P-38 flight file follies - opinions and input sought

[Captain Kurt]

Hi Ivan

You are absolutely correct that adjusting drag coefficients to obtain a top speed at critical altitude will mess with the speed at sea level. So will obtaining the distance desired at maximum cruise settings (in this case the mean of the pilot's manual speeds between 250 mph and 300 mph) mess with the distance obtained at "long range" settings (again a mean between 165 mph and 230 mph). Obtaining the two former will cause too high values on the latter. Obtaining desired numbers for the latter will cause too low top speed and short max cruise distance. I think CFS2 flight algorithms are too generalized or unsophisticated to get everything in sync no matter what adjustments are made. You can have every aircraft parameter correctly entered into the air and cfg files and the result is an aircraft that bears no performance resemblance to the real aircraft. Hence we (me anyway) are stuck with tweaking and compromises.

To answer some of your questions

Yes I have engine power dialed in with RPM / Hg settings and military power overboost Hg.

Bear in mind that in a campaign or mission using the warp function, it does not matter what the player sets the engine and throttle setting at just before he warps. The speed programmed into the waypoint determines how fast the player flies to the next waypoint. And the game determines what the throttle settings would be for waypoint to waypoint based upon the aircraft's flight files. The player's throttle setting only matters if he is flying the whole distance in real time. Auto mixture is turned off in the air file so it the player wants to fly a distance by hand he can adjust the throttle, mixture and RPM to obtain greater distances.

I have played with a large number of prop angles (pitch) and moment of inertia values. Only slight performance differences seem to be obtainable which is just not real world. The best combo I have come up with still requires a compromise of a 10% scalar increase in the cfg file to get documented aircraft performance. Again maybe it is me, but after all the work I have put into it I come back to the conclusion that something in CFS2 flight programming algorithms is too generalized.

I have the performance tweaked to the point where only a negligible amount of added fuel is needed. see below

So here are what my final performance figures for both the G and J model P-38's are:

P-38G
Distance at 275 mph cruise is 600 miles. (exact distance)
Distance at 195 mph cruise is 1179 miles (should be 800 miles)
Top speed at 25000 feet is 395 mph (exact speed military power)
Top speed at S/L is 362 mph (22 mph too high)
RPM spool up time to 10% both engines is 14 seconds from engine #1 start (original was 35 seconds)
Added 4 gallons of fuel

---------------------------------------------------------------------------------------------------------

P-38J


Distance at 275 mph cruise is 840 miles. (exact distance)
Distance at 195 mph cruise is 1630 miles (should be 1175 miles)
Top speed at 25000 feet is 408 mph (exact speed military power)
Top speed at S/L is 362 mph (12 mph too high)
RPM spool up time to 10% both engines is 14 seconds from engine #1 start (original was 35 seconds)
Added 28 gallons of fuel

----------------------------------------------------------------------------------------------------------
The small amount of added fuel and its weight have zero effect on performance.

Other performance improvement

- Engine RPM spool up to 10% is now 14 seconds from engine 1 start. (originally 35 seconds)
- Player can take off once engines are equalized at 10% without wing men running over him.
- Wing men can stay in formation at full throttle top speed.
- AI aggression performance much enhanced.
- AI land, taxi off runway and stop.

Unless someone has a solution for the negative cross effect between top speed at high and low altitude and distance performance cruise settings, I am at a loss. I am at the point of accepting the compromise and go with the top speed at critical altitude and maximum cruise distance as these most affect the campaign game play, and live with the too high values at the other end.

[Ivan]

Hello Captain Kurt,

I would do some local testing but I don't happen to have a current installation of CFS2 at the moment.
I lost two computers to MB failure and HDD failure about a year ago and haven't recovered all the software I used regularly much less the software I DIDN'T use regularly.

What are the RPM, Manifold Pressure, and Altitude for the early (P-38G) and late (P-38J) Lightning to achieve the same cruising speeds? The early Lightning was a lower drag aircraft without the intercoolers under the engines and tended to cruise faster with the same throttle settings. This was one of the problems I had when working with the P-38F and P-38J AIR files. There wasn't a noticeable difference at cruise power with my AIR files.

I should test my own Lightnings at various cruise throttle settings to see what they doing, though I would much rather do that AFTER I have some good numbers from a manual as to what their speeds should be at cruise settings. I had not wanted to do any serious detail testing until after I had built proper Propeller Tables for the P-38 because although the stock tables that I am using now give "reasonable" straight line performance, they have some serious issues in some areas.

If you feel an absolute need to tune the Sea Level Maximum Speed without affecting the Maximum Speed at Altitude, it is possible to do this via the Compressibility Tables. I believe some of the CFS stock aircraft have some non-linearities in those tables at low speeds. This isn't really the right way to do it, but doing it via propeller tables isn't nearly as easy.

Moment of Inertia values in the Propeller Parameters won't do a thing for performance.
Messing with the pitch range if that is what you mean isn't really useful if the propeller tables are anything like what we typically find in CFS1. If you are interested in what I have tried, the "Flying Swallow" thread gives a pretty good description though there is a lot more that has been done since that hasn't been described there.
You probably already know this, but changing the way the simulator selects pitch is a matter of adjusting values in Table 512 (Propeller Power Coefficient). It takes a fair amount of calculation to get this even close to correct.

The problem with mixture in these simulators is that there really isn't a "correct" way to simulate how things were done.
In real life, a pilot would just use Auto-Rich in combat and Auto-Lean for economical cruise. A smart pilot might be able to get a bit better performance by doing some manual tuning, but if we allow that in the AIR file, then we also have to manage the mixture corrections for altitude changes and most countries had carburetors that could handle that situation.
I haven't yet found a way to have both manual mixture and automatic corrections for altitude but I am working on something to address that situation, though I still haven't figured out how I want everything to work yet.

- Ivan.

[Captain Kurt]

Hi Ivan

Max cruise engine settings are almost identical
275 mph true air speed
25,000 feet altitude
28 Hg for the G, 29 Hg for the J
2950 RPM for the G
3000 RPM for the J

These are with CFS2 cockpit gauges.

I freely admit I'm and idiot and have no idea how to calculate a compressibility or propeller power coefficient. For that matter, I don't know which table(s) relate to compressibility. I have discovered that the MOI of the prop does have noticeable effect on thrust by trying different numbers (basically just hacking at it plugging in numbers from the stockers and other aircraft). Sadly mostly the differences I tried seemed to be degrading performance, not improving it. But again I have no ability to calculate an appropriate value or how to accurately compare differences. I ended up just using the best test result number but as I said it still is falling short of good performance.

BTW do you have a link for the Flying Swallow thread? I'd like to read through it but I'm not finding it via the forum search.

[Ivan]

Hello Captain Kurt,

I am a bit surprised by the settings you are using for range testing.
Attached is the SEFC for the P-38J.
As you can see, although the "Rated Power" is also a Max Continuous setting, it is not going to work in terms of getting optimum range because it is using Auto-Rich mixture.
It is also at a much lower RPM than what you are using: 2600 RPM.

The actual Maximum Cruise is at 2300 RPM and 35 Inches Hg and Auto-Lean mixture.

The Compressibility Table can be found at Record 430.
The Nose Tuck Table is found at Record 433 and is quite relevant for the P-38.

The "Flying Swallow" thread about Ki 61-Id Hien can be found here:
http://www.sim-outhouse.com/sohforum...Flying-Swallow

There is a lot of stray stuff if what you are looking for is propeller tables.
I was also doing a bit of work with propeller tables for the P-40N recently until I got stuck.
I am debating adjusting "reality" a bit for the curves in order to lessen the effects of interpolation.
The problem with a lot of the monologues in the CFS1 forums is that when I make a serious mistake, no one ever calls me on it.
(Actually Smilo has a couple times, but he is very polite about it.)
If you see a thread stop for a long time, I myself may have noticed a serious mistake.

- Ivan.

[Captain Kurt]

Hi Ivan

You've hit the screw on the head (see below) It's another CFS2 compromise. I'm aware of the CFS2 engine settings vs the real world. However as someone else said recently "It's a combat sim after all". Players will most often simply use throttle to adjust their speed and not go through the additional steps of adjusting RPM and Mixture as well, especially in missions, campaigns and quick combat. CFS2 throttle control by itself uses the max RPM stated in the air file and adjusts the manifold pressure up or down. It doesn't touch the mixture setting. On the keyboard that is by 10% increments unless the plus / minus keys are also used.

It is still possible to use throttle, manifold pressure, and mixture controls if the player is flying the plane "by hand" and not using the warp feature to cover long distances. However, when the warp feature is used, the game evidently uses the speed that is set to go to the next waypoint and the variable Hg/max RPM speed control is the setting it uses to calculate fuel burn. That may be why fuel burn while warping can be so high and reported aircraft range not being met.

What I am trying to accomplish is the closest match to performance in a mission / campaign environment where the warp feature is used consistently. I am attempting to match how the game controls the engine with reported speed and distance performance, so that it works in the CFS2 combat sim environment. At the same time, the plane needs to perform realistically while the player IS flying the airplane "by hand". A nice conundrum.

If this was for a non-combat sim then the scenario would be totally different and the real world engine control settings would be paramount.

So far I have top speed at altitude and reported max cruise distance solved when warp is used, but not at S/L speed or for "long range" distance. Those are overkill, but not especially important in game play so I'll stop fighting those for now. But BTW when I do fly the plane by hand and adjust throttle, manifold pressure, and mixture to the handbook settings, the S/L top speed comes close to matching reported speeds. I haven't taken the hours necessary to test how that affects the max distance possible.

Next I need to adjust control responses closer to reported performance than what I have now. That will involve more compromises. The P-38 was an energy fighter using speed and "boom and zoom" fighting tactics. Of course (sarcasm here) CFS2 game programming only uses turning dogfighting tactics for the AI. It thinks every plane is a Zero. In order for the AI P-38s to not simply get slaughtered it will need some more maneuverability than the real plane had because the AI won't exploit the strengths of the real plane. Another nice conundrum.

I guess you could say I have to use a hammer to drive a screw, but the screw still needs to get installed.

[Ivan]

Hello Captain Kurt,

I do understand where you are coming from. These are all "Combat Flight Simulators".
My own approach is a little different though which is why I am still in CFS1.
I look at this as an experimental lab for flight dynamics testing and even after nearly 20 years, I find that I am still learning new things that can be done with an ancient simulator.
For the most part, there is very little reason that one cannot get VERY close to actual flight performance with the exception of handling multi-speed superchargers and terminal dive performance because of propeller table limits.

I personally have not flown more than about a dozen mission over the years, so I don't really know the differences that are encountered.

As for the P-38 being an energy fighter, it did have the engine power to give advantages there, but it could also be flown as a turn fighter with the exception of a rather mediocre roll response. With the Fowler Flaps out, it could turn pretty well and with differential engine power, it could swap ends much faster than one might expect. Both engines were critical engines on the P-38.

- Ivan.

[Captain Kurt]

Hi Ivan

From what I read that's true to an extent. Sounds like it also depended on the mark. Earlier models had a larger turn radius and slower roll rate than the later J model and once the boosted ailerons were installed it truly did become a good all round fighter. One trick pilots of the earlier models reportedly used if they got into a turning fight was to drop 10 degrees of flaps. Made them slower but increased the rate of turn.

Contrarily I also read some accounts given by Luftwaffe pilots who had fought the P-38s in Italy, who said their Bf109s were out-turned by the P-38s. Of course they were likely flying R6 trop versions which had a somewhat lower performance than a clean 109. And pilot skill is always a factor. But still, that contradicts what you read elsewhere about the P-38s performance. In the end it's all pretty anecdotal so I guess a guy can pick what he thinks is credible. As we know, even the official test reports don't agree from report to report.

Are we having fun yet?

[Ivan]

Hello Captain Kurt,

From the accounts I was reading recently in "America's Hundred Thousand", by Francis Dean, the low speed roll rate of the P-38 was never particularly good and the lag before the aircraft rolled through the first 10 degrees was particularly disconcerting.
This was from pilot accounts gathered during the fighter conferences I believe in 1943 and 1944.

The Fowler Flaps on the P-38 came pretty much straight out to increase wing area in the first few degrees unlike Gouge Flaps which come down immediately, so there wasn't that much drag for the first few degrees.

As for turn performance, it matters a lot how fast the two fighters are flying at. Then again, the stall speed of a P-38 isn't greatly different from a Spitfire which is amazing considering the size and weight differences.

Glad things are working out.

- Ivan.