SUBSIM Radio Room Forums
Glue, paint, razors...what could go wrong?


SUBSIM: The Web's #1 resource for all submarine & naval simulations since 1997

Go Back   SUBSIM Radio Room Forums > Wolfpack: the new multiplayer co-op U-boat simulation > Wolfpack
Forget password? Reset here

Reply
 
Thread Tools Display Modes
Old 05-06-2021, 07:25 AM   #1
Efshapo
Watch
 
Join Date: May 2021
Location: France
Posts: 29
Downloads: 3
Uploads: 0
Default Lateral Speed Method

Hi everyone,

This is my first post on this forum. I wanted to share something with you all.

In February, I discovered a new method to acquire data needed to shoot torpedoes. I only tested it on Wolfpack -that's why I post in this section- but it should work on any sub game, and even in real subs.

I searched a bit and so far, I haven't found this technique anywhere else. Maybe some of you could show me sources of what I think I merely rediscovered.

I will try to explain it in a few words:

Usually, one need at least AOB and target speed to shoot. Range is also needed if gyro angle is high.

Once the ship has been identified, range is quite easy to estimate. Speed can be trickier if you're measuring it from an AOB that is far away from 90°. And AOB must be the hardest one since you have to either plot target course, measure width/height ratio, or maneuver to a position where target features are visually aligned (kingpost at AOB 90°, some masts at AOB 45°, target track at AOB 0°...)

With my method, I only use range and lateral speed (perpendicular to the bearing). The point is that you can acquire these data from any AOB. Here is a step-by-step procedure:
- Measure range from target with your preferred method.
- Choose a vertical feature on the target (usually a mast) and measure the time it takes for this feature to cross a certain amount of horizontal graduations (centiradians) on the periscope at 6x magnification, I often use 8.
- Use the following formula to compute lateral speed : lateral_speed (kts) = 0.486 * range (hm) * nb_centirad / time (s)
- Enter 90° for AOB input in TDC (starboard or portside, depending on the direction of traveling of the target).
- Enter lateral speed as speed input in TDC.

Here is a video to illustrate (jump to 3'20s):


This previous method only works for one target. I refined it to work on a whole convoy:
- Choose one representative target among the convoy.
- Measure its lateral speed as seen above.
- Visually estimate its AOB.
- Use this formula to compute target speed estimation: speed = lateral_speed / sin(AOB)
- Enter estimated AOB in TDC.
- Enter estimated target speed as speed input in TDC.

There also is a simple trick to get a good estimation of AOB, using this same refined method:
- Choose one target among the convoy.
- Measure its lateral speed as seen above.
- Also measure its real speed with your preferred method.
- Use the following formula to compute AOB : AOB = asin(lateral_speed / real_speed)
- Enter this AOB input in TDC.
- Enter real target speed as speed input in TDC.

Here is a video to illustrate (jump to 3'30s):


In all these three procedures, the main data from which other are derived is lateral speed, which is actually the one that needs to be accurate for the TDC to yield a good gyro angle.

What do you guys think? Have you seen something similar used elsewhere?

-Efshapo

Last edited by Efshapo; 05-06-2021 at 07:59 AM.
Efshapo is offline   Reply With Quote
Old 05-06-2021, 08:10 AM   #2
Aktungbby
Gefallen Engel U-666
 
Aktungbby's Avatar
 
Join Date: Jul 2013
Location: where I'm at presently
Posts: 23,972
Downloads: 21
Uploads: 0


Default welcome aboard!

Efshapo!
__________________
"Only two things are infinite; The Universe and human stupidity; And I'm not too sure about the Universe"
Aktungbby is online   Reply With Quote
Old 05-06-2021, 08:17 AM   #3
derstosstrupp
Crush Depth Teammember
 
derstosstrupp's Avatar
 
Join Date: Aug 2008
Posts: 1,150
Downloads: 480
Uploads: 0


Default

It’s basically Auswanderungsverfahren, which is determining speed by bearing change. It’s nothing new, I’ve done several videos on it, German surface ships did it too (it’s outlined in MDv 304 Torpedo-Schießvorschrift). It’s not exactly like this, but the principle is the same. You can calculate target speed using an estimate of target AOB and an estimate of range, and then a measure of bearing change. The only difference here is that you are using centirad marks versus bearing change which is fine, and you are also setting AOB to 90 and inputting the lateral speed across the LOS into the TDC, which is also fine. Mathematically it’s the same.

Not a primary U-boat method, method of choice in history was simply matching course and speed (“Ausdampfen”), or they plotted (“Koppeln”), and to the extent they couldn’t do that, doing the fixed wire method based on an estimate of target length (“Durchwandernlassen”), or simply estimating speed by bow wake.
derstosstrupp is online   Reply With Quote
Old 05-06-2021, 08:28 AM   #4
Efshapo
Watch
 
Join Date: May 2021
Location: France
Posts: 29
Downloads: 3
Uploads: 0
Default

There are some similarities with the constant bearing method since my technique exploits the same mathematical constant which is lateral_speed = speed * sin(AOB), the gyro angle being directly related to lateral_speed.

In my opinion, the real new trick here is the use of the horizontal centiradians scale which is really faster than maneuvering the sub to get a constant bearing. This is what I haven't seen anywhere else, and would like to know if it had been documented before.

I understand one could do it using direct bearing change but I wonder if bearing measurement is as accurate as the horizontal scale...
Efshapo is offline   Reply With Quote
Old 05-06-2021, 08:34 AM   #5
derstosstrupp
Crush Depth Teammember
 
derstosstrupp's Avatar
 
Join Date: Aug 2008
Posts: 1,150
Downloads: 480
Uploads: 0


Default

Quote:
Originally Posted by Efshapo View Post
Yes, there are some similarities with the constant bearing method since my technique exploits the same mathematical constant which is lateral_speed = speed * sin(AOB), the gyro angle being directly related to lateral_speed.

In my opinion, the real new trick here is the use of the horizontal centiradians scale which is really faster than maneuvering the sub to get a constant bearing. This is what I haven't seen anywhere else, and would like to know if it had been documented before.
I’m not referring to the constant bearing method though. Auswanderungsverfahren, as I said, deals with bearing change. Like I said, the only difference between your method and the traditional method, is that you are measuring the bearing change In centiradians and not in degrees of bearing.

I’ve never seen it anywhere else done with centiradians, that part is new, but the overall method is as old as the 1920s when this was developed in between the wars.

Here’s an old video of me doing it with degrees:
derstosstrupp is online   Reply With Quote
Old 05-06-2021, 08:43 AM   #6
Efshapo
Watch
 
Join Date: May 2021
Location: France
Posts: 29
Downloads: 3
Uploads: 0
Default

From TVRE ( http://www.tvre.org/en/acquiring-torpedo-firing-data ):

Quote:
The main disadvantage of the Ausdampfverfahren method was the relatively long time needed for suppressing changes of target bearing. That's why, on the basis of this method, another procedure was developed, the so called Auswanderungsverfahren method. This procedure was less accurate than Ausdampfverfahren, but required taking only two target bearings 1 minute apart and measuring (or estimating) the distance to the target during the taking of the first bearing.
Yes, thank you! That's indeed really close to what I'm doing.

Again, I guess bearing reading is probably not as accurate as horizontal scale reading but I'm glad someone gives me a document (or rather the name of a procedure) closely related to what I proposed!

We already had the 4- and 3-bearings method, I guess we could call this one the 2-bearings method.

Thanks again.
Efshapo is offline   Reply With Quote
Old 05-06-2021, 08:45 AM   #7
derstosstrupp
Crush Depth Teammember
 
derstosstrupp's Avatar
 
Join Date: Aug 2008
Posts: 1,150
Downloads: 480
Uploads: 0


Default

Of course, happy hunting!
derstosstrupp is online   Reply With Quote
Old 05-06-2021, 09:07 AM   #8
Efshapo
Watch
 
Join Date: May 2021
Location: France
Posts: 29
Downloads: 3
Uploads: 0
Default

I just noticed that you had actually already replied this same answer on my video. And I also already thought you were talking about the constant bearing method, I'm so sorry.

Thanks again for you obvious expertise!
Efshapo is offline   Reply With Quote
Old 05-06-2021, 09:10 AM   #9
derstosstrupp
Crush Depth Teammember
 
derstosstrupp's Avatar
 
Join Date: Aug 2008
Posts: 1,150
Downloads: 480
Uploads: 0


Default

Quote:
Originally Posted by Efshapo View Post
I just noticed that you had actually already replied this same answer on my video. And I also already thought you were talking about the constant bearing method, I'm so sorry.

Thanks again for you obvious expertise!
No worries at all! German is famous for loooionnnnngggg words and the words for these methods look similar at a glance.
derstosstrupp is online   Reply With Quote
Old 05-06-2021, 09:31 AM   #10
Efshapo
Watch
 
Join Date: May 2021
Location: France
Posts: 29
Downloads: 3
Uploads: 0
Default

BTW, there's another huge advantage in using the centiradians scale instead of turning the periscope, which is that you can use several chronos at the same time.

Of course, you can do it like I did in my video, measuring real and lateral speeds simultaneously. But you can also multi-check your measurements! I recently started to use one chronometer to measure lateral speed from -4 to +4 centirad, as in the video, but also from -5 to +3, and -3 to +5, which gives me 3 time values! I can then check that they are similar, and take the median one, or an average. I then do the same for real speed, using -1, 0 and +1 marks. It's a real gain in accuracy!
Efshapo is offline   Reply With Quote
Old 05-13-2021, 08:03 AM   #11
mobucks
Seasoned Skipper
 
Join Date: Jan 2010
Posts: 713
Downloads: 209
Uploads: 0
Default

If you really did figure this out on your own, I tip my hat to you. Very interesting! Thanks!
mobucks is offline   Reply With Quote
Old 05-15-2021, 01:24 AM   #12
ljqcn101
Sailor man
 
Join Date: Jan 2013
Location: Hong Kong
Posts: 50
Downloads: 74
Uploads: 3
Default

Pretty quick and useful method. Thanks.
ljqcn101 is offline   Reply With Quote
Old 05-15-2021, 06:26 PM   #13
Efshapo
Watch
 
Join Date: May 2021
Location: France
Posts: 29
Downloads: 3
Uploads: 0
Default

Quote:
Originally Posted by ljqcn101 View Post
Pretty quick and useful method. Thanks.
Thank you for your feedback!


Quote:
Originally Posted by mobucks View Post
If you really did figure this out on your own, I tip my hat to you. Very interesting! Thanks!
I did, thank you!
I also developed a TDC simulator in Excel to measure the error injected by the False AOB 90 method.
Attached Images
File Type: jpg TDCSim.jpg (97.7 KB, 22 views)
Efshapo is offline   Reply With Quote
Old 05-05-2022, 11:14 PM   #14
VonHoffman
Watch
 
Join Date: Oct 2019
Location: Near crush depth. Maybe.
Posts: 30
Downloads: 6
Uploads: 0
Default

Interesting technique.

I use something similar. What you're doing is instead of measuring bearing angle change by rotating the periscope, you are letting the target pass the more accurate graduations on the scope itself to yield a more accurate measurement.

The only issue I see is that your own ship needs to be stationary.


What I do is stop my boat, and time how long it takes for the entire visible length to pass the centerline of the scope. Regardless of AoB I know the target speed.

Example:

Target is directly abeam (90 degrees) and the target is 30 meters long, if it takes 10 seconds for the bow to stern distance to pass the centerline of my scope, then the target is travelling at ~ 6 kts.

If the same target is at relative bearing 120, and the AoB of the target is 30 degrees, how long does it take the ship to cross the centerline of the scope? IT STILL TAKES 10 SECONDS!

It appears to be moving slower, but the overall length of the ship still moves 30 meters in 10 seconds! As I'm measuring it crossing bow to stern across a fixed point, I can still get a good fix on its speed. Also, you can use any magnification you like to get a good view for the measurement as you only require the ship to pass a fixed mark - the center of the scope.

To get accurate AoB, I can then cheat:

If I know its range (let's say 10 hm), I know that a 30 m long ship should take 3 centiradians if it is at 90 degrees AoB. If I only see 1 centiradian (or 4 centiradians at 6x), its AoB is asin(1/3) = 20 degrees.

To parallel its course and remain in the same relative position:

180 - 120 = 60 off my stern
90 - 60 - 20 = 10 degree AoB less than I should see
I need to turn left 10 degrees and run at 6 kts to maintain relative position.

If I was already parallel to his course and he's 60 degrees off my stern I should have an AoB of 90 - 60 = 30 degrees. In the example it's not, so I must correct his relative bearing for the AoB seen.

Last edited by VonHoffman; 05-06-2022 at 04:12 PM.
VonHoffman is offline   Reply With Quote
Old 05-09-2022, 04:00 AM   #15
Efshapo
Watch
 
Join Date: May 2021
Location: France
Posts: 29
Downloads: 3
Uploads: 0
Default

Quote:
Originally Posted by VonHoffman View Post
Interesting technique.
Thank you. :]


Quote:
Originally Posted by VonHoffman View Post
I use something similar. What you're doing is instead of measuring bearing angle change by rotating the periscope, you are letting the target pass the more accurate graduations on the scope itself to yield a more accurate measurement.
Indeed. Only later did I discover a historically accurate technique involving a moving periscope existed ("Auswanderungsverfahren", as indicated by our friend Stosstrupp).


Quote:
Originally Posted by VonHoffman View Post
The only issue I see is that your own ship needs to be stationary.
Technically, you don't have to be stationary. You just have to take into account you own lateral speed. You can either compute it, that would be own_speed * sin(relative_bearing), and add it to the result. Or you can maneuver to make sure it's negligible, either by stopping as you stated, or by going at any speed you want provided you're observing your target from a relative bearing of 0 or 180.


Quote:
Originally Posted by VonHoffman View Post
What I do is stop my boat, and time how long it takes for the entire visible length to pass the centerline of the scope. Regardless of AoB I know the target speed.

Example:

Target is directly abeam (90 degrees) and the target is 30 meters long, if it takes 10 seconds for the bow to stern distance to pass the centerline of my scope, then the target is travelling at ~ 6 kts.

If the same target is at relative bearing 120, and the AoB of the target is 30 degrees, how long does it take the ship to cross the centerline of the scope? IT STILL TAKES 10 SECONDS!

It appears to be moving slower, but the overall length of the ship still moves 30 meters in 10 seconds! As I'm measuring it crossing bow to stern across a fixed point, I can still get a good fix on its speed. Also, you can use any magnification you like to get a good view for the measurement as you only require the ship to pass a fixed mark - the center of the scope.
That's what people usually do, yes. I actually use this in a step of my second "refined" method, as a way to get an accurate AOB. You can see it in the video.


Quote:
Originally Posted by VonHoffman View Post
To get accurate AoB, I can then cheat:

If I know its range (let's say 10 hm), I know that a 30 m long ship should take 3 centiradians if it is at 90 degrees AoB. If I only see 1 centiradian (or 4 centiradians at 6x), its AoB is asin(1/3) = 20 degrees.
The method I propose here to measure AOB is more accurate because the "sensor" used is the stopwatch instead of the horizontal scale. I only use an exact number of horizontal divisions so I don't have to estimate horizontal subdivisions. And if I manage to get my measurement under a minute, the stopwatch displays 1/100th of seconds which gives a far better sensor resolution.


Quote:
Originally Posted by VonHoffman View Post
To parallel its course and remain in the same relative position:

180 - 120 = 60 off my stern
90 - 60 - 20 = 10 degree AoB less than I should see
I need to turn left 10 degrees and run at 6 kts to maintain relative position.
I'm not sure I'm following you here... If the target is at relative bearing 120 and its AOB is 20° left, then if you want to run parallel you should turn 180-120-20 = 40° left (and then observe the target at relative bearing 160).


Quote:
Originally Posted by VonHoffman View Post
If I was already parallel to his course and he's 60 degrees off my stern I should have an AoB of 90 - 60 = 30 degrees.
Again, I disagree. Running parallel to a target observed at relative bearing 120 means its AOB should be 60° left.
Efshapo is offline   Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -5. The time now is 06:39 PM.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2022, Jelsoft Enterprises Ltd.
Copyright © 1995- 2022 Subsim®
"Subsim" is a registered trademark, all rights reserved.