Best World war two warships?

delcyros said:

That´s explainable by the increased accuracy and the different environment of firecontroll post war. I can assure You that the closest you could come with ww2 era firecontroll is hitting at direct fire only (no FC required, only point blanc range) or -most usual- straddling at anything else involving FC. With a mean salvo deviation of +-200 yards and more in range (68% of all falls are in within this distance to the MPI), there is simply no possibility to ask for a hit. You either straddle a target or You adjust the MPI acordingly to straddle. The hit on the other hand is a random event in within the probability limits of the straddle itselve. You may manipulate the probabilities a bit (that´s why half salvos are preferable to full salvos), but You can´t tighten the dispersion pattern beyond the limits of the gun / turret / projectile / propellant combination. The dispersion pattern for major calibre projectiles at realistic fighting range, however, is always much larger than any ship buildt in ww2.

Click to expand...

Del

the differeneces are not that great, all of these variables you are talking about exist in the post war environment as they did during the war. that was not the point of my post. It was merely to point out that a straddle was not a hit, and should not be read as a hit. Moreover increasing the chances of a hit are connected to accurate FC. If the MPI is not estimated correctly, or it takes a long time to determine range speed and course, then the probability of a hit under those conditions will drop .

Re-read my original post, and you will find the measure of success was not that a straddle was obtained, but the number of hits over the number of rounds fired. That we are arguing about how to measure success in gunnery surprises me to be honest A straddle was a step in the right direction, but is still not a hit. We incidentally still needed to determine the target range course and speed to achieve a hit, and this often resulted in a straddle before registering a hit. I dont think I communicated this addequately in my previous posts. I was merely reacting to the inference that a straddle or a near miss was "as good as" a hit. Its not, either statistically or operationally.

The measure of accuracy in my opinion is how quickly a hit can be registered after fire is commenced. In Barents sea, Lutzow opened fire on Obdurate at just before noon, and the damage on the ship observed at 1204. though not a hit as such (splinter damage only) that suggests to me about 3 to four salvoes before any result was obtained. the range was 15400 metres, and visibility was reported at 14000 metres. I am unsure if there is an error in the range visibility reports, or if Lutzow was using radar to control her guns.

It would be interesting to compare that admittedly impressive effort with the RN cruisers efforts against the Eckoldt. I forget the details, but the time between the eckoldt being engaged and her sinking were very short.

Parsifal,

Eckoldt was sunk in an odd engagement. It basically resembled the 1880 -1910 era of fighting ships, not ww2. In this era you should view the QF guns as the main armament and high volume of fire would overpower eventually any opponent from close range. Both RN cruisers were firing from point blanc range and rapid, direct fire under local turret controll except for the first directed ranging salvo. This condition is not directly comparable with either HIPPER or LÜTZOW. It is not an indication for good firecontroll but demonstrates exzellent situtational awareness and the skill to exploit a tactically favourable condition.
Altough the measure of success often is called a hit, it gives a wrong impression about the accuracy and firecontroll ability. A hit is a random event in within certain limits of probability. You may get lucky (PoW was at Denmark Street, it got three hits our of three straddles, making each straddle a hit) or unlucky (Bismarck got one hit out of each two straddles in the same engagement). The number of hits compared to the number of shots fired is therefore not a useful description of how good firecontroll was in a comparative approach. In my example it makes PoW looking much better than she indeed was (regardless of the outcome of DS), criss crossing her target several times and unable to keep a good track of her target. Bismarck however, consistently straddled both her targets, even if she did not achieved the the high straddle to hit efficiency out of her many more opportunities to get hits. In within the range of probabilities of the small sample size, both conditions are "true". But to expect that in a prolonged action PoW would always enjoi such an amazing hit to straddle probability would be highly unrealistic.

You cannot exactly determine the correct position for the MPI in within a scatter of shots which may be placed randomly in within the dispersion pattern. You may get an idea where the MPI may be in prolonged actions but unless You are firing against immobile targets, You hardly can be sure of this due to a number of involved factors. Once You have straddled You are already right on target with ww2 FC. In within the operational doctrines a straddle in ww2 was a correct mark on target. A repeated straddle always marked the begin of effective fire.

Also operationally, a straddle may result in a hit. The 11in base fuzed HE used by Lützow against the RN DD´s have a very large splinter radius. They don´t need to hit the soft skinned target physically to inflict damage. During River Plate, Ajax received a damaging near miss (actually from GRAF SPEE´s 11in HE straddling her), which put a turret our of action for the remainder of the action. Nathan Okun demonstrated once that even from 1000 Kal. miss distance (e.g. 283m or ~300 yards), the 11in HE still has chance to pierce the ships unarmoured hull due to splinter forces. The probability for such an event is remote but 300 yards +- is quite a distance in excess of salvo dispersion patterns of this period. Putting the hit closer to the ship increases the danger a DD (or for that matter a merchant as well) is exposed to.

I still have to respectfully disagree about the number of Hits being a poor measure of firing accuracy. Theoretically if all the known variables are accounted for, we would not need a probability model at all. We would know all the variables, and would achieve a hit with every shot. However the more unknown variables there are, the less we know about the problem, and therefore try to compensate for this by using innaccurate mathematics which we call "probability". The more that the unknown variables can be reduced, the greater the certainty of achieving the intended outcome.

Applying this theory to the gunnery problem, the more accurate the information on the target, the higher the chance of achieving a hit. Therefore, despite what you describe a hit as a random event, in fact it is not a random event, its just an event that cannot be predicted with the tools at hand. However if the tools at hand are made more accurate, then the "Unknownness" (which you call randomness) of the hit event drops, until at some point it is not really an unknown variable anymore

Now the question is this, were the tools available during the war capable of eliminating a hit as an unknown outcome. I dont think they were, because in the post war environment, with superior methods and technology we still could not eliminate the randomness of achieving a hit. This part accords to what you are saying...

However the chances of achieving a hit can be increased, if the accuracy and precision of those tools can be improved. You mentioned the circumstances of the Eckoldts loss, which is very true (though I believe she was still some thousands of yards away from the british Cruisers when lost....her consort beitzen did escape....if it had beeen an 1880s style engagement, it was unlikley the beitzen would escape), nevertheless this was still part of the probability matrix for a successful hit for wwii gunnery....its just that the probability of a hit in those circumstances was so close to one, as to effectively render a hit at those ranges a certainty. At longer ranges the same could be achieved if the techniques the training and/or the technology could be improved so as to also increase that prob to 1.

So what are the ingredients that can improve the probability of a hit. This is obviously a matter of opinion, but my belief is that the main elements in the probability matrix are not limited to those you mentioned, they include....

1) The standard of crew training if human errors creep into any part of the operation, this would have an overwhelming effect
2) The speed that target data can be transferred from the point of data collection to the wepons platforms
3) The size speed and manouverability of the target
4) The target range
5) The visibility of the target This is where radar has its greatest effect
6) The sea state and the stability of the firing platform
7) The inherent accuracy of the guns firing
etc etc
My belief is that inherently in the latter part of the war, the British were more accurate than the Germans in their gunnery. I am sceptical that the Germans held some sort of technological advantage over their opponents, but even if they did, this is just one part of the equation. maybe, for example, German ships were less stable than the British counterparts, or their gunnery crews lacked proper training at the end of the war. I dont believe, however, that the Germans possessed as great a chance of hitting a target in the latter part of the war as their allied counterparts. Thats an opinion, its open to debate and disagreement of course

Some of this may depend on the ranges being fought at. At long ranges with angles of fall being high the ships salvo is a pattern. All shots with in a circle (or perhaps oval) so may yards or meters across. it was often double or triple the length of a ship let alone the beam of a ship. With a steep angle of descent the width of a ship was a bigger target than the side of the ship.
The size of the pattern was not adjustable. Once they had a straddle that was it. A few overs, a few unders and hopefully a hit or 2. A second straddle in a row showed they had the course and speed figured out. after that is a matter of luck or probabilities. All shells from a salvo fall in area XXX across for an area of YYY. target ship occupies ZZ% of area YYY. Firing 6 or 8 or 9 shells per salvo, what is the likely hood of one or more shells scoring a hit?

Better training or optics or radar cannot reduce the size of the pattern of dispersion at a given range. It may allow for figuring out ( or guessing) if 1 under and 8 overs means that the target ship, while straddled, is on the edge of the pattern and a correction needs to be made but if the number of overs and unders is fairly close that is as good a fire control solution as can be had.

DO not confuse this with firing "ladders" or other techniques of firing groups of guns (less than a full salvo at times )at different ranges in close timing in order to confirm the range quickly.

If the range is short to medium the size of the shot group's dispersion may be ship sized or close to it. In this case the number of hits should go way up if the salvo "straddles" the target but there is no way to tell if the 'shorts' or 'longs' are caused by differences in the shells or propelling charges, interference between shells in flight (reason a number of ships fitted delay coils to center gun in triple turrets), minute differences in ignition/barrel time or even (in combat) misalignment of the guns. Since the cause of the overs or unders is not known there is nothing the crew can do about it in action.

Hello Wavelenght
Quote:" There may well have been a conflagaration in the forward magazines, but in my opinion probably not…"

Yes, it isn't possible to say for sure, and to be exact, I meant a sort of conflagration in one of A-turret magazine, a total conflagration in A-turret magazines would probably have been fatal.

Quote: "Reading up on the various accounts of the battle, it looks like there were possibly two hits forward at that time; one vs the turret and one to the forward section on the tween deck. This is just forward of A-turret The tween deck hit started a nasty cabling fire that was the main source of the smoke."

Thanks for the info, That was new to me!

Quote: "A hull hit high up abreast A turret and passing over the belt, may have penetrated the barbet"

That is also my line of thinking.

Quote: "The Germans did have ready use ammo in the turrets…"

IMHO, not being BB specialist, especially main armament shells were so heavy that that would not be very practical. Of course the difference between 8" and 11" shells is clearly smaller than that between 8" and 14"-15" shells, especially when British 14" and 15" shells were heavy type. So in BBs the loading was wholly mechanical and having two ways to do it would made the system more complicated. After Jutland British were very careful with their ammo systems so it is entirely possible that RN didn't have ready use ammo in their main armament turrets of BBs but KM had.

BTW, on the British penetration tests, IIRC RN noticed that their heavy long body shells wobbled a bit when fired markedly lower MVs, in order to get right striking velocity for simulated long range hits, than that for which they were designed and because of that didn't give too much attention to the results.

Hello Delcyros

Quote: "Regarding the Scharnhorst account...With the shell decapped, the impact on the barbette would then result in full shatter, which might still cause significant local damage (up to including damage behind the barbette) but not a classical textbook penetration of the whole projectile in a condition fit to burst."

Entirely possible, but as Hood's 15" hit on the Turret II of Dunkerque at Mers-el-Kebir shows, full penetration wasn't necessary to produce substantial hole and cordite fire.

Juha

Very fine and informative discussion. Please keep up the good work!

Parsifal, I think your arguement that "perfect data produces a perfect solution" is accurate but not viable to a naval engagement. Delc's point that straddles are desired as they increase the chances of a hit are a realistic interpetation of naval gunnery, especially in something so unpredictable as a sea battle in the far North Atlantic.

Even an absolutely accurate weapon (if such a thing existed or exists) will miss due to the variables that are beyond the control of ship's crew. Sea state, scatter from the mounts/turrets firing, inefficiencies in the gunlaying equipment, haze, smoke, you name it. All of them would create problems with pinpoint accuracy to the level that a salvo straddle would be a viable and desired outcome.

Similar to artillery blanketing a target with rounds before an infantry attack, perfect information, perfect solutions, ect are almost impossible to obtain. Blanketing the target (in the methodical way a salvo does) is the best solution.

Again, IMHO.

Radar does make possible quicker and more consistent straddling at longer ranges. The range accuracy of even the best optical range finders is typically 1% of the range. Radar is much more accurate. The American 40cm Mk3's accuracy was 0.1% of the range give or take 40 yards. The British 50cm Type 284M's range accuracy was 240 yards, or by using a corrective template to account for distortion toward the perimeter of the CRT; 120 yards.

Another advantage of radar over optics is the ability to better determine errors in the MPI of the salvo patterns. This boiled down to the ability to discriminate between the target and the shell splashes. The range resolution of the Mk3 was 400 yards, so echoes of the target and splashes could not be readily determined if they were within 400 yards of each other . The Type 284M was markedly better than the Mk3 with a range resolution of 150 meters. It could be better determined if the MPI wasn't squarely on the target, and correction could be made, improving the chances. Late war equipment with shorter pulse durations was better yet. Post war advances would have made possible fine corrections in MPI errors easier.

The German Seetakt radar was actually well suited to firecontrol because of its unique methods of measuring and displaying range data. In practice the operator held the pip on the null mark on the CRT and the correct range was continuously displayed, updated, and transmitted directly to the central firecontrol. The operator could zoom in on and closely examine any portion of the time base for observation purposes.

delcyros said:

For the thicker Panzerdeck a material with high strain is the preferable application. Thinner plates could be treated for higher hardness without a significant drop in strain. The 30mm samples from TIRPITZ in UK tests are much harder than the 50mm deck plates too (probably for splinter bulkheads).

best regards,

Click to expand...

Looking into this interesting issue further, I find that too much emphasis is probably being placed on elongation as a reliable indicator of ballistic resistance and scaling effects by modern analysis. Most of the British ballistics tests of the Tirpitz Wh plates were done with 1900 lb+ 15-inch armor piercing capped projectiles. Yet the Wh plates still required slightly higher necessary velocities than the equal thickness British NCA in the same testing conditions. The British NCA was probably about 24% elongation, while the Wh would have been about 20% elongation. If the shells had been uncapped then the differential may have been greater yet. Krupp engineers stated that relatively minor variations in the elongation metric had a negligible effect on penetration of homogenous armor by capped projectiles. Another interesting finding by Krupp was that an equal weight of standard grade aluminum actually provided greater resistance to penetration by heavy projectiles than the same weight of steel homogenous armor. The elongation before fracture of aluminum is actually only 11-14%. The elongation of steel armor will be from 18% to 25%. Elongation % probably has some effects but I think some modern analysis greatly overstates it.

The Norwegian examinations also indicated that the splinter bulkheads were much harder up around 270brinell. Which would help them resist splinter perforation much better without resorting to greater thickness and increased weight expenditures.

I thought I understand quite a bit of this topic but let me express that I doff my hat for You, wavelength.
Thanks for Your inputs and sharing Your knowledge!

I reconsidered recently the RN tests of armour material coming from TIRPITZ and there are some interesting findings in them. Particularely if You compare the results with Nathan Okuns facehd series computation models. Did the norwegians also undertook ballistic tests with the samples or did they relied mainly on mechanical and chamical sampling and what were their findings on them? I am eager to learn more from this interesting source.

best regards,

I'm learning a lot about WWII aircraft from you and many others on this site. It has always been a special interest of mine, but I have much to learn.

As far as I know the Norwegians did not conduct any ballistics tests of the Tirpitz plates. The British tests of the Tirpitz plates are the definitive tests in my opinion. The Americans got different results from testing German armor and these are what I think Mr. Okun bases his calculations on. George Elder told me several years ago that the Americans tested a homogenous plate that was 437 mm thick. As you have stated, the quality of very thick plates is usually sub par compared to thinner plates. I would not expect this 437 mm plate to be representive of the Wh used on the warships at all. Also, late war, the Germans were forced to use ersatz homogenous armor in place of their earlier chromium/molybdenum based material. They were able to keep up the ballistic resistance within acceptable tolerances, by using vanadium in place of molybdenum, and greater carbon content, and so forth, but they never considered these ersatz materials to be as preferable as the original material.

The British always got different results from Americans in their tests, and I question some of the American testing procedures. Often the Americans were forced to test very limited supplies of enemy plates that may not have a large enough area to keep the impact far enough from the edge of the plate and so forth.

In a related issue, the British did not find samples of Japanese armor to be of poor quality but the Americans did. The Japanese armor was actually almost identical to the British materials in terms of chemistry. British naval shells had sharper points than American naval shells and would interact with the armor differently depending on the testing conditions. This would mean that the British shells would require less penetration energy than the American shells if the striking angle was close to the normal (right angles) but the American shells would perform better at more oblique striking angles than the British shells. This could result in very different test results.

The American Class-A armor will do better against smaller caliber soft capped shells comparative to the German KC, because of a thicker hard face portion. But the German KC will do much better than the American Class-A against larger caliber projectiles. In the British tests, the Tirpitz KC performed within about 3% of the British cemented armor.

The Tirpitz Wh did even better against 8-inch uncapped comparative to British NCA, than it did against 15-inch capped projectiles comparative to NCA. Does this mean a greater scaling effect with Wh than with NCA? Perhaps so, however in my opinion, it is the result of the Wh exceeding 80kg/mm2 tensile and is to be expected based on the Krupp finding that it takes greater energy for uncapped shells to penetrate when the tensile strength of the homogenous armor exceeds 80kg/mm2.

The German Wh armor seemed to be suited to a de-capping system. It would perform best in such systems. In a single plate system one may want to use a different material. Perhaps Ww, or the American class-B, or British NCA, would give better results against capped shells in a single plate system. I would almost bet they would. Comparing different armor materials among different systems can be very complex. It's just different ways of doing things, and one approach isn't always better than another.

Thought this would be a photo of of interest to this topic, found this photo online. Gives a good idea of the size difference between the Iowa and Alaska classes!

Battleship Missouri, battlecruiser Alaska, escort carrier Croatan, and destroyers at Norfolk Naval Base, Aug 1944
Source United States National Archives

I'll throw in my preferences as well. The only ones I can talk about are the following (my warship knowledge is not that great).

Battlecruiser - I've only heard about Hood, Renown and Repulse. I'll put in a vote for Hood, only because of it's fame as a capital ship.
Battleship - The greatest ones ever built, from what I know, I will say either the Tirpitz or Yamato because of what others have said on here - that they both withstood a tremendous amount of punishment from everything the Allies could throw at them. The sheer amount and size of the weaponry on Yamato was just unbelievable.

what a fine discussion this was. I miss these guys. I learnt a LOT from them, plus they were great guys to boot.

Hats off guys

Most WW-II tests of armor were useless by modern standards, which probably goes along way to explain why you get such wildly divergent results. I try not to take them too seriously and I don't think you should either.

Any formulation based on these results starts out on relatively thin ice and should be taken with large grain of salt. If the results reported do not include methodology of tests and decent standard deviations on results, then those results should not be trusted either.

1.) Battleship - Iowa
2.) Battlecruiser - Kongo
3.) Heavy cruiser/Armored cruiser - Baltimore
4.) Light cruiser - Sendai
5.) Destroyer - Sumner
6.) Submarine - XX1
7.) Patrol Boat or MBT - Elco type
8.) Aircraft Carrier - Essex

The only questionable choice here has to be Sendai as the best Light Cruiser. Why would you pick that class over so many others, such as the british Belfasts, the Italian Ciano Classes, Soviet Kirovs or Chapaevs even the Clevelands. by comparison the Japanese cruisers were old, slow and undergunned and very lightly protected and constructed

It might depend on wither you consider "british Belfasts, the Italian Ciano Classes, Soviet Kirovs or Chapaevs even the Clevelands' As true light cruisers or heavy cruisers armed with 6" guns to get around treaty restrictions.

In any case the Sendai doesn't stack up all that well against the British Arethusa class which is only a little bit bigger in size, or even the later Agano class cruisers.

Well, the ships i mentioned (esxcept for the Kirovs) were all consistent with the treaty limits for a Light Cruiser. If the treaty limits were removed the ideal light cruiser becomes the US Worcesters, the ideal Heavy Cruiser is the Alaskas, or the modified Baltimores with automated 8in guns, depending on the role being envisaged for your cruiser force. If you are considering capital ships as your main targets, you would build Alaskas, if the main prey to be Destroyers, you would build fast firing Baltimores. Possibly the best cruiser combination would have been a combined Alaska/Worcester force so as to engage or shadow both heavy and light forces targets....

I have to say that the USS Worcester is pushing the definition of WW2, she wasn't commisioned until 1948.