PoW salvos to hit Bismarck?

[dunmunro]

I took another look at this topic.

The first 2 salvos were spread 500 yds, the 2nd pair were spread 775 yds and the 3rd pair were spread 950 yds, which produced a hit.

The 2nd salvo would have had to have been spread 1500-2000 yds to have bracketed the target, and even then it is probable that the next two salvos would have been a further refinement. There is no logical way for PoW to have hit prior to the 5th or 6th salvo, without accurate rangefinder or radar ranges.

[Steve Crandell]

The first salvo was way over, probably because of weather/wind conditions and her short base rangefinder in the fore top. Radar at that point in the war really only gets you a more accurate first salvo, because resolution and displays weren't up to determining salvo MPI. After that first salvo, the next one should have been corrected to produce a short half salvo. This enables quicker straddling and potentially interferes with the enemy spotting effort. It was common practice in all the navies of the world. You are pairing up the salvos, which makes half salvos somewhat meaningless. Why not fire full salvoes if you are not going to correct each half salvo? If for some reason you are pairing them up, at least one of the second two should have been short, since both of the first two were over.

I am not the first person to notice this poor practice on the part of PoW. It was, in my opinion, the result of not having enough realistic gunnery practice prior to the engagement. I have seen examples of shooting where two or even three salvoes in a row were short or over (not great either), but I think this is the first time I've ever seen six overs in a row.

Bismarck, on the other hand, fired three quick half salvoes, all in the air at once, each one at a different ladder range. Standard German procedure. One or more was over and the other(s) were short. I don't remember which. The next half salvo was straddling and she commenced firing "good rapid" and sank the Hood. If PoW had shot that well the engagement might have turned out differently. She was very lucky because all three of her hits had important long term effect on Bismarck.

[Dave Saxton]

Radar at that point in the war really only gets you a more accurate first salvo, because resolution and displays weren't up to determining salvo MPI.

At least among the Allies I take it.

[Steve Crandell]

Radar at that point in the war really only gets you a more accurate first salvo, because resolution and displays weren't up to determining salvo MPI.

At least among the Allies I take it.

Well, the only one I know of is USN and that is beginning in '43 with the Mark 8 Mod 1. Before that you could kind of do it, but the range had to be less than 20,000 yds. You really need a plan form display to show MPI. An A-scope isn't going to do it, or any other display which only shows range.

[Steve Crandell]

I may not have been perfectly clear with my last post.

The Mark 8 mod 0 introduced the B-scope, but in practical terms was unable to spot targets or shell splashes over about 20,000 yds. This performance was similar to the previous Mark 3. It was first installed on USS Indiana in Sept '42.

The Mark 8 mod 1 increased output power from 25kw to 50kw and receiver gain by 8 db. It was capable of spotting out to 35,000 yds or so. It was first demonstrated in March of '43 and then installed on USS Iowa.

[dunmunro]

The first salvo was way over, probably because of weather/wind conditions and her short base rangefinder in the fore top. Radar at that point in the war really only gets you a more accurate first salvo, because resolution and displays weren't up to determining salvo MPI. After that first salvo, the next one should have been corrected to produce a short half salvo. This enables quicker straddling and potentially interferes with the enemy spotting effort. It was common practice in all the navies of the world. You are pairing up the salvos, which makes half salvos somewhat meaningless. Why not fire full salvoes if you are not going to correct each half salvo? If for some reason you are pairing them up, at least one of the second two should have been short, since both of the first two were over.

I am not the first person to notice this poor practice on the part of PoW. It was, in my opinion, the result of not having enough realistic gunnery practice prior to the engagement. I have seen examples of shooting where two or even three salvoes in a row were short or over (not great either), but I think this is the first time I've ever seen six overs in a row.

Bismarck, on the other hand, fired three quick half salvoes, all in the air at once, each one at a different ladder range. Standard German procedure. One or more was over and the other(s) were short. I don't remember which. The next half salvo was straddling and she commenced firing "good rapid" and sank the Hood. If PoW had shot that well the engagement might have turned out differently. She was very lucky because all three of her hits had important long term effect on Bismarck.

Firing 1/2 salvos whilst ranging gives you two data points instead of one, and I suspect that the theoretical hit probability was greater with two 1/2 salvos than one full salvo, and so it saved ammo and produced greater hit probability.

PoW and Hood were both supposed to be engaging the same target, so PoW could only fire during her allotted time sector. This reduced her ability to fire multiple salvos since they might begin to overlap into Hood's time sector. In theory with both ships engaging the same target and exchanging range data, the correct range would have been found very quickly, but as we now know, Hood was engaging PE, and for some reason did not relay her gunnery info to PoW over the gunnery wave radio link yet PoW probably still had to respect the time sector allotment since she didn't receive info to the contrary from Hood.

Hood's gunnery woes may have effected the outcome of the battle in more ways than we have previously considered.

[Steve Crandell]

Firing 1/2 salvos whilst ranging gives you two data points instead of one, and I suspect that the theoretical hit probability was greater with two 1/2 salvos than one full salvo, and so it saved ammo and produced greater hit probability.

There are a couple of problems with half salvoes. One is during the critical early parts of an action not all guns are at the same temperature and you often correct one salvo with guns of a different temperature and they go somewhere else. Then you try to correct that salvo with guns from the first salvo that are a different temperature, and so on for the first few salvos ... it takes longer for all the guns to reach their common hot state. If you are firing two or more half salvos as one correction, that wouldn't be a problem.

The other problem is related to radar ranging. Full salvos give you a better MPI determination. With only three or four splashes, one outlier can have a significant effect on where the fire control officer thinks the MPI is and so effects his adjustment. Also, sometimes you lose a splash because it is too close to the target, and that effects MPI determination. I'm aware of one practice shoot where a US battleship didn't get a radar observation of the salvo at all because all three rounds were very close to the target and were lost in it's radar blip. That is why US battleships used full salvoes in combat. Unfortunately they often (not always) used less than that in practice for economy reasons and it affected the quality of the practice.

[Dave Saxton]

Well, the only one I know of is USN and that is beginning in '43 with the Mark 8 Mod 1. Before that you could kind of do it, but the range had to be less than 20,000 yds..

We have KTB records of the Germans accurately directing shore batteries at night with Seetakt at considerable ranges as early as 1940 using the method Zielbeobachtungsschiessen. Z.B. Schiessen is spotting the fall of shot relative to the target with radar and correcting MPI. For example, shortly after midnight on June 18, 1941, Battery Grosser Kurfurst (4-11" guns) successfully took a British convoy under accurate fire using a GEMA radar for blind firecontrol. The method used was Z.B. Schiessen and the range was 336hm (36,700 yards).

You really need a plan form display to show MPI. An A-scope isn't going to do it, or any other display which only shows range.

While a plan form presentation lends its self well to showing MPI it is not always required. Seetakt used a unique type of bearing track presentation that could indicate fall of shot relative to left or right of the target, and of course the fine range presentation could indicate the fall of shot for range. The AVKS examination of Bismarck's radar equipment noted that the EMII devices could spot the fall of shot relative to the target, but to make MPI correction more precise; the AVKS recommended that graduated scales should be added to the fine range and to the bearing indication.

The manual for Mk-3 notes that the fall of shot could be spotted for range but it was extremely difficult to spot the fall of shot for bearing. A skilled operator could do it using the A-scope in lobe switching mode and noting which trace presented the larger pip.

The British Type 284 could spot the fall of shot for range but once again spotting for bearing was extremely difficult if not impossible. With Type 284M the trace flickered if the target was off line but held steady if the target was on line. British Type 274 could not spot the fall of shot at all, so a special spotting radar was added to Vanguard in 1947.

I have copies of the technical manuals for Mk-8. The 1942 edition states that 16" splashes could be spotted to max range of 20,000 yards. The 1943 edition doesn't give a max range for spotting but notes that 16" splashes had been spotted to 20,000 yards.

[Steve Crandell]

Obviously a radar mounted on a hill somewhere high above the surface of the water is going to be able to spot better than one at 150 feet. In fact, at ranges over 30,000 yds the aft director on the US battleships had more missed spots due to it's lower height.

The Mark 8 mod 0 (and the Mark 3) was essentially unable to spot splashes consistently over 20,000 yds. However, Bill Jurens has documented a number of actual gunnery exercises using the Mark 8 Mod 1 and above which observed the fall of shot at ranges above 30,000 yds. In fact, in early Dec 1943 Adm Lee had five fast battleships (all of the South Dakotas plus Washington) each fire ten salvos at a mean range of 25,600 yds and ten salvos at a mean range of 32,200 yds. The purpose of the test was to determine the ability of the then new Mark 8 Mod 1 to spot at long ranges. There were some rounds which were not observed, but the test was generally successful. At that time it was generally understood that optical bearings were superior to those generated by the Mark 8 radar, however. Ranges were of course better with radar.

[dunmunro]

Firing 1/2 salvos whilst ranging gives you two data points instead of one, and I suspect that the theoretical hit probability was greater with two 1/2 salvos than one full salvo, and so it saved ammo and produced greater hit probability.

There are a couple of problems with half salvoes. One is during the critical early parts of an action not all guns are at the same temperature and you often correct one salvo with guns of a different temperature and they go somewhere else. Then you try to correct that salvo with guns from the first salvo that are a different temperature, and so on for the first few salvos ... it takes longer for all the guns to reach their common hot state. If you are firing two or more half salvos as one correction, that wouldn't be a problem.

The other problem is related to radar ranging. Full salvos give you a better MPI determination. With only three or four splashes, one outlier can have a significant effect on where the fire control officer thinks the MPI is and so effects his adjustment. Also, sometimes you lose a splash because it is too close to the target, and that effects MPI determination. I'm aware of one practice shoot where a US battleship didn't get a radar observation of the salvo at all because all three rounds were very close to the target and were lost in it's radar blip. That is why US battleships used full salvoes in combat. Unfortunately they often (not always) used less than that in practice for economy reasons and it affected the quality of the practice.

RN half salvos always use the same guns per salvo so temp changes are consistent.

OTOH, firing half salvos minimizes dispersion because of the greater separation between barrels, and changes in target range rate are easier to follow since there is more feedback with more frequent salvos landing near the target. The target is also kept under a more constant fire

[Byron Angel]

..... If memory serves, my study of the Solomons night fighting turned up data to the effect that FC/FD radar was reputed to be able to spot fall of shot as follows -
5in @ 12,000 yds
6in @ 13,000 yds
8in @ 14,000 yds
14-16in @ 20,000 yds

[Dave Saxton]

One of the more important aspects of being able detect small objects and shell splashes to greater range is the signal noise ratio. Indeed such factors as transmitter power, receiver sensitivity, wave length, antenna height...ect.. pale in importance to signal to noise ratio for detecting shell splashes. An even more important aspect of the signal to ratio issue is how much more signal is required to detect a target than is the level of noise.

When the British captured their first Wuerzburg, they were not that impressed at first. They noted that the power output was only about 8kw and the receiver sensitivity and band width were relatively low. Yet when when they tested it it tracked planes to 38km, and still detected targets even when the signal to noise ratio was virtually 1:1. A couple of years later MIT/Rad LAB scientists sent them their latest findings; reporting that to attain the lowest signal to noise ratio required for detection, the band width of the receiver in megahertz should be the inverse of the pulse width in micro seconds. Sure enough the bandwidth of Wuerzburg was 1/2 megahertz and the pulse width was two micro seconds. Indeed all the German radars followed this rule (the bandwidth of the GEMA sets was adjustable).

Allied designers not knowing this had given many of their designs rather large mismatches of bandwidth to pulse width. Mk-3/4 and Type 284 have large mismatches and require, therefore, large signal to noise ratios to detect such things as shell splashes at longer ranges.

Mk-8 has a very short pulse width and therefore less of a mismatch using a high band width receiver. Nonetheless, the unexpected low detection range, at times, for MK-8 for shell splashes, is not really that puzzling. The transmitted beam was scanned back forth through a 30* sector several times per second. This is what made the desirable Type B or plan position presentation possible. However, this also meant that the target only got painted by relatively few pulses per second. A build up of a strong return signal for each target and therefore a good signal to noise ratio was not attained.

Bell Labs gave Mk-8 a slow scan mode for this very reason. In this mode the beam scanned through the entire 30* sector only once every 60 seconds. This meant that a target would be painted by hundreds or thousands of pulses every second and a build up of a strong echo signal and a strong signal to noise ratio was possible. This allowed Mk-8 to track small targets or detect shell splashes at greater range more consistently. The trade off was that it did not register all targets through the 30* sector at once but only those targets that happened to be about where the beam was at at any given time. This was not good for MPI spotting. The operator, could go to fast scan mode for a wide field of view but this might mean that small targets no longer registered. The operator could also stop the beam on a selected target, but shells falling out side the beam would not be registered.

This was why Type 274 could not spot shell splashes. The transmitted beam was not scanned and the beam was less than 1* wide. Shells would fall outside the beam just about every time.

[Byron Angel]

Thanks for posting that, Dave - sehr interessant.

Byron

[Thorsten Wahl]

performance of MK 8 1944 February
at Truk vs Nowaki
"From experience of fast US battleships"

from Iowa action report
range to target between 35700 and 39000 yards

Iowa fired 40 projectiles within 5 salvoes
the splashes were easily discerned by eye and radar
by the time the third salvo fell, the target began to disappear in the sun glare, haze and distance ... and disappeared from the radar screen

spot observation
salvo one with cold gun correction - no correction smoke and flash were spotted (Comment in the document "wishful thinking because target was not hit")
salvo two up 2 left 3
salvo three up 1 left 2
salvoo four could not be spotted out of visual and radar range for spot
salvo five could not be spotted out of visual and radar range for spot
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
From New Jersey report (wich fired 18 projectiles)
range to target between 32200 and 35000 yard

Th Mark 8 operated in low speed precision scan
there was no difficulty in discerning splashes for radar spotting
Air spot was provided by CAP

[Thorsten Wahl]

When the British captured their first Wuerzburg, they were not that impressed at first. They noted that the power output was only about 8kw and the receiver sensitivity and band width were relatively low. Yet when when they tested it it tracked planes to 38km, and still detected targets even when the signal to noise ratio was virtually 1:1. A couple of years later MIT/Rad LAB scientists sent them their latest findings; reporting that to attain the lowest signal to noise ratio required for detection, the band width of the receiver in megahertz should be the inverse of the pulse width in micro seconds. Sure enough the bandwidth of Wuerzburg was 1/2 megahertz and the pulse width was two micro seconds. Indeed all the German radars followed this rule (the bandwidth of the GEMA sets was adjustable).

Just to add - the german Gema Seetakt radar sets had two signal amplification units wich manipulated the incomming signals.

If the detection unit (Eingangsstufe)receives a reflected signal an oszillating unit fires a synchronised signal wich was added to the received signal so the signal strenght of the reflected return was increased without amplification of the white noise. This signal was then normally amplified and then forwarded to a second signal amplification unit(Mischstufe) amplified again and after a further amplification fed into the display device.