thickness exceeding 150 mm in any German BB designs, with the preference being to beef up the outter layers. Dave may have discovered one rational why this might have been the case. Some very good work here.
What we realley need is a comprehensive technical treatise as to exactly why Wh was evolved. I mean, we need an archival document that details precisely what properties were being sought and why they were being sought from a terminal ballistics perspective. I believe Dave is on to some neat ideas, and I would like to hear from some 1930-era German designers on this. Ulrich, where can we get this information?

George

[marty1]

Thanks for the link Ulrich -- and thanks for your translation work on the DELIVERY REGULATIONS FOR ARMOR MATERIALS.

[Ulrich Rudofsky]

When you read it, you may find some things that are incomprehensible or could be improved or corrected. Afterall, I was a pathobiologist, and this type of stuff is quite difficult for me to understand.

[Dave Saxton]

. Was there a similar design approach on the part of the Germans with Wh or Ww armor? I guess what I’m driving at is what would typical hardness ranges for Wh and Ww have been.

2) Did the Germans employ a Brinell system for measuring hardness – or did they employ something of their own particular flavor? Same question for Charpy – were they routinely conducting notched impact testing on armor steel?

3) You mentioned that Wh was a low nickel content armor. Was this a function of strategic alloy shortages during the war? Did the pre-war Wh recipe change during the war such that nickel content was reduced or did the low nickel content have noting to do with any shortages of nickel in war time Germany?

On St-52 – I’ve seen a fair amount of modern long rod penetration test data in which St-37/52 was employed for the target blocks. Elongation for the experimental data I have examined is typically indicated as 22% (St52) to 25% (St37) -- with BHN being indicated as 180(St52) and 135(St37).

Are you familiar with a more modern armor steel material HzB,A? No elongation data was provided but hardness was in the range of BHN255.

Hi Marty,

Ulrich, is right about the German concern about impact toughness. All of the German research testing data from the period, that I have seen, include toughness trials. During that period they used a metric measurement: kgM/cm2. This is usually at a given temp. Ofton it was at -20*C. (BTW elongation ratings are also at given temps) The modern metric toughness measure is Jouls. St52 impact toughness measures from the 30's and 40's run about 12kgM/cm2. Assuming 12kgm/cm2 is eqivilent to 120 J, then this translates to about 90 ft-lb's. This seems right in line. The krupp material that closely matches the composition for the Tirpitz Wh, from 1931; tested at 18.8 kgM/cm2. This would translate to ~133 ft-lb's.

I'm familar with HzB, A, at least I have seen ballisitc welding research using that material. 255 Brinnel, seems a reasonable estimate for a material like Wh. IIRC, it was in the mid 200's Brn. The .17% Cu would allow a greater hardness, with no loss in ductilty, in the normalized condition. Ww would be significantly softer. Probably in the low 200's, at about 115 ksi UTS. Most German hardness data from that period use the Brinell system.

Is the 22% elongation for St 52 in 8", or in 2", or is it a transverse axis measure? Most, long in 2" measures for ST 52, I have seen in German research from the WWII period are in the upper 20's to lower 30's, but the the transverse rating is in the lower 20's. If it is for 2" long, then it most likely is in the normalized condition. St-52 utilizes much more copper than Wotan. Indeed, it appears the principle micro alloy. Most St 52 Cu concentrations exceed .3%. Some German composition tests for St 52 from the WWII era have about .45% Cu. According to Mathews and Rosenthal's findings, this is more than enough to cause significant copper percipitation strengthing, with a gradual ductilty reduction by normalizing. Most likey in normalized condition, the UTS of ST 52 would be about 62 kg/mm2 (88 ksi).

I'm kind of sceptical about the shortage theory of the German switch from nickel alloy homogenious armour, to molybdenum alloy armours. First of all, the Krupp research into high yeild strength Cr/Mo alloys dates from at least 1931. This was in peace time, and even before Hitler came to power, and began planning for another war. Moreover, molybdenum, is far more scarce, harder to obtain, and more expensive than nickel. Most of the worlds entire supply comes from the climax mine Colorado. If the Germans were so concerned about possible war time shortages, then they would not have developed such dependence on the material in the 30's. The truth is that Mo is indespensable for what they were trying to do. Meyers and de Boer's comment: "The concept of introducing better strengthing mechanisms instead of using pearlite if the carbon percentage is decreased.." weighs heavily here.

Wotan was suppossed to be a weldable material, and this requires the reduction in C content. This would put a cap on the potential UTS of about 115 ksi with about 3-4% Ni and 2% Cr, because nickel intensifies the effect of the carbon on strength. Chromium can replace the lost carbon to a large degree in this regard, and Mo will work with the remaining C to produce the percipitation strengthing effect of acicular ferrite. Increasing the Ni content with decreased C, doesn't do much. Steels alloyed with only Ni, in as high as 9%, only reach a UTS of about 110 ksi.

[marty1]

Dave:

The material I have for long rod penetration trials vs. St37 and St52 does not detail the rolling direction from which the elongation sample was derived, nor the length of the samples used in the pull test. I assume what you are asking is whether the elongation test sample was extracted along the same plane as the rolling direction or perpendicular to the rolling direction.(?) I’d guess that the more critical axis to penetration would drive the direction of the elongation sample. In the case of penetration via erosion I think elongation might only play a significant role in the back surface plug development. So my guess would be that the elongation sample is perpendicular to the rolling direction. But this is just a guess.

Keep in mind that these are St37 and St52 samples from circa-1975 – 1977.

UTS is indicated as 0.61GPa for St-52, and 0.46GPa for St-37.

What the authors – V. Hohler & A.Stilp” -- indicate is:

“The target materials are steel St37 (BHN = 135 +/- 20 kp/mm^2, sigma = 46 +/- 7kp/mm^2), St52 (BHN = 180 +/- 20kp/mm^2, sigma = 61 +/-7kp/mm^2) and armor steel (HzB20, BHN=260 to 330kp/mm^2, sigma = 90 – 110 kp/mm^2). The targets are square plates with thickness which are always greater than twice the penetration depth p.”

[Dave Saxton]

Sometimes the context for "transverse" elongation ratings, seems to actually be the through thickness direction. At other times it may be the direction perpendicular to the rolling direction?

The French place much less importance on elongation than do the Germans, or especially the Americans. For the French, the more important ductility measure is reduction of area. This may be because they got bit in the past, by depending on longitudinal elongation to indicate the ductilty of a material? They had a material that measured 23% elongation at 110,000 psi tensile, but it's transverse elongation, at the same time, was only 12%. Later research determined that the problem was caused by manganese sulfide inclusions. They now ofton use a ductilty measure that is a ratio of the reduction of area by the transverse elongation.

St52 from the 70's is essentially the same as German St52 from WWII. Most compositions charts are virtually the same. Sometimes the C content is lower, but I have seen WWII era St52 composition charts as low as .12%C too. Modern and WWII St 52 seems to run about 610N/mm2 UTS in most cases-that is 61-62kg/mm2. The 70's tensile and brinell hardness data seem as expected.