Page 1 of 1

Accuracy of Iowa Top Spot optics.

Posted: Tue Jun 09, 2015 11:42 pm
by Dave Saxton
From tests in 1984, the accuracy of the Iowa's foretop range finder was compared to indicated range on the Mk13 radar. It was within 100 yards deviation from the Mk13 radar range all the way out to 44,000 yards. The accuracy of MK13 was 0.1% of the range with a tolerance of 5 yards. So at 44,000 yards the accuracy of the radar was about 50 yards. This means that the accuracy of the optical rangefinder was about 150 yards at 44,000 yards. Not bad at all. Compare to the German tests of rangefinders posted by Thorsten here:

viewtopic.php?f=13&t=6321&start=15

The German large range finder was 10 meters wide. The Iowa's was 8 meters wide. Yamato's was 15 meters wide.

The same 1984 test also found that the optical range finder could not determine the pattern of the fall of shot of a salvo beyond 26,000 yards, but it could still determine if the MPI was long or short (but not by how much) all the way out to 40,000 + yards. The optics were better at determining if the MPI was left or right of the target than even the MK13 radar. Near misses for deflection would show as hits on the radar. The recommendation even with Mk13 was to use radar for range, and optics for bearing if possible.

Re: Accuracy of Iowa Top Spot optics.

Posted: Thu Jun 11, 2015 8:43 pm
by dunmunro
Do you mean 1944?

What's your source?

Re: Accuracy of Iowa Top Spot optics.

Posted: Fri Jun 12, 2015 2:29 pm
by Dave Saxton
No, I mean 1984. This was from tests conducted during Iowa's 1980s career.

The source is from the end notes of a Yamato vs Iowa scenario analysis by Commander Strafford Morss UNSR. Sometimes the end notes are more interesting than the text:
In 1984 Iowa conducted a comparison of optically determined ranges from the Spot One (Top Spot) 26.5 foot range finder and radar determined ranges from the Mk13 fire control radar located on top of Spot One.... At 44,000 yards the compared ranges differed by 100 yards...
The ability to determine salvo pattern limited to about 26k, but the ability to still determine MPI long or short at extreme ranges, was from a 1985 report by Iowa's Weapons Systems officer.

From these reports the author thought it likely that the Yamato with its significantly larger optics, considerably higher up, could determine range and bearing at 35,000 yards as well as the Iowa could, radar or not, given favorable conditions.

Re: Accuracy of Iowa Top Spot optics.

Posted: Mon Jun 15, 2015 6:09 am
by Bill Jurens
That relative performance of optical vs radar spotting, though the quotes are probably correct (Cdr Morss does not make things up), would certainly be considered atypical for regular day-to-day performance. Even in the 1980s the Iowas habitually used radar spotting for range, optical spotting for deflection. The navy very carefully compared the relative accuracy of optical vs radar, vs aircraft spotting during 1944-45, finding that at least for range, radar showed considerable superiority. It's important to discriminate, however, between establishing range as an absolute value, and spotting the M.P.I. which are really two different things.

Bill Jurens

Re: Accuracy of Iowa Top Spot optics.

Posted: Mon Jun 15, 2015 5:58 pm
by Dave Saxton
Bill Jurens wrote:That relative performance of optical vs radar spotting, though the quotes are probably correct (Cdr Morss does not make things up), would certainly be considered atypical for regular day-to-day performance. Even in the 1980s the Iowas habitually used radar spotting for range, optical spotting for deflection. The navy very carefully compared the relative accuracy of optical vs radar, vs aircraft spotting during 1944-45, finding that at least for range, radar showed considerable superiority. It's important to discriminate, however, between establishing range as an absolute value, and spotting the M.P.I. which are really two different things.

Bill Jurens
I agree. Even as you say the atypical results here show radar to be more accurate. I just found the relatively small differential rather interesting, and unexpected.

Re: Accuracy of Iowa Top Spot optics.

Posted: Mon Jun 15, 2015 6:31 pm
by dunmunro
I have no doubt that an optical RF could achieve excellent accuracy under ideal conditions but conditions were seldom ideal in actual combat. Consider PoW versus Bismarck: In the 1st engagement, PoW could not (although Norfolk did) obtain any optical ranges, while KM optical RF performance is unknown. In the 2nd engagement PoW was able to obtain optical ranges with a 15ft RF out to ~33k yards while Bismarck, according to the Baron (and seemingly verified by Bismarck's poor shooting), obtained none at all. The other factor is the conditions under which test optical RF ranges were obtained; Was the target and/or ranging ship moving at high speed, as is typical during combat, so that optical ranging performance is degraded by ship motion and vibration, or was the test performed under more static conditions? Were atmospheric conditions typical or were they judged abnormally transparent with steady seeing conditions?

Re: Accuracy of Iowa Top Spot optics.

Posted: Mon Jun 15, 2015 8:38 pm
by Dave Saxton
Actually, the Baron wrote that the problem in that particular case was observation (sun glare and smoke specified) of the fall of shot rather than in obtaining ranges.

This is in keeping with a 1985 Iowa gunnery Officer report which describes how conditions can indeed make ranging and spotting difficult. The most difficult condition being the shock from firing the guns. This was described as "ungodly" in the foretop director. And surveying the reports of radar sets being knocked out on Iowas during the late war period, when firing 16" guns, it was more often than not a foretop or foremast located set that was knocked out. Additionally, it was described as "imperative" that the ship not fire when shot was falling around the target, and also that the billows of smoke generated made using the optics much more difficult.

Re: Accuracy of Iowa Top Spot optics.

Posted: Wed Jul 01, 2015 6:08 pm
by Brad Fischer
While optical ranges can indeed rival radar ranges at times, they typically exhibit a large variability from “cut to cut” which is why ranges are plotted and a mean range line is derived from this plot. Using an error budget* derived from fleet experience an Iowa class BB using all 5 range finders might produce an error of:

40Kyds +/- 357yds
35Kyds +/- 275
30Kyds +/- 205
25Kyds +/- 139
20Kyds +/- 102
15Kyds +/- 72

Radar not only has smaller average ranging error but provides ranges continuously; typical time to receive ranges, plot then derive a mean range line is 15-30 seconds depending on the density of ranges. Radar retains a fairly significant advantage in range spotting at extended ranges as well. Wartime results indicate the average error for direct spotting was 50-80yds regardless of range. The average direct spotting error for an experienced RF operator was typically 1.5 Unit of Error (UoE.)

Regarding Yamato’s range finders, it is unknown how many are stereoscopic. Nippon Optical Company had a very strong relationship with Barr & Stroud and almost exclusively produced binocular coincident RF for surface gunfire applications. One thing the IJN did not have was stabilization; this can lead to significantly longer times to produce ranges per unit of time and increases the variable error (which in turn reduces ranging accuracy.)

Finally with regards to shock to radar systems it was a problem but was, relatively speaking, fairly rare. We did a survey of all the action reports and gunnery exercises for Mark 8 equipped battleships during the war and the average failure rate was 4.7% for a single unit and a 0.22% failure rate for a two unit failure. This data was compiled from 300 firing events.

Brad Fischer




*To be succinct the error budget consists of:
RF 1 Unit of error = 58.2 x Range ÷ Baselength x Magnification
Ranging Accuracy = (√(Fixed Error²) + (Variable Error²)) x RF Unit of Error
Transmitted accuracy = √(Ranging Accuracy²) + (Transmission error²)
Avg Transmitted Error = (Transmitted accuracy1 + Transmitted accuracy2….n ÷ n)² ÷√n
(Where “n” is the number of rangefinders)
Total System Error = √(Avg. Transmitted Error²) + (plotting error²)

Re: Accuracy of Iowa Top Spot optics.

Posted: Thu Jul 02, 2015 9:03 am
by Francis Marliere
Brad Fischer wrote:Finally with regards to shock to radar systems it was a problem but was, relatively speaking, fairly rare. We did a survey of all the action reports and gunnery exercises for Mark 8 equipped battleships during the war and the average failure rate was 4.7% for a single unit and a 0.22% failure rate for a two unit failure. This data was compiled from 300 firing events.
Mr Fischer, is the data for Mk8 representative of WWII radars or can we expect the earlier sets to be less reliable ?

Best regards,

Francis Marliere

Re: Accuracy of Iowa Top Spot optics.

Posted: Thu Jul 02, 2015 5:11 pm
by Thorsten Wahl
sample of optical ranging excercise
Image

Re: Accuracy of Iowa Top Spot optics.

Posted: Thu Jul 02, 2015 10:58 pm
by dunmunro
Thorsten Wahl wrote:sample of optical ranging excercise
Can you interpret the names of the x and y axis?

Re: Accuracy of Iowa Top Spot optics.

Posted: Sat Jul 04, 2015 9:28 am
by Thorsten Wahl
x axis time starting with 0 to 30 minutes

y axis
upper graph measured distance
-black true distance measured by Langbasisgerät

-green 10 rangfinder Vormars
-orange 10 m rangfinder aft
-yellow 10 m rangfinder turret A
-blue 10 m rangfinder rangfinder turret B
-dashed line average from the rangefinders

lower graph rate of change of distance
-black true rate of change
-red rate of change from the E Mittler

Re: Accuracy of Iowa Top Spot optics.

Posted: Sun Jul 05, 2015 2:42 am
by Brad Fischer
Francis Marliere wrote:Mr Fischer, is the data for Mk8 representative of WWII radars or can we expect the earlier sets to be less reliable ?

Best regards,

Francis Marliere
That was for Mark 8 radars only, other sets such as the SG-A et. al. are not inclusive and weren't researched. The earlier Mark 3 did suffer from a much higher failure rate but we were unable to obtain reliable data on those units. I know from the various fleet correspondence to BuOrd they felt that the Mark 8 was much more durable than the Mark 3, but how much more I don't have enough data to quantify it.

Brad Fischer

Re: Accuracy of Iowa Top Spot optics.

Posted: Sun Jul 05, 2015 2:46 am
by Brad Fischer
Thorsten Wahl wrote:x axis time starting with 0 to 30 minutes

y axis
upper graph measured distance
-black true distance measured by Langbasisgerät

-green 10 rangfinder Vormars
-orange 10 m rangfinder aft
-yellow 10 m rangfinder turret A
-blue 10 m rangfinder rangfinder turret B
-dashed line average from the rangefinders

lower graph rate of change of distance
-black true rate of change
-red rate of change from the E Mittler
Thank you very much for this data! Very interesting indeed. I did a quick analysis of it and came up with an average error of 127m (for average rangefinder range) which is about 0.72% of range. What's even more interesting is the inherent lag in that range compared to the rate change of the true range. Some of the major changes in rates lagged by a minute plus. Still a very good showing undoubtedly.

Brad Fischer

Re: Accuracy of Iowa Top Spot optics.

Posted: Sun Jul 05, 2015 6:12 pm
by Thorsten Wahl
the typical time lag of about 1 minute for rate change was considered by explicitely anticipation by observation of ships movements

magnification of the 10 m RuM on board Deutschland was 25 the improved 10 m RuM used on board of the Scharnhorst-class and Bismarck-class had max magnification of 50
table of [Mindestfehler] ->[unit of error?] based on resolving power of the eye
Image