FM Tweaking Tutorial--anyone interested?

[Herr Prop-Wasche 7]

You are right, hairyspin. However, while you can't use Airwrench to save certain information while in evaluation mode, you can always use it to observe what happens when you make any changes to the .cfg file.

 

You are also right about AirEdit being an airfile editor. While the .cfg file and the .air file both have entries that control identical variables, each file also has unique variables. For instance, using AirEdit, you can edit a variable in the .air file called "Cdo_fuselage_drag_coeff." This variable does not seem to be available in the .cfg file. Might this be key to controlling dive speed?

[hairyspin 1]

I would suggest looking at the .cfg file of a roughly similar plane to yours and using that as a template.

 

 

A sound and sensible notion!

 

Regarding .air and .cfg files, there's a degree of overlap between the two, but which file takes priority in the flight model varies from sim to sim. FSX (sorry!) takes most of its data from the .cfg, early versions of FS and CFS give priority to the .air file. I'm no expert, just an interested meddler who's been reading what he can find on the topic for a few years, so I expect the dive speed is controlled by a number of variables, possibly quite a lot of 'em!

 

I think I said this was almost another hobby in itself ....

Edited April 27, 2010 by hairyspin

[Herr Prop-Wasche 7]

Hi Dave!

 

Odd that you can fly in QC but not in campaign. I have found that the altered planes either fly in both or not at all! Are you saving your settings using Airwrench or are you making your changes directly in the .cfg file? I found that doing only the latter will work.

[Bullethead 12]

Odd that Airwrench is only allowing you to edit the contact points, BH. What plane were you working on, or was it one of your own creations?

 

I wasn't working on any plane. The Mudpond site where you download Air Wrench has this caveat splattered all over it concerning the demo. Until you buy the full version, the ONLY thing that it will save is contact points.

 

So, I took the site at its word and haven't tried to do save anything. Once I get to the point of flighttesting, I'll buy the full version and see how it works. In the meantime, I find the demo quite useful for determing the best L/D gliding speeds for OFF planes :).

[Guest British_eh]

Interesting note olde chap. As a proponent of the Dr.I, I am happy to say that the Dev's did a great job, but with their excellent patches, etc, the Dr.I FM now, is olde hat. I have quite a few compelling reports, including one from Mikael Carlson, owner/pilot of one of the only Dr.I with an original 9 -cylinder Le Rhône 110 hp rotary engine which truly replicates the handling characteristics of the original craft. Unlike the OFF model, it would be somewhat difficult to put the Dr into a spin, and 1 1/2 rotations to pull it out. Near a stall, the aircraft hangs and hangs, somewhat like the Tripe, only longer. Perhaps in P4 they will address many issues. There is so much we want , and only so many Dev's, to deal with the herculean task of CFS3 code, etc.

 

Would be nice if there was some mod in the interim.

 

 

Cheers,

 

British_eh

 

 

Besides, everyone knows that the DR1 is "porked!

 

Regards,

Royce

Edited April 29, 2010 by British_eh

[Duce Lewis 3]

Besides, everyone knows that the DR1 is "porked!

 

Good un Brit!

I know you've been flying the Dr1 for quite some time

Nobody knows her better than you, I'm sure

 

I tried her a couple times

How do you see past all that wingage I wonder?

 

I doubt we'll ever see perfect FM as long as we're tied to the cfs3 engine

Just too many limitations

 

I like what Herr Prop-Wasche is doing

Maybe we can tweak the best outta her

[Herr Prop-Wasche 7]

Looks like my suspicions were correct about the "Cdo_fuselage_drag_coeff" variable. Currently, this variable for the Sopwith Triplane is set to .60. I raised it to .135 and observed a reduction in the dive speed from 270 mph down to approximately 190 mph. OTOH, it also lowered the climb rate, so more tinkering will be necessary.

 

Curiously, the two triplanes are near the front of the pack in terms of lowest drag coefficients. I would think they should be higher, given they have an extra set of wings with braces and struts to slow things down. I think the next thing I will try is to lower the drag on the N28 and see if I can overstress her in a dive.

[Fortiesboy 3]

Hi Dave!

 

Odd that you can fly in QC but not in campaign. I have found that the altered planes either fly in both or not at all! Are you saving your settings using Airwrench or are you making your changes directly in the .cfg file? I found that doing only the latter will work.

 

I have played around a bit in Air edit, but I have made my changes just using the .cfg file.

It seems that when an important change is made, OFF realizes that and then can see that it is not a "certified craft"- In fact when you go to select it to fly , it doesn't load and a message comes up to say it is an uncertified craft. This happens in campaign and ordinary QC, but in CFS3 QC you can get it to fly.

[Herr Prop-Wasche 7]

'allo Dave.

 

Try this. Edit the .cfg file for the plane you are working on. Save the .cfg file. THEN, start Airwrench and load the .air file for that aircraft. Briefly look around to see if your changes "took" and then exit Airwrench. This should update the .air file with the new .cfg setting. If you don't start Airwrench and load the changed file, the .air file does not get updated and you get a "mismatch" or whatever and OFF thinks it is an uncertified craft. This is the process I have been using lately with success in both QC and campaign.

 

Let me know if this works for you or not.

Edited April 29, 2010 by Herr Prop-Wasche

[Guest British_eh]

Great stuff here HPW,

 

The Dr.I could have flown quite nicely without the middle wing, which did add drag. JFM seems quite knowledgeable on the craft, and I believe his new book has a lot of technical data. I am unsure of the dive speed of the Sop Triplane, but I have researched what appears to be quotable dive speed of the Sopwith Camel, at 120 mph, and the Dr.I at 130 mph. Pol had mentioned that there are different parameters for "dive"speeds, as in the discussion about the Albatross's dive speed. My take is that it is power on, and straight down. I am unslure of any other way to think about it. Dive speeds are listed using the V designation, with V - small D as the design speed and V - small DF as the demonstrated dive speed. The Drag are of the Camel comes in at 0.81 M sq and the Zero lift drag coeficient at 0.0378. Surpisingly the Dr.I comes in at 0.62 M sq and 0.0323, less than the Camel, and thus the aircraft that I would want to be in, fighting a Camel! ( or any other foe of that time frame)

 

 

Cheers,

 

British_eh

 

 

Herr Prop-Wasche Looks like my suspicions were correct about the "Cdo_fuselage_drag_coeff" variable. Currently, this variable for the Sopwith Triplane is set to .60. I raised it to .135 and observed a reduction in the dive speed from 270 mph down to approximately 190 mph. OTOH, it also lowered the climb rate, so more tinkering will be necessary.

 

Curiously, the two triplanes are near the front of the pack in terms of lowest drag coefficients. I would think they should be higher, given they have an extra set of wings with braces and struts to slow things down. I think the next thing I will try is to lower the drag on the N28 and see if I can overstress her in a dive.

[Herr Prop-Wasche 7]

Interesting information! The DrI's lower drag coefficient probably comes from its shorter wings than the Camel's, giving it an overall lower surface area. Off the top of my head, drag coeff. is probably affected by some of the following factors: length of wing, wing chord, number of wings (all 3 contributing to overall surface area), "straight" vs. tapered wing, thickness of wing, and dihedral. Can you think of any others?

 

Do you know of any readily available sources that might give me some of this information, especially data about zero lift drag? In the game, the aerocraft with the lowest drag coefficients are (with lowest listed first) the Sop. Pup, Triplane, Camel, and Se5a viper, followed by the DrI, Spad XIII, and the Bristol Fighter. The planes with the highest drag coefficients are (from lower to highest) the Nieuport 17, Fokker DVIIF!!, Halberstadt, Nieuport 16, Nieuport 11, and Fokker EIII.

 

What do you think? I'm not sure I agree with the ordering of the planes in the game in terms of drag. However, this is only off the top of my head, as I do not have any hard data in front of me to compare.

[Herr Prop-Wasche 7]

Pol is probably referring to several factors that most likely have an affect on dive speed, including maximum speed, cruising speed, zero lift drag, aircraft weight, etc.

 

Drag seems to have some of the most effect on a number of factors. It not only effects dive speed, but also level speed, climb rate, and even ceiling, as demonstrated when I lowered the drag coeff. of the Alb DIII early. I was able to attain an altitude of 19,000 ft.--a full 1,000 ft. more than it should be capable of!

 

Almost too many factors to play with.

[Herr Prop-Wasche 7]

A small update...

 

Only after starting the tutorial did I realize how little I actually knew about all of the variables in the FM. Thus, why I haven't posted anything in this thread recently. Since then, I have done a little more studying and experimenting and feel that I have a much better grasp on things.

 

A few of the things I have discovered: In case of an inconsistency between the .cfg file and the .air file, the values in the .air file take precedence. Previously, I thought it was the opposite. I discovered this after trying in vain to change the speed at which an aircraft takes damage in a dive by altering the .cfg file. Instead, I found you have to change an entry in the .air file (which is a hex file) to make this variable take effect. Now, I can make the Alb DIII start to take damage from over-speeding in a dive. Also, taking damage in a dive is independent from taking damage due to high G maneuvers, so I think I can more realistically portray the Alb's tendency to shed a wing in a dive while making it less likely to shed a wing in a high G turn! Good news for Olham!

 

I am also working on re-working the dive rates for various craft. This can be changed by altering another .air file entry for parasitic drag. Many of the aircraft in BHaH seem to have incorrect dive speeds. I have already mentioned the Sop. Tripe, for example.

 

For those who are interested, here is a list of some of the other variables that can be altered in the .cfg and .air files: engine HP (among many others), center of gravity, stall speed, induced drag, positive and negative G tolerance, max airspeed, cruising speed, maximum dive speed, climb rate, roll rate, turn radius, crash "tolerance" for fuselage/wings/tail, among many others. As you can see, lots of variables to play with and keep one busy for awhile!

[Herr Prop-Wasche 7]

Another brief update, of sorts...

 

Due to an unfortunate series of events, I have been without home internet access for the last several days. This situation is likely to continue for the next week or so, so I will not be able to post very often for the forseeable future. However, I am still working on the FM and making (slow) progress. I will also try to post a second installment of the FM tweaking tutorial within the next couple of days.

 

Cheers,

 

HPW

[Duce Lewis 3]

Another brief update, of sorts...

 

Due to an unfortunate series of events, I have been without home internet access for the last several days. This situation is likely to continue for the next week or so, so I will not be able to post very often for the forseeable future. However, I am still working on the FM and making (slow) progress. I will also try to post a second installment of the FM tweaking tutorial within the next couple of days.

 

Cheers,

 

HPW

Excellent HPW!

Great work M8

[Herr Prop-Wasche 7]

Still experiencing internet access problems, so my access to the forum is limited. Still, I have some news to report. I have finished making some FM tweaks to the Albatross DII and DIII series and to the N28 and plan to upload the files later tonight or tomorrow. Look for them in this thread or in the download section.

[Herr Prop-Wasche 7]

Finally, the FM tweaks are done!

 

Attached, please find two zip files. One is for FM tweaks to the Albatross DII, DIII early, and DIII late; and the other for the N28.

 

The Albatross files are in reply to a request from Olham and others that there is not much of a difference or improvement from the DII to the DIII series of aircraft. I tend to agree, since the Alb DIII series, with the exception of its wing shedding problems (represented in the FM tweak!), was recognized as one of the best aeroplanes of the mid-war period. Likewise, the N28 files are an effort to somewhat improve the performance of the N28.

 

Although there are many changes to the FM's of each aircraft, the main ones have to do with parasitic drag. Parasitic drag affects the performance of each aircraft in many ways, including climb rate, cruise speed, stall, and aircraft ceiling. IMO, the parasitic drag of many of the aircraft in OFF is not quite correct. For example, the Bristol Fighter, with the heaviest weight and greatest wing area of any plane in OFF, has one of the lightest drag settings. In contrast, the Nieuports have drag settings that are sometimes two or even three times the drag of the F2b. As a consequence, other settings have had to be changed in OFF to get the aircraft to perform according to historical patterns. I have tried, with my first FM tweaking attempt, to remedy some of these problems with a few aircraft. I will leave it to you to decide if I have been successful. So far, the tweaks are only to be used in QC or CFS3 QC. If most of you approve of the changes, I will make another update available for the campaign.

 

Here are the instructions for installing the FM mod: 1) Open your aircraft\AlbDII_QC1, Alb_DIII_QC1, Alb_DIII_early, and Nieuport_N28_QC1 folders. 2) BACKUP OR RENAME THE *.AIR AND *.CFG FILES IN EACH OF THESE FOLDERS. 3) Unzip the zip file(s) into your OBD Software\aircraft folder. If the program asks to overwrite files, click NO, and make sure you have renamed or backed-up the correct files. 4) Play a QC battle. Enjoy!

 

Please be sure to inform me of what you like and dislike about each of these changes. FM tweaking is hard work (my sympathies to OFF), and if I do not hear from anyone, I won't be able to continue making FM tweaks to the aircraft that may need some adjustments. I hope you enjoy the files, and recognize the labor of love that these changes represent. Thank you!

 

 

 

Edited May 15, 2010 by Herr Prop-Wasche

[Duce Lewis 3]

Finally, the FM tweaks are done!

 

Attached, please find two zip files. One is for FM tweaks to the Albatross DII, DIII early, and DIII late; and the other for the N28.

Great job HPW!

 

I'm away right now but I'll definitly check this out

I'll report back results in a day or three

[DukeIronHand 8]

Thanks for your hard work - will give these modified FM's a spin.

[Herr Prop-Wasche 7]

I will be looking forward to your reports! Don't be shy about sharing the bad as well as the good.

 

Here is a question for everyone to ponder: do you think the planes in OFF are, in general, easier to fly than in real life, about the same as in real life, harder to fly in real life, or easier or harder depending on the particular airplane? Regardless of whether you think the planes are too easy to fly or too difficult to fly, what aspect, in general for most planes, is most out of line with your expectations: speed, turning ability, climb, stall, etc.? I am curious about others opinions for guidance about what direction to take with my FM tweaking.

[Herr Prop-Wasche 7]

Bumped in effort to restore thread.

[Duce Lewis 3]

HPW,

 

Sorry for not getting back sooner

I was away a bit

 

I took the N.28 up for a test and observed the following

Acceleration is better

She'll manuever better too, due to reduced drag

 

At 5k feet:

I could perform a barrel roll but it was slow and a little sloppy

She'll loop about the same but after 3 loops the upper wing failed

I tried a tight turn , laying her on her side and pulling with elevators ...but the upper wing failed again

Nose down brings better performance

 

At 10k feet:

Barrel rolls are very sluggish and not worth the effort

Loops bring a stall and roll off at the top

 

Biggest improvement was in climb (though I didn't try climbing her before the mod)

I could pull the nose up 30^o and she'll still climb OK

 

Testing brought forth a few observations/suggestions:

1. I was able to fly OFF QC but it was necessary to create a USA Pilot with a Spad 13 as default A/C

 

2. Testing will be better and more accurate if we create a standard test routine

Altitudes, Manuevers, Climd, Dive, etc.

 

3. Documenting FM Mods and Testing in txt files would secure data

 

Cheers,

[Herr Prop-Wasche 7]

Thanks for your feedback, Duce.

 

The wing failures can probably be cured by adjusting the maximum_g value. I will try that tonight and get back to you. Keep in mind differences in "stick handling" of course. Although I put the N28 through some turns, I did not go out of my way to stress the airframe in combat conditions. On the other hand, I don't want the plane to become indestructible.

 

Roll was set pretty low in the original FM. I didn't increase the floor, but instead increased the range of values. I can probably increase the base roll rate a little to see what effect it has on performance.

 

I agree about the testing. In general, I try to find the maximum ceiling and the cruising speed at 500 feet. Climb, turns, and dive are done pretty much by "feel' as I don't have a decent stopwatch right now.

 

(I have just recently discovered a complicating factor in setting the drag in the air file. Turns out that in addition to adjusting the climb rate, ceiling, and dive speed, it also has a major impact on the cruising speed. This leads to strange results, such as a Triplanes with high cruising speeds and very high dive speeds. In fact, all "fast" planes have relatively low drag compared to slower planes. The only way to adjust the cruising speed independent of the drag factor is to adjust either a value called "power_scaler" in the engine performance section or "thrust_scaler" in the propeller section. However, I don't think these variables have anything to do with real world physics and are instead "cheat" values used in CFS3 to control for unrealistic performance specifications--if that makes any sense. I'll have to test changes to both of these variables to be more sure.)

[Duce Lewis 3]

HPW,

MS probably designed a set method for creating FM's

There are likely more independent Base Values

Then those that interact for tweaking

 

I think MS released an SDK foe modders of cfs3

Maybe there's some recommended methodology there

Dropping a line in a cfs3 forum might get us some good recommendations too

 

I'd really like to create some mods myself too

But unfortunately, I'll be very tied up for the next couple weeks

Memorial Day is approaching (May 31st) and my Dad and I have to prepare to place flags on veteran's graves

 

I'm sure I'll manage to get some testing time in

...but please be patient with me if I take a little time to respond

But I promise I'll get back

[Herr Prop-Wasche 7]

No problem and take your time, Duce. It's nice what you and your Dad are doing and puts in perspective the real meaning of "Memorial Day." Olham is also helping with the FM and he is, of course, on vacation until the end of the month.

 

I've checked into the CFS3 SDK and although it's helpful in some areas, it only mentions a few critical files and variables. Most are left for us to guess at. I don't have a current copy of CFS3 on my computer. If you do, you might look at the cfg files for a few of the WWII airplanes to see how they differ from OFF's WWI planes. Might be interesting to look at.

Edited May 19, 2010 by Herr Prop-Wasche