Page 4 of 7

Input from B. Claire.

Posted: Sun Feb 27, 2005 12:41 am
by George Elder
BC works for DoD as an analyst and has 3 degrees, one of which is in marine engineering. I explained the problem. In his view, the current's effects on the shell splash vs the ship is insignificant, but the wind could be a factor due to its likily stronger effect on the splash than the Bismarck. Like Foeth, he thought that it would be very diffucult to produce a low error margin using this methodology for reasons that go well beyound wind and currents. Indeed, he found Foeth's error estimates very likily, and opined that it might be even greater.
However, Bill had a great idea that might work. He felt one plausible means was to examine the hull's wave-pattern, as in the typical peaks and valleys the water makes as a ship travels through it. He suggested comparing this hull wave form with that seen in the ship's trials and its known wave profile for various speeds, which are done for all warships. These data may yield some useful results, although it will be difficult to determine the peaks and valleys in the hull wave as revealed by the film. He felt this could be a major problem in the exercise, but this may be an area we want to examine. Moreover, he expected a fairly substantial error margin for this method, but said a multitude of speed assessment means might lead to a more supportable "general" conclusion.
As for other methodologies, he had little to offer at the moment. He found the problem fascinating, and thought some intel tools might exist that could provide an estimate. But with regard to obtaining a resolution of 0.5 degrees, he was not optomistic.
Antonio and I are currently preparing a contact letter for a few Penn State researchers, and we'll see how that goes. But the more expertise we gather, the more likily we are to find a workable idea -- if this is indeed possible. This is lots of fun.



Posted: Sun Feb 27, 2005 10:37 pm
by Randy Stone
...Jose, you are quite right. We should both grow up and it should not happen again.
José M. Rico wrote: To Randy and George: Enough of all this already.
I am certain George will begin to address me in the respectful and civil manner which decency and this Board expects.

And I will do the same.

Randy Stone

Rolling up our sleeves...

Posted: Sun Feb 27, 2005 10:56 pm
by Randy Stone
Antonio, foeth, George and Iankw

...let’s review the record.

We have before us the question of Bismarck’s speed as observed on film and the light this sheds on the issue of the Winklareth Theory.

George and I have been around and around about fluid dynamics and the issue of whether or not the oceanic current is of any consequence in this case. Happily, as George posted, foeth will help us out. And so he has.
foeth wrote: As far as the current is is not important, as all things are relevant to each other... foeth
foeth wrote:I’d start by ignoring both wind speed and currents... foeth
foeth wrote:I think you don’t need fluid dynamics but uncertainty analysis... foeth
foeth wrote:Apparently somebody thought it would be a good idea to calculate it (the water column) using FEM (Finite Elements Modeling), but why ? (All quotes Thursday, 2-25-05/ 9:17am) foeth
Why indeed ? These are all very good points and underscore precisely topics I have previously addressed.

We must recognize that we are not modeling a scenario but are making educated judgements by observing events captured on film. Fluid dynamics is not a viable solution as it does not address the issue at hand. As foeth pointed out, we are actually looking at an analysis of error in defining the characteristics of the film under study.

This is exactly what Antonio and I have done, independently of each other, and we have derived sufficiently similar results as to provide a confident validation to the numbers. The amazing thing about George and I going on about this as though we were in some sort of great disagreement is that we agree, when our respective posts are evaluated, that Bismarck was traveling between 29.5 and 30.0 knots. Therefore, I agree that it’s silly to waste so much time on needless recriminations and I apologize to the Board on behalf of the both of us for not pointing this out earlier.

Insofar as we now have a consensus that the oceanic currents and fluid dynamics play no role in this exercise we can now move to the issue of determining the speed of Bismarck and what effects this has on the Winklareth Theory.

First, allow me to make some comments about the following statement:
foeth wrote:The speed can be given as 30 knots +/- 3.6 knots. ANY speed in this range is CORRECT. Even if (the) estimate in time is perfect and the error in position is only 5m (2%) you’re off by 1.2 knots. (Thursday, 2-25-05/ 9:40am) foeth
I believe there are some practical impediments to assigning a speed of 30 +/- 3.6 knots to Bismarck which compel us to review this estimate. First, the highest speed I have seen anywhere for Bismarck and upon which I can rely is 30.8 knots. Second, the lower end of the stated value – 26.4 knots – is not consistent with other testimony which places Bismarck as steaming at least 28 knots during the Denmarck Strait action. If we consider these facts as reasonable then I would submit foeth’s estimate must contract to 28.0 to 30.8 knots (or an even 31.0 knots for sake of simplicity). Thus practicality would suggest we are presented with a speed of 29.5 +/- 1.5 knots.

On this basis foeth pardon me in assigning to you this estimate of 29.5 +/- 1.5 knots.

Reviewing what George has written,
George Elder wrote:I am content that (Bismarck) must have been traveling at least 29 knots...(but) I am not yet prepared to say she was going at 30 to 31 knots... (Thursday, 2-25-05/ 6:42pm) George
Here, having posted his opinion that Bismarck must have been traveling between 29 and 30 knots, I will assign George his estimate of 29.5 +/- 0.5 knots.

I have previously posted that my own estimate yielding a speed of 30.0 +/- 0.5 knots.

Now, in response to foeth’s calculations which he covered in detail,
foeth wrote:...I introduced a 0.5 s(econd) error. I cannot assume that the speed of the camera runs true or that the number (Antonio’s film counter – Randy’s emphasis) is read accurately enough, the point in time of the watercolumn is correct, etc. The error in distance is about 5%. (Thursday, 2-25-05/ 9:40am) foeth
I would point out that I did assume the camera to be functioning correctly throughout the sequence. This is as much a judgment call as foeth has made and likely as correct or incorrect as his judgment may be. However, the time interval is sufficiently abbreviated as to preclude the introduction of any significant error, in my opinion.

In addition, I eliminated the error in timing the sequence by the simple installion of a digital chronometer in parallel with the laser reader of the DVD player. With shuttle operation this could theoretically read to 0.001 seconds but for practical purposes it would obviously be far more reliable to settle for a reading in the region of 0.050 seconds, a neglible factor for our calculations here. The same measurement applies to the water column which is, after all, nothing more than a positional reference...a football goal post, as it were.

As for the error in distance, as it happened I had chosen 5 meters, which places us at 2%, which is, of course, exactly what foeth had pointed out. Thus I arrived at 29.45 to 30.45 knots. Since this number reminds one of a statistical statement which defines a family as 3.7 people (whoever heard of 7/10ths of a person ?) I simply reduced the value to a more digestable 29.5 to 30.5 knots, or 30 +/- 0.5 knots.

Thus between Antonio, foeth, George and myself, we have arrived at a speed band of 28 to 31 knots for Bismarck and a speed band of 29.5 to 30.0 knots around which all four of us have expressed agreement. Antonio has posted his calculations. We would note that foeth has posted a value for Bismarck which he may or may not agree I have ‘corrected’ properly given the circumstances. George has expressed his agreement and a more important statement. Let’s revisit his posting, as follows:
George Elder wrote:... With regard to speed estimates...I am content that the ship must have been traveling at least 29 knots given all we know from the record. I am not yet prepared to say she was going at 30 to 31 knots, and especialy 31 knots...As I have admitted, I am not at all sure how one can accurately judge the Bismarck's speed given what we now have available. I think we should search out a method that will have an acceptable error margin, say +/- 0.5 knots at most. That is a demanding standard, and perhaps it cannot be obtained. (Thursday, 2-25-05/ 6:42pm) George
It is an interesting to read George’s comment that an error of “...+/- 0.5 a demanding standard...” when, in fact, he wrote immediately prior that, in his opinion, Bismarck was traveling “ least 29 knots...” and was not “...prepared to say she was going (more than) 30 knots...” In other words, he supports the contention that Bismarck was steaming at 29 to 30 knots or – put another way – 29 +/-0.5 knots, the very demanding standard he deems necessary for the discussion !

Looking back on all that has been written, I can not understand what George and I are even arguing about since we are in such agreement as we are. And although we have established estimates to within +/- 0.5 knots, it is also quite unnecessary for the purposes of our problem as I can readily demonstrate. All of which leads us back to the point where we can discuss the effects of these calculations with regard to the Winklareth Theory, which is the object of this exercise.

If we agree, as obviously we do, that Bismarck was traveling between 28 and 30 knots, it can be demonstrated that there is no need to worry about whether we can define Bismarck’s speed to within +/- 0.5 knot. Although I believe we have, and have so demonstrated, the distinction is quite meaningless.

It is meaningless because even if we assume that Bismarck began the action only 2000 yards astern of Prinz Eugen (my own estimate which many believe is artificially low) and that she was traveling at our agreed upper limit of 30 knots there is simply no physical way Bismarck could have overhauled Prinz Eugen in the timeframe defined by the Winklareth Theory. On the other hand, if Bismarck was further astern of Prinz Eugen (very likely from the evidence I am seeing) and was steaming at our agreed lower limit of 28 knots, then the case supporting theWinklareth Theory is further undermined and, actually, is quite hopeless.

In conclusion we have established:

That, for the purposes of our exercise, the effects of current and wind are negligible and the application of fluid dynamics modeling has no value in determining Bismarck’s speed,

That the methods employed by Antonio, foeth, George and myself in determining Bismarck’s speed and the variation of our respective determinations is acceptable, valid and verifiable for our exercise,

That the sea speed of Bismarck in our collective determination is 29 +/- 1 knots, a speed envelope fully sufficient to address our question, and

That considering this speed envelope as determined for Bismarck, in no case will Bismarck overhaul Prinz Eugen as postulated in the Winklareth theory.


Posted: Sun Feb 27, 2005 11:11 pm
by iankw
In a nutshell!!

Posted: Sun Feb 27, 2005 11:22 pm
by foeth
In a nutshell? Didn't you see Austin Powers?

Just to point out:

1) I only gave comment on the measurement technique. This technique gives an estimate and an uncertainty. Doesn't matter that B can't go faster than 30,8 kts; if you'd found say 31,5 kts as an estimate you'd know you'd be an the high side. I used Antonio's numbers without thinking. I only showed with a small calculation that such an uncertainty is important to take into consideration without caring much for the result.

2) The +/- 1,5 kts seems plausible. The error margin is not as small as 0,5 kts, but it's not over 5 kts, for example.

With shuttle operation this could theoretically read to 0.001 seconds but for practical purposes it would obviously be far more reliable to settle for a reading in the region of 0.050 seconds, a neglible factor for our calculations here.
Of course, I meant the original film!

If Antonio's and Randy's numbers are reasonaly correct, Bismarck should be sailing near top speed. This top speed may be a little lower than the 30,8 kts (well, this is where wind may matter a bit :) ).


Posted: Mon Feb 28, 2005 12:56 am
by George Elder
Nearly all the experts thus far consulted note that both current effects and wind effects do not each cancell out with regard to the their relative influences on the Bismarck and the shell splash's locations. Some experts opine that the shell splash may be considerably more impacted by wind effects than is the Bismarck, although current effects are deemed to be small. Measureing the effects of wind on the position of the shell splash location relative to the Bismarck does not seem feasible in that we know so little about the water column's mass, etc..
Thus far, most of the experts deem that the present shell splash methodolgy will produce a considerable inherant error margin for reasons not related to the current or wind effects, and do so as Foeth has indicated. A further caveat has recently been profered regarding the Bismarck's axial orientation relative to the camera -- perhaps more on this later. I have also been advised that course machinations are likily to reduce speed.
One possible additional means of speed determination has been proferred, that being the hull wave form. Letters are going out this week to find other possible means. It is hoped that finding two or three means may result in a more likily and supportable conclusion than will a single method -- although each method must be subjected to error anaylisis.
At present, it seems that we cannot reduce the error margin to 0.5 knots using the existing means, despite claims to the contrary, or perhaps even 1.0 knots. Most deem this very unlikily, but we will see what evolves from this week's contacts and from future research. There is thus no conclusion, and especialy one from me that deems the speed is 29.5 knots. The matter is now being looked at by experts, and we have not received enough input to make anything like a precise velocity conclusion.
Lastly, I consider the shell splash method the exclusive brain child of Antonio. He discovered and developed it. He quantified it, and credit is due to he alone for its use. It may turn out that this method, when combined with others, leads to a more refined velocity estimate. This seems to sum things up at present. Much left to do.


Speed estimates

Posted: Mon Feb 28, 2005 2:28 am
by Bill Jurens
This is a re-write and slight expansion of a previous post, sent about four hours ago, which does not seem to have gotten through.

Although the methods suggested to determine Bismarck's speed in this thread show some ingenuity, we must remain very careful not to read more information in to the photo than is really there.

In particular, one must keep in mind that the timing on the sequence as shown on television is likely to be fairly considerably in error. "Television time" is not the same thing as "film time", and the conversion of one to the other usually incorporates some error, usually of no consequence, but of potentially considerable consequence here as timing is being used as a basis for measurement.

In addition, we must also keep in mind that the camera ship is ALSO moving during the sequence in question, which in and of itself will almost certainly lead to misleading parallax effects. Further, insofar as the exact position of the splash in range is problematical with reference to Bismarck's projected track line, this has the potential to lead to considerable errors as well.

It's well to note that most of these problems would disappear, or at least be somewhat alleviated, should the images be measured from the original negatives, or a contact print therefrom. Attempting to derive precise measurements from a television image, while better than nothing, is, in reality not MUCH better than nothing.

For what it is worth, I would estimate the error band for speed derived using this sort of technique as +/-10% for 90% probability. Certainly measurements to a half-knot or even a full knot are likely to be the result of over-analysing artifacts hidden in the noise. In that regard, clever (and correct) mathematics, no matter how well applied cannot be expected to yield a result which is better than the imagery itself. And the imagery is generally of very poor quality indeed...

The idea that Bismarck's speed can be derived from the wave profile around the hull is interesting, but also highly problematical. In the testing tank, this works just fine, given some knowlege about hull geometry etc., but at sea, where wave effects per se are superimposed upon wave effects due to motion itself, the resulting profile is unlikely to be capable of delivering more than a very broad approximation of speed. In addition, the resolution of the images is not, in my opinion, likely to permit anything other than a 'guestimate' of the profile to begin with.

I hope this helps...

Bill Jurens

Excellent points.

Posted: Mon Feb 28, 2005 4:38 am
by George Elder
Hi Bill:

Your views coincide with those of the other experts we have consulted, and especialy with regard to a new area that Bill Claire brought up. You note:

"In addition, we must also keep in mind that the camera ship is ALSO moving during the sequence in question, which in and of itself will almost certainly lead to misleading parallax effects. Further, insofar as the exact position of the splash in range is problematical with reference to Bismarck's projected track line, this has the potential to lead to considerable errors as well."

Claire was grasping for methods with the waveform profile because I was pressing him on the matter. He said it would be very difficult to even see the complete profile, and I am not at all sure where the waveform drawings reside. But we will complete the exercise.
Perhaps the cloak & dagger folks have a few tricks, but without the origional material -- we may be engaging in exercises that will not get us very far. Still, the quest is fun, and some of us are learning a great deal in the interum about an area we have never studied in any depth.


Posted: Mon Feb 28, 2005 1:56 pm
by Ulrich Rudofsky
I have been following this discussion for some time. I think that Foeth and Jurens have finally brought up two very important points: The mathematical rules for significant numbers and the determination of statisical significance.

I'm enjoying all...

Posted: Mon Feb 28, 2005 4:51 pm
by Randy Stone
...of what has been written.

As I posted, the effort is rather meaningless when one takes either 28 or 30-31 knots and applies them to the distance to be covered by Bismarck in overhauling Prinz Eugen.

In either event, the deed can not be accomplished per the Winklareth Theory.

So an analysis, however flawed and unnecessary, can not produce the evidence to support the Theory, and

the practicalities of steaming at either 28 or 30-31 knots can not produce the result either.

Thus the conclusions stand and the Theory falls.

Not much question of it and little left to do on this front but to freshen up the bows and ribbon.

I do agree with foeth's statement that Bismarck would be steaming at near top (optimal) speed for simple reasons I'll detail later but, in any event, backing into the question yields the same result: namely, that the distance between Bismarck and Prinz Eugen could not be traversed in the time postulated by the Winklareth Theory no matter what.


Posted: Mon Feb 28, 2005 5:04 pm
by foeth
Indeed, and I hope my little exercise indicates that

a) With good data the error is already large
b) With bad data it's more

Calculation of error margins is something only done by experimentalists I suppose and gives you an indication of your analysis. Even with numbers hitting the mark, you have to be fair and admit the estimate can be "luck".

Note that without getting much into the actual value and implication of my "answer", the error margin is large. For those who want to interpret this number, 30 kts +/- 3,6 does NOT mean that 30 kts is the average. It means that 26,4 kts is ALSO the average and then there's a 5% chance that it's worse.

The wave-pattern method is even MORE inaccurate, for several reasons

a) The calculated wave pattern can be calculated fairly if not very well for a known configuration. Bismarck is not yet hit in the bow, but still, what's her average draught? But unfortunately she's also moving, so we have the incident wave, the refracted wave (by ship motion), the diffracted wave (wave changed by presence of the hull) and her own wave system (which is a function of time due to the ship motions).

b) Getting a location on a shell splash is difficult. Now correlate a wave pattern which is even less clear? Note that key waves in the system, such as the bow wave, may be higher or lower at different speeds, the actual shape of the wave pattern will be fairly constant at the speed range of interest. As a matter of fact, the angle of the bow wave with the ship will not change over her entire speed range. With the given picture quality I think this is not an effective approach.

The error will certainly be larger than with the shell splashes. I think Bill J. is correct in assuming that the base material is not of sufficient quality.

But lets see what other people come up with.

Posted: Mon Feb 28, 2005 6:16 pm
by José M. Rico
There is one thing that it seems to me you guys are forgetting. The Bismarck has already been hit, and her top speed reduced to 28 knots (according to KTB), in the time frame picture of the film you are analysing.


Which brings us...

Posted: Mon Feb 28, 2005 7:17 pm
by Randy Stone
...right back to the approach which asks:

1.) Whether Bismarck could overhaul Prinz Eugen at the highest possible speed (31 knots) from the shortest distance (2000 yards), or

2.) Whether Bismarck could overhaul Prinz Eugen at the slowest speed (28 knots) from the longest distance (~3000 meters).

The second question is quite superfluous really, I know.

The question is really quite simple even if no one is satisfied with the present estimates; the fact that Bismarck can be reasonably assigned speed values of 28 to 31 knots settles the question as to whether the Winklareth Theory has any legs on this score.

It doesn't and that pretty much sums up the issue.

So any discomfort over calculations, which nevertheless arrive at a range of 28 to 31 knots regardless of errors, are misplaced.

As for foeth's assertion that Bismarck could be running in excess of 30 knots, I can accept that. Here we have a vessel with little bottom fouling, able to easily fight the sea, said sea being cold, with a powerplant in optimal condition and tactically forced to flee. There is every incentive to run at the highest possible speed.

If Bismarck was traveling at better than 30 knots (at least at some point in time), I would not be surprised; of course this would further erode the Winklareth Theory but Condition 1 above is satisfactory enough on this count.

I am not forgetting Jose, that Bismarck was already hit forward and increasing flooding was taking it's toll. After all, she was raising a propeller out of the water at one point. This flooding can only hurt the case for the Winklareth Theory; but I am imposing ideal conditions (namely the highest possible speed) -- any subtraction from which will only further undermine the Winklareth Theory. Condition 2 above allows for that probability.

This is why I have concluded an exact measurement of Bismarck's speed is meaningless in this context because it is unnecessary. Bismarck could not overhaul Prinz Eugen at any speed of which she was capable, undamaged or not, given the initial distance between the two vessels.

Glad to see you posted the picture in its correct orientation :wink: .

As an aside, I first read "The Bismarck Chase" when first released in North America (I can check my invoices for the precise date) and it impressed me as highly speculative and improbable then. These discussions serve a useful purpose in confirming these initial conclusions.


Speed and photos

Posted: Mon Feb 28, 2005 11:29 pm
by Antonio Bonomi
Ciao all,

just few addittional things from my side :

to Randy :

1) No matter which speed you choose for Bismarck the initial first salvo photo of Bismarck that shows her between 1800 and 2500 meters back on Prinz Eugen wake ( we will see later the correct estimation ) clearly makes impossible Mr. Winklareth theory as said.
This is another evidence that goes on top of several others, I will suggest you not to loose anymore time with that, it is all over since a long time already.

to Jose :

1) the photo we are evaluating is not that one ( which is Nh 69728 ) but the one were last PoW shells falls ahead of Bismarck bow ( I sent you a copy so you can post it ).

2) YES, surely Bismarck was hit already by PoW but to me that was not such an immediate effect on her running asset and speed and assuming the hit on the bow was taken at 05.58 it was only 7 minutes before.
On the photos till 06.20 we still see Bismarck with the bow very high on the water and with big waves on her bow that shows full speed ( those can be all used to make Bismarck speed determination if one wants to play with that too).
Later at 10.00 am when Bismarck will be looked at by Prinz Eugen for oil leaking and 2 photos were taken than we clearly see the bow low on the water and the stern high and abnormally out as we know it happened ( Nh 69732 ).

3) As Nh 69722 ( Bismarck first salvo ) shows Bismarck lets say 2000 meters back on Prinz Eugen, than the second photo available for the battle is Nh 69728 ( the one you showed above ).
It is very clear that Bismarck did gain a big amount of distance ( evaluated around 1000 meters ) as Nh 69728 shows Bismarck at around 1000 meters on Prinz Eugen starboard side still back to the heavy cruiser of course.
Prinz Eugen sailed straight at 27 knots that is for sure, and Bismarck can be there only if she sailed 30 knots and for 10 minutes she gained around 1000 meters which is exactly justified by that speed difference.

You will agree that it is impossible for Bismarck to be aside of Prinz Eugen or ahead surpassing on port side, that is a pure invention and demonstrated been not possible, it is an elementary consequence.

If Bismarck was sailing only 27 knots or a bit more it was going to be impossible for her to gain such a distance from Prinz Eugen ( around 1000 meters ) within only 10 minutes, this is obvious and elementary too.

The call of Bismarck still capable to sail 28 knots ( which at this point seems a ''declaration of limitation in speed '' compared to previous performances, ... at least in my opinion :D ) recorded into the KTB was made after the battle end during some communications between Bismarck and Prinz Eugen talking damages and shells used, after 06.30 for sure.

to all :

I agree that errors and tolerances can be ( and are ) there, but I have simply made an elementary calculation of speed looking at a fixed image that showed with good approssimation an advance of Bismarck bow versus a fixed object ( the water column in front of her that was not moving in the air nor on the sea ).

The Bismarck image was stopped on my TV and was as said failrly big ( 15 cm ) and accurate, than the counter was the film from Bundesarchiv and not my own clock, and I took either and intermediate measure ( at 1/3 distance gain ) as well as the final one.

Both drove me to evaluate Bismarck speed a bit more than 30 knots, we have the logic above that shows a huge delta between PG at 27 knots and Bismarck that to be there must have run at least at 30 knots, ... than everything seems logic to me currently.

If anybody do have better ( and please only possible ones ) explanations to the evidences as they currently appear please let me know.

Of course this is what we have currently, if better things will become available than we will see in the future.

Ciao Antonio :D

Re: Speed...

Posted: Tue Mar 01, 2005 12:45 am
by Randy Stone
Antonio Bonomi wrote: Randy :

...No matter which speed you choose for Bismarck the initial first salvo photo of Bismarck that shows her between 1800 and 2500 meters back on Prinz Eugen wake (we will see later the correct estimation) clearly makes impossible Mr. Winklareth theory as said...I will suggest you not to lose anymore time with that...
Precisely. "...No matter which speed you choose for Bismarck...," that's the 800 pound gorilla in the room and there's no getting around him.

Calculate all we want, debate tolerance bands all we choose, the bottom line is that for any reasonable speed one arbitrarily assigns Bismarck, she just can't make up the distance between herself and the Prinz.

I'm not losing any time with it...just curious why that gorilla is being ignored when this question has been solved and there are more pressing issues to the destroyer attacks on Bismarck's last night :D .