Cast vs. Rolled Steel Tank Armor

Perhaps someone with materials engineering or metallurgy background could help me understand a few things about tank armor.

If you look at German tanks the armor, at least on the glacis seems to be constructed of rolled plate welded together. This is particularly evident on the Tiger II and Panther tanks. If you look you'll see that on the Tiger II the bow armor is dovetailed into the hull to make the welds even stronger. The turret of Tiger I looks to be a composite -- cast sides with welded roof. The King Tiger turret (at least the Henschel model) looks to be a composite as well - shield, roof and rear welded to cast sides.

In contrast American and Russian tanks tended to be cast, though the original marks of the Sherman had a cast hull, some later M4s had welded glacis armor. From the M26 on to the last M60s the Americans used cast armor exclusively, hull and turret, as did the Soviets.

Now it seems to me that casting as a manufacturing technique ought to be less likely to produce high tensile strength. Brittleness comes to mind.

I'm reminded of battleship armor. USN battleships were tougher than their displacements would suggest, partly due to the use of so-called Special Treatment Steel in the belt, turret faces, etc. in the South Dakota and Iowa classes (maybe the North Carolinas as well) I understand this steel was rolled and tempered.

Was the reason for casting for the ability to create curved surfaces? Or was it faster and/or cheaper?

 

Not quite my area of expertise, but it is my understanding that the large-scale welding of plate on the Pzkpfw 5 was part of the reason it took so long to make. Casting would be faster, although castings of that size present their own technical headaches.

 

Welds can also be weak points especially if a lot of vibration is involved.

 

That is a sort of "yes and no" here. I've been welding, stick, gas, and wire since I was a kid, and if done properly with the correct material the weld is many times stronger than the material itself. That was one of the tests I had to pass to start working at Research-Cottrell, weld a six inch section on two flats, and have the two sections bent to breaking. If the weld broke you flunked, the steel broke around the weld, you passed. This of course doesn't apply to ships like the Liberty type, since they occasionally mis-matched the rod and flux to the steel plate and when they had extreme temperature variations plus stress the welds would separate.

 

Agreed. That why I said "can". My impression is and you probably can verify or correc this that the thicker the material the more difficult a quality weld becomes. Als if a corner is formed then even if the weld is initially stronger the vibration can cause stress fractures either at the corner or adjacent to the weld depending on where the flex point was. In this regard, again, my impression is that the Germans had some problems especially late war where they were using slave labor in signficant parts of the process.

 

I believe that is quite true, toward the end of the war the Germans were having difficulty in a number of areas concerning steel hardening and their quality suffered in the face hardening as I understand it.

I just ran across a very illuminating article discussing armor from the period which covers the advantages and dis-advantages of AFV construction types.

Goto:

WWII tank armor (John W. Schaefer; Robert Livingston)

The second and third pages are the best as per this topic, but the others are good as well.

 

Dear Clint,

Thanks for the tank armor link. Fascinating read and very valuable.

 

Thought the Tiger l had rolled plate turret? Maybe Tigers horseshoe shaped turret was designed that way to impart the most practical protection available at the time it was designed. Don't think any other tank had a single piece rolled plate turret.

 

The choice is largely one of what you have as manufacturing capacity. Castings allow you to make complex shapes far more easily than trying to weld a number of flat plates together. It is also generally easier to do than bend or shape a thick flat plate into a shape. The Tiger I's turret was mentioned. That 'horseshoe' is a rolled (ie., formed by rolling / bending) plate. The machine to do that had to be as big as a three storey house.
You can bend plate cold or hot depending on its condition at the time of bending and how it was manufacturered. Some plate steel will bend cold, other varieties will crack if you try. The later have to be heated to near orange heat and then bent in a soft annealed state.
Heat treatment for hardening would be done after forming. Machining would follow this for the most part as heat treatment would change the shape very slightly.
Welding would also limit the type of alloys available as you have to match the rod to the plate for maximum strength. Alternatives like bolting and rivetting are obviously inferior methods.

For castings, the final product would be a finished piece as far as heat treatment etc., went. That is, you cast it, it cools, then you do your final machining on it. The biggest problem with a casting is controlling the cooling rate so that you get an even and fine crystal structure in the metal. Without good quality control a casting could have serious weak spots in it if the crystal size varied greatly or cooling was done unevenly (that's why 19th century and earlier cannon sometimes blew up).


As for protective qualities, flat face hardened or hardened plate has a 3 to 8% (roughly) advantage over homogeneous or cast armor plate up to about 3 inches (75mm) in thickness. From there up face hardening really doesn't buy you much.

In terms of manufacture, a large casting is probably faster to produce than a welded assembly, particularly if the shape is complex. Simple flat shapes would be easier to fabricate from plate however.

 

Hi Terry,

Isn't it the case that making thick cast plates was easier than thick rolled plates?

 

This doesn't answer any questions, but I took this snapshot last week & thought it may add interest : -

View attachment 13240

 

Yes and no. Plates are rolled after forming and usually go through a number of processes to take stresses out and make grain / crystal size more or less uniform. Castings are harder to control in that respect often having a varying grain size that can introduce weak spots. X-raying or magnafluxing the casting will tell you where these are and then you can spot heat treat the casting to remove this problem.

Castings are largely going to be limited in size to what you can build a mold of. Plate generally will be made in a uniform size dependent on the machinery available.

 

Small casting with welding is a far better process, and produces more uniform quality for much less effort. The Panther was the same cost as the IV after all, while the KT was the same cost as the Tiger. It makes it significantly easier to manufacture a vehicle, and the strength of each part is improved greatly.
In general the difference between rolled and cast armor is about 15%, with rolled being between 230-260BHN, while cast being 250-400bhn, which means its usually 15-50% less effective. The worst German armor was about 300BHN which is about 25-30% weaker, while standard soviet armor was almost exclusively in the 350-400 bhn range and so you see these great designs like the IS2 failing because they have half the equivalent armor thickness. Thats why you see IS2 being killed frontally by kwk40 at medium range, and 88mm at 800+m sometimes. Many of them only had 50mm equivalent armor strength.

This is also why you see some technically impossible feats by certain guns, such as a 75mm which penetrated the side of a Tiger I at 200m. This is only possible because the armor the 75mm was tested on was actually 15% better than that on the Tiger.
In general, this is a very large problem for comparing penetration performance. The Russians had no fixed armor quality for testing, they simply lined up plates gathered from enemy tanks and used them as reference. It makes sense in some ways, might as well test against what you actually see, but tank quality can vary hugely within the same model. This was particularly an issue in the testing of the 122mm of the IS2, which was tested against plated from some fairly brittle German tanks which had also possibly burned.
The Germans themselves tested against armor5 of around 275bhn, so the penetration values are not comparable to the west. The US used a standard 240bhn, so its values will naturally be lower, yet will guarantee a more reliable penetration.
There is also the issue of how likely a shell is to penetrate. Some countries require 50% of shells to penetrate in order to be considered the maximum penetration amount, while other countries will accept as low as a 25% level to consider a wepaon capable of penetration. Those charts are anything but guaranteed to work.

 

Jager, the difference is only about 3 to 8% up to about 75mm / 3". That is, as armor plate gets thicker the difference between cast and rolled plate armor slowly disappears. What is far more important is the manufacturing process and quality control. You have to carefully control cooling to minimize crystal size in the metal's structure as smaller crystals make tougher armor. Heat treating is normal for all armor. Face hardening only penetrates the metal for the first .1" or so. Behind that there is a bulk of tough but softer metal.
This is why I mentioned tests like magnafluxing or x-ray. Tank manufacturers today do these on each plate along with other tests. You can also do things like make the plates with powdered metal (car manufacturers started doing this with engine blocks) as this will make the crystal size far smaller and more uniform.
Brinell numbers alone don't tell the whole story on armor effectivenenss. There is alot more to this than just hardness. For example, too hard a plate will shatter under impulsive loads so, you have to strike a balance between hard and tough.

 

That is only a half truth, you will notice that it is about 8% less resistant for any given BHN, but they never made cast parts FLEXIBLE enough to be more resistant. Thats where that extra few dozen lower BHN makes a real difference.

When it comes to the hardness, Im not sure why you thought I was saying harder is better, because as you can see from my post, softer is more resistant.
BHN is not really hardness, but more about flow. They test it by pushing an object against the metal and seeing how far it flows outwards. They just measure the width of the crater in the metal. They dont measure bending or anything like that, so BHN is actually a kind of composite score of strength and hardness, so its very good at telling resistance to shells. As far as I have seen, no other measurement gives such an accurate general resistance in typical armor thickness of WWII.