[WIP] Improved Ship Physics

[TorpX]

Originally Posted by Johnners

Hi Torpex, and thanks for your great mod.

I hope I'm not asking a dumb question, but after a prolonged period of running submerged and returning to the surface, I went to move the "battery expert" to an other compartment while the battery recharged (as per instructions), only to find him apparently sleeping (little z's above his icon) while the rest of his watch were still active. Is this usual and does he still need to be moved, our should I have him keelhauled for slacking off???

No, not a dumb question.

It is normal for him to have a different sleep cycle, since you have to move him around, and this can interrupt his sleeping. You should still move him out of the engines compartment; the game is made so 'passive' abilities are always on, even if the crewman is sleeping.

As your patrol continues, you will notice crew sleeping patterns change as their fatigue levels will differ.

[Johnners]

Thanks for the reply TorpX and again for the great mod

[TorpX]

A warning to those SH4 players, who still like to think of the game as a reasonably faithful submarine simulation. The following information may damage their enjoyment of the game. (It did mine.)

 

I was working on adjusting the crew slots of the different classes of boats to get a better/different level of efficiency, when I decided to do some testing of how the boats work with reduced crew levels.

As a first experiment, I put most of the duty shift into Hogan's Alley, so they would be out of the picture. Then, I proceeded to order a number of typical actions: diving, turning, speed changes, torpedo launches, etc.

I found, to my disgust, most of these actions do not require the compartments to be manned at all.

There were a couple of exceptions. Engine operation requires crew in that compartment (especially with ISP), Torpedo loading requires crew in the torpedo room, but launching does not. Of course, we can both fire and load the deck gun, without crew. Diving and surfacing (usually thought of as a complicated and demanding task), is accomplished as easily without the crew, as with them.

I did not try to test damage repair, as I can't easily damage the boat in a controlled manner. Frankly, I wasn't all that eager to know, at that point.


Most of the this green bar stuff seems to be monkey points. Why, for Pete's sake, would Ubi put in all this fussbudget stuff with efficiency levels, crew abilities, and fatigue, when it hardly matters?

I suppose it's anyone's guess.

I hoped maybe that there was a simple problem in the files that could be changed to fix this. I spent quite a few hours trying to re-engineer the *.upc files to force the game to take account of the crew staffing, all to no avail.

I might not have even posted this, but there is always the chance someone who has a good understanding of computers/software can fix this in the future.

I made the initial tests with RFB +... + ISP. To check to make sure it was not an ISP issue, I repeated the test with RFB alone, and then with stock. Results were the same.

[ColonelSandersLite]

Honestly, that doesn't really bother me at all.

 

Crew efficiency effects:
sensors (definitely)
reload rates (definitely)
damage control (pretty sure)
gun accuracy (maybe, not really sure tbh)

While theoretically possible to encounter the situation you mention, the only time I don't have any crew in an area is my guns when I don't need them. Technically, you don't need crew in them to use them yourself, but reload rates are atrocious then so that's a kinda pointless distinction. Yes, it's possible to strip a lot of crew out of the sub at port but why would you even actually do that?

On paper, it's possible to have enough casualties that this can be a serious problem. In practice, I doubt it's really possible to survive that anyways.

[TorpX]

I've been fiddling around with the torpedo files, in the course of releasing AutoTarg. I found I can add 2 modes of failure to a torpedo:

1. a torpedo can 'sink', as in a cold run. The torp keeps going deeper and deeper, until it hits bottom and explodes.

2. a torp can 'porpoise' and run on the surface. Not sure if enemy ships would see a surface runner any easier or not.

Also, I looked into the RFB 2.0 patch file, and it seems that the dud rate is on the high side. Can anyone who plays a lot of RFB shed some light on this? Do you have a very high rate of duds?

I am thinking of either doing a separate torpedo mod, or a new version of ISP with some changes here.

[swdw]

Originally Posted by TorpX

2.0 patch file, and it seems that the dud rate is on the high side. Can anyone who plays a lot of RFB shed some light on this? Do you have a very high rate of duds?

I am thinking of either doing a separate torpedo mod, or a new version of ISP with some changes here.

Been gone a while. Couldn't run SH4 for 2 years. Have now built a WinXP game machine and hope I can get back to playing again. I led the RFB mod team until turning it over to Luke FF.
Now onto some stuff you've written about.

Dud rate early in the war is historically accurate. I did a bunch of research on this before spending almost a month to get the torps tweaked in to match patrol reports. If no one tweaked the files after I quit working on it, then if you look closer at the files, you'll see you're dud rate is less at 45 degrees or greater, which is also historically accurate. You'll also see the dud rate lessen if you shoot them at slow speed. Also historically accurate.

If they got tweaked after my mods, I can't answer for that.

Going back to an older post on the sub physics. The acceleration, turning radius, and other parameters were pretty close to accurate. We could not make them dead on due to the fact some of the settings also affected the dive rate too much. (drag being one of them)

Same thing with surface ships.

One thing to keep in mind. The physics engine does not take into account different propeller profiles. Besides the number of blades, the surface area and twist also affect the thrust. Not modeled accurately in the game. SO you often have to go with unrealistic numbers for HP and other parameters in order to get more realistic performance.

I exchanged a number of e-mails with the developers and learned what the limits of the engine were. Then I had to work around them. I didn't make it through all of the surface ships at the time. By the end of it I probably had a better understanding of the physics engine when it came to ship performance than the developers.

As for the height of the subs in the water. Prior to the U-boat mission release the subs sat WAY too high on the surface. If you can find my original post, as this was one of the first mods I did, then you can see the difference. When you start fussing with the height in the water, it can also affect dive rate, acceleration and turning radius. Requires a lot of testing to get right.

When ubisoft released the U-boat mission add on, my understanding is they used the RFB values to get the subs to ride correctly.

SO you know, when we were setting how the boats rode on the surface, we had 4 ex diesel boat sailors on the team as testers. We came up with a good all around height. The real boats ride lower with a full fuel load and then ride higher as they burn through fuel. There are other things including a full stores load that affect this. So we got feedback from the ex diesel boat testers and came up with a good compromise height.

[TorpX]

Originally Posted by swdw

Been gone a while. Couldn't run SH4 for 2 years. Have now built a WinXP game machine and hope I can get back to playing again. I led the RFB mod team until turning it over to Luke FF.
Now onto some stuff you've written about.

Welcome back.

Going back to an older post on the sub physics. The acceleration, turning radius, and other parameters were pretty close to accurate. We could not make them dead on due to the fact some of the settings also affected the dive rate too much. (drag being one of them)

Well, I don't mean to be critical, I know you and the others in the RFB team had a huge task before you, but I had to do a lot of work on these things - acceleration and turning. Turning radii were ok, but the turning speeds were way too low. Acceleration was too fast. I was able to fix some of these things.

I exchanged a number of e-mails with the developers and learned what the limits of the engine were. Then I had to work around them. I didn't make it through all of the surface ships at the time. By the end of it I probably had a better understanding of the physics engine when it came to ship performance than the developers.

That's the problem. You, I, and anyone who has worked on the physics probably had a better understanding than the developers. I've come to the conclusion that they really weren't interested in the matter. There are too many problems for me to think otherwise.

There are things they could have easily corrected, if they had been paying attention.

Dud rate early in the war is historically accurate. I did a bunch of research on this before spending almost a month to get the torps tweaked in to match patrol reports. If no one tweaked the files after I quit working on it, then if you look closer at the files, you'll see you're dud rate is less at 45 degrees or greater, which is also historically accurate. You'll also see the dud rate lessen if you shoot them at slow speed. Also historically accurate.

If they got tweaked after my mods, I can't answer for that.

I understand that there was an effort to make them historically accurate, but afaik, there are no tests that state at A angle you get X% duds, and at B angle you get Y% duds, etc., etc.

Below are notes from the Torpedoes_US.sim file in the RFB
'patch'. The RFB 2.0 file is entirely different. In fact the dud
chances track in the opposite direction (i.e. large 85 chance at
large angle, going to small 10 chance at smallest angle).

[In the files, an angle of '0' means a perpendicular impact, 90
deg. would be a glancing hit. To be sure of this I had to verify it by experiment.]

1/1/1939 to 6/23/1943
dud chances

angle 72 to 90
chance 5.0

angle 54 to 72
chance 25.0

angle 36 to 54
chance 50.0

angle 18 to 36
chance 75.0

angle 0 to 18
chance 100.0

dud reduction speed 34.0
dud reduction rate 50.0

premature chances
waves 0.0 to 40.0
chance 10.0

6/24/1943 to 9/29/1943

same dud chances.....
no magnetic detonation or premature..

9/30/1943 to 12/31/45

angle 72 to 90
chance 1.0

angle 54 to 72
chance 2.0

angle 36 to 54
chance 2.5

angle 18 to 36
chance 3.5

angle 0 to 18
chance 5.0

same dud reduction speeds and rates...
no prematures...

First off, a 100% chance seems high by any standard. Second, the next period shows no improvement, when there should be one. Maybe the 100% chance of a dud is justified by Lockwood's drop tests(?), but ships were being sunk in this period, and later in the war, the sinkings improved, but not that dramatically.



From a gameplay standpoint, I dislike a 100% chance, as this is likely to cause players to just turn off duds altogether, or incorporate ahistorical tactics, and thus render the matter pointless.





Anyway, I want to thank you for your, and the others' contributions. Without the major mods, the game would be nearly unplayable.

- TorpX

[swdw]

Originally Posted by TorpX

I understand that there was an effort to make them historically accurate, but afaik, there are no tests that state at A angle you get X% duds, and at B angle you get Y% duds, etc., etc.

- TorpX

Someone changed the files after I turned it over. just from the numbers I can tell that wasn't my work. Like you, I tested it to make sure the numbers went in the right direction.

As for dud rate, here you go, from several sources.

1. Tests were carried out by COMSUBPAC's gunnery and torpedo officer, Art Taylor. Taylor, "Swede" Momsen, and others fired warshots into the cliffs of Kahoolawe, beginning 31 August. Additional trials, supervised by Taylor, dropped dummy warheads filled with sand from a cherry picker raised to a height of 90 feet (27 m), producing a 70% failure rate. A quick fix was to encourage "glancing" shots (which cut the number of duds in half),until a permanent solution could be found.

2. The torpedo controversy came to a head in July 1943 when the USS Tinosa received intelligence that a large Japanese tanker would pass
through her patrol area the next morning. They fired four torpedoes from 1,000 yards. The sound man could hear them hit, but no explosion resulted.
The skipper was about to cry and the XO and I said ‘Captain, this ship was tracking right on course with the speed and course we got it exactly right.’ He said ‘We will fire two more torpedoes at its stern and I will angle my periscope.’ We fired at it at 4,000 yards which is two miles with one miss and one that hit its stern and blew its stern up and it could not move again. Well, he sat there. We fired, over the next three to four hours, 12 more—one at a time. We fired one side; we would go round to the other side. Consternation and frustration was extreme. None of the 12 torpedoes exploded. The Americans were finally
chased away by Japanese ships sent to help the beleaguered tanker.

The Tinosa's Captain, Dan Daspit, saved his last torpedo as conclusive evidence that something was very wrong.

Early reports of torpedo action included some dud hits, heard as a dull clang. In a few instances, Mark 14s would strike a Japanese ship and lodge in its hull without exploding. The contact pistol appeared to be malfunctioning, though the conclusion was anything but clear until running depth and magnetic exploder problems were solved. Daspit's experience was exactly the sort of live-fire trial BuOrd had been prevented from doing in peacetime. It was now clear to all at Pearl Harbor the contact pistol was also defective. Ironically, a direct hit on the target at a 90 degree angle, as recommended in training, would result in a failure to detonate; the exploder only functioned when the torpedo impacted the target at an oblique angle.

3. Lockwood's men replaced the TNT in several warheads with cinder concrete and attached the normal contact mechanism. Test torpedoes were then dropped 90 feet along a wire suspended from a crane into an empty drydock where they landed squarely on steel plates. A direct, 90-degree hit produced a dud seven out of 10 times -- a 70 percent failure rate almost two years into the war. By adjusting the target plates to a 45-degree angle, the failure rate was cut in half. At a still greater angle, the exploders worked without fail. Lockwood immediately directed his boats at sea to launch their torpedoes from large, obtuse angles. They were ordered to improvise, to use anything but the textbook 90-degree track.

So yes there are actual numbers. 70% for a perfect shot and 35% at about 45 degrees. so you can take it from there but my understanding was that it was not a perfectly linear drop off and there was only a small change in the dud rate between 90 and 70 degrees.

BTW, there were some VERY pissed off people when it came to the dud rate when I released earlier versions of RFB. Maybe that's why it was changed for RFB 2 ??

As for the boats, wish I could compare the sub files I had compared to what was in the final release of RFB 2, but I no longer have them. There were complaints that I made the boats accelerate too slow, so maybe that was changed.

Also, when you talk about turning speed and radius of the boats, are you testing them both on the surface and underwater?

I did find a limiting factor when playing with the capital ships. If I got the acceleration, deceleration and coasting too realistic, the AI had collisions all the time. Because of this I wanted to go through and make an initial adjustment on all of them for consistency and then go back and make further adjustments. I think I only made it through about half the Japanese cruisers and battleships. The plan was not to include the changes until all the BB's and cruisers were completed. So the changes probably aren't in RFB

Never got a chance to do much with the destroyers. This was a frustrating issue as they act like a speedboat at times

[TorpX]

Originally Posted by swdw

Someone changed the files after I turned it over. just from the numbers I can tell that wasn't my work. Like you, I tested it to make sure the numbers went in the right direction.

I guess there have been so many revisions of RFB, it would be hard to know the where and why of all this. I played some with the earlier versions, but I didn't do any modding, and just accepted things 'as is' in those days.

3. Lockwood's men replaced the TNT in several warheads with cinder concrete and attached the normal contact mechanism. Test torpedoes were then dropped 90 feet along a wire suspended from a crane into an empty drydock where they landed squarely on steel plates. A direct, 90-degree hit produced a dud seven out of 10 times -- a 70 percent failure rate ...

So yes there are actual numbers. 70% for a perfect shot and 35% at about 45 degrees. so you can take it from there but my understanding was that it was not a perfectly linear drop off and there was only a small change in the dud rate between 90 and 70 degrees.

This is new to me, as I didn't know they tabulated specific percentages. This is very helpful. It does stand to reason there would not be a large difference from 90 to 70 deg., as the impact forces would not be much different.

Also, when you talk about turning speed and radius of the boats, are you testing them both on the surface and underwater?

Yes. Certainly!

The radii were good, but the speeds were very low. It is a stock issue, the game engine seems to want to slow everything down a lot in turns. I had to play with things a lot to get what I wanted. For general types of ships this may not be a big deal, but for a submerged sub, it's critical bad news.

I did find a limiting factor when playing with the capital ships. If I got the acceleration, deceleration and coasting too realistic, the AI had collisions all the time.

I haven't noticed a problem there. It might have been a stock issue. From what I've seen, whether one uses a physics mod or not, convoys will be jumbled to some extent when they are spooked, or attempt to evade, but I think they are ok, if they have adequate spacing to start with. I used 1000 yds. between columns, and 500 yds. between rows, and they did ok. If they were too tight, I could see how there might be problems. Maybe earlier versions of game didn't work as well?

Never got a chance to do much with the destroyers. This was a frustrating issue as they act like a speedboat at times

Fixed that.

None of the ships or subs had a value for the max force. Nobody knew what/how to use this. The value is supposed to be in tons of force, but I know it doesn't really work that way. They were all zeroed out. If a suitable value is used, the acceleration is much more realistic. That Ubisoft didn't do anything with this makes me think they miscoded something, and rather than track down the problem and fix it, they did the easiest thing, and simply zeroed everything out. This was ok for them, but left us with terrible physics.

Battery performance was awful. I'm sure you already knew that. Ducimus came up with a fix for that. That's important.



Anyway, I've done about as much as I know how to do for the game. The torpedo thing was kind of a chance thing I noticed after I had an earlier idea. I was surprised about the high dud rate. I had expected no more than 70%, or so.

There were some other things I tried to fix, but most of these ideas didn't work out.

[TorpX]

I am constructing a torpedo.sim file based on the 70% max. dud rate discussed below. I'm also using the two additional modes of failure I tested before.

I tried to add a 'bow splash' to the torpedo, so a porpoising torp would look better, but it didn't work. Either I didn't do it right, or it just won't work. I happened to get one in a test with a shot at a DD. As far as I can tell, they cannot be seen any easier than a normal torp. Too bad.

I had to do quite a few tests to get a feel for what wave level cutoff makes sense for the premature rate. In the file prematures are a function of 'wave height'.

I ran a mission with ISP's scene.dat and had the wind at 13 m/s level. I set the premature parameters to 0% premature chance, wave 0 to 20, 100% chance waves 20 to 1000. None
exploded.

Then tried 0%, waves 0 to 10, 100% waves 10 to 1000. All 6 exploded, including one surface runner.

Set divide at 0 to 15 and 15 to 1000, got mixed results; 9 exploded, 15 completed run at high speed. Depth setting errors normal for mk 14; some I set shallow, some default 10 ft. I didn't see any difference. I had expected all or none would premature, since I had it at 100% chance. It isn't clear how the game does the mechanics for this. Without tabulating any figures, it seems most prematures occur soon after arming. If they can make it through the first half of their run, they usually will complete it.

Set divide at 12; 16/16 exploded.

Set divide at 13; 15/24 exploded. Using the same parameters in a mission with 11 m/s wind, 15/16 completed runs (I believe one sank as there was no surface explosion). With these parameters in 15 m/s wind, 12/12 exploded. I fired 4 with M.I. feature turned off; none exploded.

From these tests, I would consider a high waves cutoff point of 12 or 13 to work well (with my scene.dat file).

Note that although the file structure seems to imply prematures are possible with contact exploders, this has never happened in my tests. Since prematures historically were associated with the magnetic pistol, this is ok, but I thought I would test to find out for sure. Different depth settings didn't seem to affect chance, as far as I could tell, but I didn't do systematic tests for this. Ideally, deeper torps would premature less often. I mostly used default depth, with a few salvos at min depth.

Finally, I tried changing the torpedo wake in the *.sim file to see if I could make them harder to see. I initially made the wake narrower, and offset it back 25m. It is easy to see the difference, but ships detect them the same. I made the length zero, thus rendering the wake invisible (to us), but a DD will evade just the same.

[TorpX]

I'm trying to finish the next version of ISP. I, perhaps foolishly, decided to take a look at the aircraft.

I put together a mission with flights of Japanese planes flying in to attack vulnerable almost defenseless American ships. Oddly enough, the IJ pilots can't seem to sink them. (This is especially odd, as they have sunk me more than once.)

The more I see, the more flaws I spot. I haven't really started changing them, so far; I just want to establish a base line of performance.

Two glaring flaws as I see it:
[I'm using the RFB+RSRDC set up.]

The Bettys dive and climb like they're driven by over-caffeinated fighter pilots. I saw one dive at an angle, at least as steep as a Val would make, then pull out abruptly, make a vertical climb, until he stalled. This isn't dangerous in the game, as he recovered very quickly, and went on his way.

Perhaps worse, is the fact that I cannot get the Kates to make a successful torpedo attack. I initially had them come in at a few thousand feet, assuming they would glide down to release the torps. HA! They just drop them from cruising altitude, so they explode at impact. Then I had them scripted to come in at low altitude, looks better, but the torps still exploded a moment after going in the water.

I looked at the Air_Torpedo_JP.sim file, and found it has no defined arming distance, (or much else). I put in a 20m arming value, then a 50m value, but they blow up just the same. I also reduced the speed of the Kates'. Nothing seems to work here. I might also point out, they drop the torps so that they hit very close to the target ship. Sometimes they do hit them. As it stands now, they are, in effect, bombs, and nothing more.

If anybody has scripted, or witnessed SH4 air attacks, and can shed some light on this, please enlighten me. I was under the impression that 'air torpedoes' were supposed to work, as advertised.

[CapnScurvy]

Originally Posted by TorpX

I'm trying to finish the next version of ISP. I, perhaps foolishly, decided to take a look at the aircraft.......


Two glaring flaws as I see it:
[I'm using the RFB+RSRDC set up.].......

TorpX, you may have hit upon the reason for your trouble with those last two sentences, and didn't even know it.

RSRDC introduces the Air_Torpedo_JP weapon to the game. It's not a stock weapon, neither does RFB add it. I wouldn't want to say for sure what the problem is with the weapon, but unless you want to re engineer Lurkers mod I'd count it out from any tests with aircraft. As a matter of fact I'd stay away from RFB too, and do any tests strictly with the stock game aircraft. Work with what you know works, before looking for trouble with someones else's modifications.

Something I've run into before with RSRDC planes is they aren't cracked up to work as expected. While using your two mods you've activated, go to the Museum and find the American Avenger plane introduced by RSRDC. Watch it for more than a couple of minutes. It flies fine at first but, slowly the nose will rise, then straight into the sea! You'll never see this plane in-game, because as soon as its spawned it will soon nose dive into the sea. Some of the other planes won't have sound. That's easy to fix, but why bother....their someone else's mod.

While in the Museum, scroll through the various nationality units as well. If you don't get a CTD with at least one of them, you'll be lucky. The Museum is a great way of checking whether a unit will create a problem in-game or not.....no CTD in the Museum, it shouldn't CTD in-game. Well, at least maybe it won't!!

Anyway, the issue with the Air_Torpedo_JP is something from RSRDC. As well as some of the other issues you've pointed out. I'd leave dead dogs lay unless you really want to fix someone else's mod (I've done it before, but no one ever notices ).

[TorpX]

Originally Posted by CapnScurvy

TorpX, you may have hit upon the reason for your trouble with those last two sentences, and didn't even know it.

RSRDC introduces the Air_Torpedo_JP weapon to the game. It's not a stock weapon, neither does RFB add it.

I'm guessing Lurker wanted to add it. I've pretty much given up on the air torpedoes. Even if I got them to work, with the AI pilots aiming for impact hits, I'm not sure it makes much difference. Oh well. They seem to get a fair number of hits, if they go in low.

Something I've run into before with RSRDC planes is they aren't cracked up to work as expected. While using your two mods you've activated, go to the Museum and find the American Avenger plane introduced by RSRDC. Watch it for more than a couple of minutes. It flies fine at first but, slowly the nose will rise, then straight into the sea! You'll never see this plane in-game, because as soon as its spawned it will soon nose dive into the sea. Some of the other planes won't have sound. That's easy to fix, but why bother....their someone else's mod.

I did as you suggested, and I think I can shed some light on this.

I followed the Avenger, and just as you said, after about 5 min. it went wobbly, with the nose and wings going up. However, it did not crash. I kept watching, and sure enough, after another 5 min., it did it again. Five more minutes, and there he goes again. Curious, I pulled up some of the other aircraft. The American 'fighter' ('Buffalo'?), the other torpedo plane ('Devastator'?), the P-38, the A6M2 'Zero', and the H6K 'Mavis'. They all did the same thing, though the timing was different. My hypothesis is that the game has them in a sandbox, and they must turn to the left when they reach the edge.

I didn't see any crash, but the turns were awkward, at best. The P-38 had it's wings nearly vertical.



I'm trying to tamp down the sharp turns and very steep dives/pull-outs; at least for the multi-engine types. I know Germany and Japan were big on dive bombing, but I don't think Betty and Mavis types should be diving that steep.

Perhaps a bigger problem than the a/c physics, is the AI pilot tactics. They seem to want to do things that they should not, or could not do. In some of my tests, they will fly level, make a steep dive, bomb, pull out ok, but get into trouble trying to reach a higher altitude in a very short interval. A real pilot would know better than to try to climb faster than their plane can manage, but the AI seemingly does not. On a similar note, I noticed that the Kate pilots will often fly themselves into nearby hills, apparently not seeing terrain. They can fly through the trees, but if they hit land, they've had it.

[CapnScurvy]

Originally Posted by TorpX

I followed the Avenger, and just as you said, after about 5 min. it went wobbly, with the nose and wings going up. However, it did not crash. I kept watching, and sure enough, after another 5 min., it did it again. Five more minutes, and there he goes again.

Hmmmmm, I wonder if it has something to do with me using RSRDC v550 instead of v575. My Avenger takes a nose dive every time.

That wobble you've seen...... I've seen it too after a few minutes of watching the Museum planes. Since the camera is directly attached to the units, it's hard to get a bearing on which way the plane is flying. With the camera having the Free Movement capability, you can see the sunlight is on the right side of the plane when you first view it. After the wobble, the plane seems to be flying back in the direction it just came because the plane is now "lighter" on the left side. The wobble seems to be when the plane makes a turn in flight. It's hard to see the turn since there's nothing to get a bearing on when the camera is flying right along with the plane and there's nothing to get a direction of (the sea is no help), except for the direction of the sun light.

I'll admit the turning radius of the planes are unrealistic.......too sharp, but I don't know if changing the .sim file will create a different outcome, or start some other unforeseen issue?

All planes travel the same scripted path for the Museum. The way I know this is that when my Avenger crashes into the sea the smoke from the crash site is visible when the next plane flies over the area!

I'll try to see if v575 has something different than v550.

[TorpX]

That could explain it.

I've been trying to dampen both the abrupt turns, and the steep dives/climbs. Trying progressively lower values for the rudder drag, and wing drag, I got down to values of 2*10^-20. Since even this didn't help, I tried zero. Either the game has some limit overriding it, or it just ignores it.

Smaller values of the max force make the aircraft less lively, but this has both positive and negative effects.

I'm thinking it isn't going to be possible to make great improvements in this area.

I've also noted that the bombs, and torpedoes actually lunge forward when they are dropped. Drag should cause them to fall behind, but it doesn't happen that way. I guess this is a fairly minor thing, but I wonder if this affects AI aiming. Looks odd, if you happen to see it.

In any case, I want to finish this aircraft business, one way or the other. There aren't that many parameters to work with, and what there are don't seem to be terribly useful.