Using a fictitious course to set TDC
 

I promised when I lead another thread astray that I'd try to make a diagram or diagrams to explain my concept of shooting based on a made-up target course. The situation is that you don't know what the range of the target is so you make one up! Taking bearings three minutes apart, you plot those bearings on your fake target track to calculate (you guessed it!) a fake speed!:D:D This is sounding promising already.....

Now, I have finished my first chart and it takes the simplest case, where the course is actually at 90º to yours but you just make up a range for fun and profit. Works fine. Here's the construction:

http://i196.photobucket.com/albums/a...ousCourse1.jpg

So how do you like THEM apples? Make sense? Actually you don't even have to shoot this as a constant bearing attack. Turn on the PK, put in the range, AoB (90º at the last bearing point) and speed. Then shoot whenever you feel like it! Fill it full of holes if I'm wrong, I can take it! I don't think I have left myself any weasel room to deny it in that case. Let's reason this sucker out together and make it work.

I thought of this when I saw this post.

http://www.subsim.com/radioroom/show...0&postcount=47

Urge

Yes, that animation started the whole revolution in fleet boat shooting here at Subsim. It was the result of U-Boat skippers gutted and aaronblood traveling to the dark side at the same time I was looking at the possibility of adapting U-Boat techniques to American submarines.

The more I look at my chart above the more unhappy I am with it. It looks too much like the Dick O'Kane chart and really should have two sightings, then the firing point at zero, then the torpedo track plotted for clarity. Don't have time to do it right now but I think the Mark II version of the chart should be done.

I also want to use it as a conventional targeting method with PK on so timing of the shot isn't critical. That would be fun too. Now that I can play SH4 again after I wiped and rebuilt my hard drives, this is a worthy first project.

Greyrider gets some credit on this one because a sonar technique will end up in the mix somewhere. Even though his 8010 idea never got anywhere he is still the first guy since I developed the Dick O'Kane sonar only method to attempt to develop a sonar attack method.

Anyone else who wants in, the water's warm! Contributions are welcome, including that of disproving the whole idea. Also if I have been unclear on any part of the concept I need to know. If the technique can't be taught it is useless. That means every step must be clear, easy to understand, easy to remember and easy to execute. There is no room for smoke and mirrors, unannounced tricks or unsupported claims in any legitimate tactic.

The fun is in covering every detail, making a set of instructions that anyone can follow successfully. The proof of success is when the student ends up shooting better than the teacher. The purpose of developing a technique is not so the originator can brag, but so that the students can brag about what great shots they are. That's the reward, right there! "I was frustrated and ready to quit playing but learning the xyz attack made me a successful skipper and saved SH4 for me" is the big payoff. I'll take a hundred times more criticism for the simplicity and lack of originality of the Dick O'Kane and John P Cromwell techniques in exchange for just one of those. Instead, I have received dozens just like it, and have repeatedly announced and shared the credit with all who contributed to the cause, especially to Neal Stevens who created the unique community in which this could take place. There are other sub simulation websites, but only one place where things like this happen.

Thanks RR, for taking the effort of making it visual.

It certainly does look like the Dick O'kane method, but with one critical flaw. It relies on an assumed target speed. If the target speed is different than the real speed then you could miss ahead of the bow or behind the stern. It all depends on how much leeway the target's length gives you. I'm afraid this technique is in no way more guaranteed than Greyriders 8010 method.

Hmm
 

It seems to me that you'd have to fire a spread to have a hope of hitting the target. I like it though.

Now be careful there! We're not just making up a speed. We're basing it on observation of two sonar vectors and projecting them on a fictitious target track. So long as the speed and track are proportional to the real one, you have created similar triangles and it will yield a valid solution.

Haven't you ever done the stadimeter thing and found through the attack map that your range is too short? As long as the impact x is on a straight line between your sub and the target, you have a hit, even though your range is wrong. You do have to reduce the speed proportionally to maintain the alignment if you're using the PK especially. That error was the inspiration for my idea. The only thing the solution has to yield is the correct gyro angle. If the correct angle is the product of an inaccurate speed and range the torpedo still makes a boom.:D

RR
(and everybody else interested)
regarding the ficticious course attempt
up to now I have this: link

Hope it is "understandable"...

Fire Away!

(going deep now ...) :lurk:

Ok my apologies, appearantly I still don't have my head screwed on correctly. I didn't realize the speed was derived from the bearing progression at this ficticious distance. You wrote it in the first paragraph, but I focused on the seemingly 'magic' 7 knots in the 2nd paragraph. Anyway yes, due to the proportional triangles the torpedo track would be the same.

Sigh, I need some R&R :zzz:

p.s. no pun intended with that.

VERY interesting. Ideally I'd like to come up with a graphical solution so that external calculators aren't necessary. But I'm thinking that placement of the fictitious course influences the solution. Your diagram is better than mine. I need to get to work on something similar, showing two bearing sights, plus the firing position, plus the torpedo track.

Then the question: how does placement of the fictitious course influence solution quality? What kind of error tolerance do we have? I'm thinking at this point that we might have to use a visual AoB estimate to fix a non-right angle position of the fictitious course.

Good work! Now you've made me burn some midnight oil.:dead:

For starters assume that you know target course.
Why? Because there is a bearings only method that allows you to do so. Implementable via passive means (sonar or optical observations no radar required). You won't like it though, as the sub must remain stationary as it collects data (at least three distinct bearing/time observation pairs). But if the target' is a medium or slow "mover" and you detect him early enough you may have time both for proper data collection and maneuvering.

{ I realy think you must go through this link: http://www.archive.org/details/maneuveringboard00unit }

Then you'll have to deal only with target's speed inaccuracies as range probably doesnt matter for small torpedo gyro angles.


.

I've read through that and it's interesting. But I don't think it's usable for the average person playing SH4 and the techniques were definitely developed after WWII. And like you say, you have to stop the boat. I understand that in real life they have techniques where they don't have to stop the boat but our nav map doesn't have the tools to allow us to do the bearing rate chart. I just hate to break concentration by pausing the game and using outside tools. I know....it's a personal problem.:D

It's the 1941 manual. It fought in WWII alright! :DL

The "keep it in the game" principle is a very ... mature one indeed. But there can be "legitimate" exceptions such as using a Maneuver Board. After all you can plot in real time (not pausing the game). There are other kinds of solutions that are not as "knowledge intensive" as proper nav work. (And no I don't mean alt-tabing to an Excel spreadsheet). I'm in a brainstorm phase already! :doh:

.

RR,
any progress?



.

Dont' misunderstand me. The in game only stuff is nothing but a personal choice. I'm not saying that using the maneuvering board, MoBo, Solution Solver, is/was, Capn Scurvy's tools or anything else are cheating. I just don't like leafing through pages of torpedo speeds, picking a row for target speed, a column for torpedo speed and plucking out a number because I'm dyslexic or something and keep making mistakes which are easily avoided by using what I do. In real life they really did all those things and had a whole bunch of other tools we don't, like bearing difference and bearing rate tables, stadimeter plots, tons of slide rules, etc. I'm just a guy who loves to boil everything down to the lowest common denominator, then explain it in english so clearly that my cat can outshoot me two out of three times.

You know, you're right about one thing. The maneuvering board is something very different from our nav plot. Real subs had both. I saved a copy of the manual as everything I have is from 1946, after the war ended.

It's not that I don't understand the higher math and I'm not afraid to share a room with a trig table. I'd just rather not if I can avoid it...:D

No progress yet. I've been wasting too much time on that other thread. Think my efforts would be better spent here.

No problem!
BTW I truly meant "mature" (it was not ironic).
But have you done anything on "ficticious"?


.

OK, what do you want to call progress?:doh: Let's just say that intuition doesn't always lead to the truth. I set up a situation in SH4 because other than MoBo, which I need to seriously do a refresher course on, SH4 has the most appropriate drawing tools for the situation. Inkscape just wasn't cutting it.

Now using the game to plot also gives me another advantage. I can kill all the trig tables, formulas, unclear hocus pocus altogether and let the TDC do the math for me. You can check my work at a glance and tell whether I've made any mistakes. I'm not here to make claims that you can't prove or disprove easily.

Here's the plot. I should annotate it, but I was pretty excited about what I found and wanted to get it posted quickly, especially since I've kept you waiting too long already.

http://i196.photobucket.com/albums/a...reenshot94.jpg


We have a real target out there moving left to right on a course of 70º at 4.5 knots. But shhhhhhhh! We don't know that. Let's take a couple of sights and stick them on a fictitious course line 1000 yards away, running at right angles to our course. You can see the two courses actually cross and this should be an accurate setup if there is such a thing.

We take our first bearing at time 0:00 and find a bearing of 300º. There's the bearing line and our fictitious position is where the bearing line crosses the fictitious course line. The real position is where it passes the skewed real course line. Shhhhhhh! We aren't supposed to know anything about that.

Three minutes later we take our second bearing of 314º, plotting that on our fictitious bearing line. You can also see the real position where it crosses the real course line. Shhhhhhhh! Don't spill the beans...

With our compass we can measure the distance on our fictitious course line as 700 yards, which translates to 7 knots. We're going to fire when the target reaches the zero bearing so our angle on the bow for an attack 90º from our course will be 90 minus zero or just 90º starboard. We put a range of 1500 yards or so (doesn't matter for this application) into the TDC and with our four parameters entered we can check out our gyro angle.

http://i196.photobucket.com/albums/a...reenshot96.jpg

Nice thing about the TDC is that you can verify that I have input the AoB at 90 starboard, speed 7 kt, range who cares. Gyro angle? 10º. Now let's enter the real numbers and see what we get.

You can see on the plot that his AoB is really 110º starboard, his speed 4.5 kt and his range 1500 yards exactly. Shhhhhhh! This is top secret information. We can enter this info into the TDC. If the gyro angle is the same, we have a hit! So with great anticipation and foaming at the mouth, Napolean (arcane reference to a badly botched attempt at a personal insult in another thread) enters the golden numbers. Hehehehehehehe!

http://i196.photobucket.com/albums/a...reenshot95.jpg

Fail! 6º is the actual gyro angle for a hit. Set by our fictitious course data, our gyro angle was 10º so we're 4º ahead of MOT (standard US Navy for middle of target. I don't make up arcane abbreviations or use non-standard variables in my formulas) if that's what we're aiming for.

OK, how bad did we miss? What the hell, let's do some hocus pocus trigonometry. Well, at 1500 yards, .0175 (approx) is the sine of 1º. Multiplying that by the 1500 yard range yields an error of 26.25 yards per degree of error. We have four of those, so we're off by 105 yards. Figuring that our average merchie is about 130 yards (390') long, taking half of that (since we are firing at MOT) we have a 65 yard allowance for error, our shot misses by 45 yards ahead and we're dog meat.:lost:

Now we could get hits with a small enough range. For instance, at 500 yards 1º of error is only 8.75 yards. We'd hit four of those, 35 yards ahead of MOT for a big boom and be a hero! But I daresay at 500 yards you can just eyeball a lead with no calculation at all and hit just about anything, so our method is just a waste of time.

We could'a had a V-8!:D

Note: anybody catching on how to rigorously prove a method? I did make a small error in the hocus pocus section. Can anyone catch it and apply the correction? Play the Jeopardy waiting music!

Now, what went wrong here. I think number one is the inappropriate use of a constant bearing technique. After all, the constant bearing techniques work at any range if you know the target speed. In other words, the solution is just the same at 1000 yards as it is at 2000 yards.

The same gyro setting hits both targets moving the same speed! However when we project bearings onto a fictitious course we end up changing the speed, according to whether the fictitious course is nearer or farther than the real target.

So though intuitively, it seems reasonable that this should work, it cannot. I will now investigate more conventional targeting methods.

I have had a bad stadimeter reading and set for a range over or under the real range. Using the attack screen I can adjust the speed to a wrong speed that is proportional to my error.

For example, suppose my input range is too large. I can input a speed higher than the target's real speed. If I get the right one, a line from my submarine through the impact point in the empty ocean intersects with the target position. That is a valid solution and I will get a hit.

The same thing applies to a range too short. By reducing the speed, I can make a situation where a line from the submarine through the impact poiint on the attack map intersects the real target. I'll hit that target with a bad range and speed.

In both instances the AoB/target course has to be nearly right.

So that's where I go next: to familiar ground and outside my normal preference for constant bearing techniques.