"5th Battle Sqdn turn away at Jutland" Topic

Ran across some RN shiphandling data in my files that might be of interest. The information dates back to the 20s (IIUC) and relates to the turning performance of the HMS Queen Elizabeth, which in turn might make it relevant to the conduct of Evan-Thomas's turn-away of 5th Battle Squadron at Jutland.

HMS QUEEN ELIZABETH
Entry Speed – 23kts
Rudder Angle – 35deg

Retained Speed & Elapsed Time after turning through:

45deg – – – 90deg – – – 135deg – – – 180deg
19.5kts – – 15.6kts – – 11.4kts – – -9.7kts
37s – – – – 56s – – – – 1m 19s – – – 1m 48s

If a correct representation, a look at the retained speed after 180deg explains how 3rd BS of the HSF, making 22.5kts was able to close the range upon the nominally faster 5BS.

Another graph of ship acceleration vs time curves suggests that a Queen Elizabeth class ship, steaming straight ahead rather than turning, would require ~9 minutes to accelerate from 10kts to 20kts and further 16 minutes to get from 20kts to 23kts.

B

Plotting the data for speed and time gives curves that are nearly linear:

PDF link

I expected the speed loss per degree to be greatest up to about 90 degrees and then less, based on previous readings. For example:

"Effect of a turn on speed
The effect of the drag of the rudder and the sideways drift of the ship will result in a progressive loss of speed while turning, even though the engine revolutions are maintained at a constant figure. For alterations of course of up to 20 degrees the reduction of speed may not be very great, but for those between 20 degrees and 90 degrees the speed usually falls off rapidly. For alterations exceeding 90 degrees the speed may continue to fall slightly, but it usually remains more or less steady."

Quoted from:

link

"In the course of turning, especially with a large drift angle, the increased hull resistance causes the ship to slow down, sometimes involving a reduction of 40 percent or more of the speed with which it approached the turn. After the average ship has turned at least 90°, conditions become steady and its centre of gravity moves at uniform speed in a circular path."

Quoted from:

link

Do modern ship designs behave so differently from WW1?

Thanks,
NCC1717

Hi NCC1717,
Plotting the data for speed and time gives curves that are nearly linear:
PDF link

I expected the speed loss per degree to be greatest up to about 90 degrees and then less, based on previous readings. For example:

"Effect of a turn on speed
The effect of the drag of the rudder and the sideways drift of the ship will result in a progressive loss of speed while turning, even though the engine revolutions are maintained at a constant figure. For alterations of course of up to 20 degrees the reduction of speed may not be very great, but for those between 20 degrees and 90 degrees the speed usually falls off rapidly. For alterations exceeding 90 degrees the speed may continue to fall slightly, but it usually remains more or less steady."

My use of the term "elapsed time" in the original previous post ws probably misleading. Also, "Time to Turn" data came from one chart, while "Speed Reduction versus Turning" from a second (related) chart; it is possible that the two data sets are not perfectly coincident. The data sets also represent overall averaged values for the class of five ships as a whole. Nevertheless, I think that the data are reasonably valid

Queen Elizabeth enters her turn at 23 kts and applies 35 degrees of rudder helm.
The breakdown of the turn by 45deg increments should look approximately like this -

Increment 1 (0 to 45deg) –
This first 45deg of the turn takes place over a period of ~37 seconds (1.2deg per sec). QE's approximate average speed is (23 + 19.5)/2 = 21.25kts. This increment is fastest because it takes a certain amount of time (10 sec?) for the rudder to be fully put over; hence less rudder drag and drift initially.

Increment 2 (45 to 90 deg) –
This second 45deg of the turn takes place over a period of ~56 seconds (0.80deg per sec). QE's approximate average speed over this arc is (19.5 + 15.6)/2 = 17.55 kts. These first two increments, taken together (0 – 90 deg), give an average overall turn rate of 1 degree per second, which is consistent with tactical expectations of the time; reduction of speed by 1/3rd after 90 degrees is also consistent with tactical expectations.

Increment 3 (90 to 135 deg) –
This third 45deg of the turn takes place over a period of ~79 seconds (0.57deg per sec). QE's approximate average speed over this arc is (15.6 + 11.4)/2 = 13.5 kts. Rudder drag + drift about the rotational axis have probably reached a maximum; rate of turn hereafter should begin to stabilize.

Increment 4 – (135 to 180 deg) –
This fourth 45deg of the turn takes place over a period of ~108 seconds (0.42 deg per sec). QE's approximate average speed is (13.5 + 9.7)/2 = 11.6 kts. This is about half the original entry speed, again (IMO) consistent with practice.

I have Time to Turn data (but not speed reduction data) for continuation of the turn from 180 to a full 360 degrees. Time required to turn from 180 through 360 degrees is 231 sec (0.20 deg per sec

Graphically speaking, an overhead plot of the turning track will indicate a spiraling movement until drag from rudder and drift will cause the ship to decelerate until the point of power equilibrium is achieved. At this point, the ship will commence to describe a regular circle (assuming a smooth sea and no wind).

- – -

Quoted from:
link

"In the course of turning, especially with a large drift angle, the increased hull resistance causes the ship to slow down, sometimes involving a reduction of 40 percent or more of the speed with which it approached the turn. After the average ship has turned at least 90°, conditions become steady and its centre of gravity moves at uniform speed in a circular path."

I think the above is agreeably consistent with this prediction.

- – -

Quoted from:
link

Do modern ship designs behave so differently from WW1?

Not being a naval architect, I hesitate to formally orate on that particular topic. But, speaking as a naval buff, I will say that, based on my reading of older period texts on naval design and architecture, turning performance appears to have varied quite dramatically depending upon very subtle features. Rudder types (numerous) were extremely influential; it is surprising how quickly rudders evolved after the introduction of engine-powered ships in the mid-19th century. This extended up to WW1 (See Gary Staff and Siegfried Breyer on the peculiar tandem rudder arrangements of the German WW1 battlecruisers). Design of the hull after body, which influences flow of water to the rudder(s) was also important. I recommend taking a look at the book "Theoretical Naval Architecture" by Attwood and Pengelly (published in many editions from 1899 to 1935); they do a good job of illustrating diagrammatically the evolution of ship maneuverability.

Blutarski,

Thanks for the response. When I asked about ship designs behaving so differently in WW1, I was speaking only about speed loss as a function of turn angle. For my miniature games, I'm not modeling the differences in helm response, turn radius or continuous turn steady state speed. I only have the data for a few of the 1900-1920 ships (thanks to Staff I have some good info on German battlecruisers) and these differences would rarely be important for the game.

What I do model is the loss of speed during maneuvers. This makes it important for players to avoid using maximum speeds for formations and to limit radical maneuvers. If they don't 'leave something in hand', as Jellicoe put it, straggling will occur. The speed loss model could also be important for tactical retreats like 5BS, or the HSF battle turns.

I'm currently using a model where there is no loss of speed for turns of up to 20 degrees, a linear 65% loss of speed between 20 and 90 degrees, and no loss beyond 90 degrees. Speeds in the code are integer values, so a more 'refined' algorithm would not matter much.

The theory in the links above seem to be based on constant engine revolutions. In my code, ships not already at full speed may recover some of the speed lost in a turn.

Hi NCC1717,
Your approach is entirely sensible. To venture into any greater level of detail would IMO be utterly needless folly.

I would concur that the QE turn example, which dates from the early/mid 20s, was likely based upon 23kts being about the fastest speed that the QE class could reliably achieve at that time in their career (before the big reconstructions).

B

Blutarski,

I have just received a copy of O.U. 5274, "Remarks on Handling Ships (1934)". The data you listed for turning performance of the QE is exactly as presented in Tables 1 and 2 of O.U. 5274. The tables include data for other ships as well.

Notes below both tables state:
"The above information has been obtained from various sources and is intended as a guide only. Data for individual ships of the same class will vary. "

From these notes and the names of the other table entries it seems that the data refers to the QE class, rather than to the QE herself.

O.U. 5274 also has a statement about loss of speed in a turn which is somewhat different from the one I quoted in a post above:
"The actual loss of speed is greatest between 20deg and 90deg, but the speed continues to fall until the ship has turned through about 180deg." I may modify my algorithm slightly as a result.

Related to another thread, O.U. 5274 has several sections on station keeping. One states:
"The distance between ships is the distance measured from stem to stem and allowance must be made for the position of the observer in his own ship, and the mark used in the other ship." These sections have a lot of information about measuring distances and keeping station in manuvers.

The station keeping sections are in a chapter titled "The Handling of Ships in a Fleet". This chapter says that the first edition was issued by Hornby for the Mediterranean fleet, 1877-1880, and that ‘his principles remain fundamentally true today'.

Hi NCC1717,
It is amazing to me that what appears to be a "standard" interval between ships was (has been?) maintained at 500 yds "stem to stem" over such a lengthy period of time that has seen capital ships triple in displacement, double in speed and nearly double in length.

I cannot help but wonder whether there were exceptions based upon special circumstances. For example, Hipper was said to have handled 1AG in "open order" when in action at Jutland.

B

Blutarski,

Organizing what little data I have gives this comparison of formation distance and length:

PDF link

This is a case where I would like to have period signal books. My impression is that the formation distance could be anything the OTC wanted, but that some default distance was used if nothing was signaled. Even the default distance could have been situational, e.g., cruising vs combat.

Here is the same data presented in a different format, and as a ratio of ship length to formation spacing:

PDF link

Thanks NCC1717,
The second format was well done – much easier to grasp (for me at least). I was particularly taken by the large interval used in connection with the HSF cruising formation.

Been busy lately studying some French tests and analyses of tacking of square-rigged sailing warships under different wind and sea conditions. Very interesting stuff.

B

I know I'm late to this party, but a word of thanks gentlemen for an erudite discussion which enhances all our knowledge of ship handling.

It is brilliant to see this sort of thing on TMP.

Nick

Hi Boz,
Very kind. I'm sure NCC1717 gives a Star Fleet salute of thanks as well.

B

Hi Boz,
Re my 23 Sep 21 1:34pm post above, I just realized that my comment –

"Very kind. I'm sure NCC1717 gives a Star Fleet salute of thanks as well."

- might have been misinterpreted.

My intention was not to meant to imply that NCC1717 would also be thanking me, but that he would also be thanking YOU for your kind acknowledgement. His contribution to the discussion was, after all, fully equal to mine.

FWIW.

B