Combined FM and AI Weight Mod Complete!

[+Polovski 414]

It took several 1000 flight tests to get player flight feel for every craft, AI flight, AI takes offs, autopilot, with and without loads, etc all working well. Crazy amount of time that took - all I can say is good luck!

[Herr Prop-Wasche 7]

I did some checking into the .cfg files and with Airwrench to analyze the changes in speed and climb to the Halberstadt caused by changing the empty weight values in the .cfg file and think I have found some interesting results.

 

First, the difference between the empty weight and the maximum weight for the Halb is quite large: 649 lbs for an early war scout. Most scouts in the game have a smaller difference between these weights. Then, the question becomes, which weight is incorrect? Or, are both incorrect? Lima noted that his source from G & T says the fully loaded weight (including ammo and gear)? of the Halb is 1696, rather than 1793 as it is now in the game. In any event, Lima chose a value of 1713 for his empty weight: 1793-80 lbs for fuel.

 

According to Airwrench, a plane flown at an empty weight of 1713 has an approximate top speed of 89.3 mph and flies approximately 81.5 at sea level, with a climb rate of only 241 ft. per minute. Note the extremely poor climb rate.

 

According to Airwrench, a plane flown at an empty weight of 1563 (150 lbs lighter) has an approximate top speed of 90.8 mph and flies approximately 83.75 at sea level, with a climb rate of 352 ft. per minute. Even 150 lbs lighter, the climb rate is still very poor.

 

By contrast, if I use an empty weight of 1378 based on an adjusted Ramp weight (pilot and fuel, less ammo and extra gear) minus 50 lbs for fuel consumed, what does Airwrench report?

(I arrived at this figure by taking the OFF Ramp weight of 1428 and subtracting 50 lbs for fuel expended). According to Airwrench, a plane flown by the AI at an empty weight of 1378 has an approximate top speed of 92.1 mph and flies approximately 86.0 at sea level, with a climb rate of 508 ft. per minute. This is more than twice the climb rate using an empty weight figure of 1713 lbs and also seems to be somewhat closer to historical figures, I believe.

 

Of course, I can't be sure how accurate Airwrench is in reporting these figures--especially climb rate. However, assuming that with current figures the Halb can only climb 241 ft. per minute at top speed, it may explain why some planes are having trouble clearing hills and trees at the end of runways.

[Herr Prop-Wasche 7]

I have yet begun to tinker!

 

Full speed ahead!

 

Damn the torpedoes!

 

Etc.

 

Actually, I have done enough tinkering with the FM for awhile, I think, except to clean up errors. I've got to get in fighting shape for P4 or OFF2 or whatever you have that will soon grace our greedy little hands--I hope it is soon!

[33LIMA 972]

I did some checking into the .cfg files and with Airwrench to analyze the changes in speed and climb to the Halberstadt caused by changing the empty weight values in the .cfg file and think I have found some interesting results.

 

First, the difference between the empty weight and the maximum weight for the Halb is quite large: 649 lbs for an early war scout. Most scouts in the game have a smaller difference between these weights. Then, the question becomes, which weight is incorrect? Or, are both incorrect? Lima noted that his source from G & T says the fully loaded weight (including ammo and gear)? of the Halb is 1696, rather than 1793 as it is now in the game. In any event, Lima chose a value of 1713 for his empty weight: 1793-80 lbs for fuel.

 

According to Airwrench, a plane flown at an empty weight of 1713 has an approximate top speed of 89.3 mph and flies approximately 81.5 at sea level, with a climb rate of only 241 ft. per minute. Note the extremely poor climb rate.

 

According to Airwrench, a plane flown at an empty weight of 1563 (150 lbs lighter) has an approximate top speed of 90.8 mph and flies approximately 83.75 at sea level, with a climb rate of 352 ft. per minute. Even 150 lbs lighter, the climb rate is still very poor.

 

By contrast, if I use an empty weight of 1378 based on an adjusted Ramp weight (pilot and fuel, less ammo and extra gear) minus 50 lbs for fuel consumed, what does Airwrench report?

(I arrived at this figure by taking the OFF Ramp weight of 1428 and subtracting 50 lbs for fuel expended). According to Airwrench, a plane flown by the AI at an empty weight of 1378 has an approximate top speed of 92.1 mph and flies approximately 86.0 at sea level, with a climb rate of 508 ft. per minute. This is more than twice the climb rate using an empty weight figure of 1713 lbs and also seems to be somewhat closer to historical figures, I believe.

 

Of course, I can't be sure how accurate Airwrench is in reporting these figures--especially climb rate. However, assuming that with current figures the Halb can only climb 241 ft. per minute at top speed, it may explain why some planes are having trouble clearing hills and trees at the end of runways.

 

Very interesting! Do these Airwrench figures also apply to the player-flown Halberstadt DII, too - if so, I wonder why it doesn't appear to suffer like the AI, given the player's plane does fly with all the same extra weight.

 

Every Halb DII campaign mission flown so far, my own flight's AI planes take off and climb sluggishly but ok and clear trees etc with no probs, but the AI-led flight goes into the same trees like lemmings. I thought perhaps I'd find my flight uses the SQ planes while the AI-led flight (the one with the aces) would use the AC planes but if there is a difference in their respective flight models, I don't see it in the .cfg files.

 

Some more testing needed. The extremely nose-heavy Halb DII I'd like to try to correct.

[Herr Prop-Wasche 7]

My belief is that, even in the campaign, the human player flies the QC version of the aircraft. Other members of the human player's flight, I believe, fly using either the QC FM or the Sqd FM, but I'm not sure which. Everybody else, including other flights and squadrons, both friendly and enemy, use the FM's in the remaining versions--SQ1 through SQ4, and AC1 through AC5. This may explain why your flight gets over the trees but the AI led flight can't clear them--you and the members of your flight are using the "lighter" QC or Sqd versions, while the AI led flight is using one of the SQ or AC versions, which utilize the heavier weight--I think.

 

That's why I'm thinking we might need to lower the empty weight of the versions that use either the SQ or AC versions of each aircraft. More testing may be needed to determine just how much weight needs to be taken off and for what planes in addition to the Halb and maybe some of the two-seaters.

 

Feedback is needed from other players using Lima's mod either with or without my mod. If you are using these mods in campaign, please report any AI take off difficulties to Lima or me. Thanks!

[+Polovski 414]

Tip - those 1000s of hours testing in sim are because you need to test actual climb rates etc etc in sim, not what Airwrench will report.

[cptroyce 0]

Prop -

 

As Pol suggested above and as I was going to ask..why don't you use the historical weights, RoC etc. as values? Also the other specs for many models, were at one time and probably still are, available from the various web sites around the net..CoG etc.

Perhaps if you plug in the historical values in Air Wrench and work backwards or forwards from those values, you can get an accurate FM for the Alb or any other models that you are not satisfied with.

 

But I have to think, that with all the work you have done on the FM and DM for OFF, you have probably already done that.

 

Regards,

Royce

 

 

[carrick58 23]

Not sure which mod you're referring to, carrick but yes they all involve over-writing some files of planes in the folder OBD Software/CFSWW1 Over Flanders Fields/Aircraft. But use JSGME and install the mods files into the OBD Software/CFSWW1 Over Flanders Fields/MOD/ folder rather than over-writing the originals, so you can back it out.

 

I've done some more testing and there is a strange problem with the Halb DII.

 

So far, other planes seem ok, But Flying a Halb camopaign with Jasta 1 out of Bertincourt which has a wood with slightly rising ground at the far end of the field, I'm seeing probs. My own flight is consistently taking off ok. But 'Flight 1''s lead plane is crashing into the trees and his wingmen are then crashing too - it's like they're blindly following him in, rather than suffering from the same problem. This has happened twise now. All the planes in my flight got airborne ok, whether I flew off first or let some of them go first. Both times, Flight 1 leader flew into the trees and all his wingies followed him. Need to repeat this a few more times to confirm it's definitely the pattern.

 

Hard to be sure, but with the Halb, I don't think the extra weight is the issue, not mainly anyway. There's something else going on. From what I've seen, OFF planes can take a huge increase in weight before they can't take off; the Empty Weight Mod added over 1000 lbs to the stock AI RE8 and it still takes off and flies ok.

 

Hard to say without a line-by-line comparison but the .cfg files for all the AI-flown Halberstadts look the same so it's a bit of a mystery.

 

One clue is that when I'm flying the DII, it's EXTREMELY nose-heavy. Much more so than any other OFF plane I've flown. Seems strange, given the wings - and with that possibly the centre of lift - are so far forward - she looks like she should be tail-heavy if anything, tho that's just a wild guess. I can't find anything about the real plane's CoG or how the Vintage Aviator DIV replica flies.

 

I've reduced the added weight by 97 lbs to correct the stock OFF figure but no difference.

 

So - thus far the only clues I've got are (i) the 'other' flight, not the player's one, being affected and (ii) the Halb's extreme nose-heaviness. Will do a bit more experimenting, to see if this is affecting some other planes too.

 

HPW - can you advise, what figures need tweaking, to push the Halb DII's CoG aft a bit, in the .cfg and .air files? I'd like to try that and see what difference it makes, if any.

 

OK Thanks, will give a try

[33LIMA 972]

Thanks for the tip Pol; I was inclined to distrust the reported Airwrench figures as they might explain the AI problems but don't seem to apply to the player-flown plane (and perhaps its flightmates) as well.

 

HPW, I had come to the same 'who flies what' conclusion altho had come to doubt that your flight-mates flew the same, lighter plane as the player, since (with the AI empty weight mod enabled) my flightmates seem to take a lot longer to catch me up after settling on course post takeoff and (ii) at least one player flying with an AI flight leader reported being able to catch up and keep up better. anyway, I think that some difference does exist between the AI-led flight and the player-led AI, and that understanding this factor is what I should concentrate on first, before concluding that adjusting relative weight is the solution.

 

Two other pieces of evidence from yesterday, partly contradictory:

 

(i) flew a campaign mission with Pups - another very low-powered plane - and saw a similar differential behaviour - the AI-led flight just about made it over the trees at the airfields end, my flight's AI had no probs; and

 

(ii) then i remembered some time back, flying a Strutter 3 Naval Wing mission in OFF's spectacular hilly terrain - this was long BEFORE the AI empty weight mod - we came across Fokkers just above which proved hard to catch. However, at one point, their leader flew into a hill - he may or may not have been damaged, can't recall - but his wingmen then flew into the same hill, seemingly quite gradually and deliberately, crashing one after the other. They had plenty of time to pul up, but 'follow your leader' seemed to take priority, at least until it was too late.

 

Putting all this together, I think an underlying 'Lemming factor' of some sort is at least part of the problem here - the AI-led flight mates are crashing, or staying low, because that's what their leader does too.

 

Next line of investigation is this 'engine boost' thing - which planes or plane models have it, which don't, and what difference that makes. I recall an OFF post or FAQ mentioning this was not the same as WEP but was parcelled out selectively to simulate engine variations etc. The OFF Halb's all already seem to have 'turbocharged=1' and 'emergency_boost_type=0'; from a quick look, at the .cfg files only, the pattern seems to be all planes have the latter setting but some have turbocharged at 0 (Camel, Bristol Scout), others at 1 (Pup, Halb). I want to check if it's applied selectively within a different aircraft (so far i think it isn't) and whether or not the planes with the same setting are the same ones with problems (unlikely but need to check).

 

Also want to check it there is any difference that applies to Ace-flown planes - on the hypothesis that it's these planes that are having some kind of FM or weight-related terrain-avoidance problem and are leading their wingies to their doom.

 

In combat, I'm finding that the AI are still quite sprightly and able to climb and/or maintain altitude, so I think reducing their extra weight would be a last resort, only if we found there was no other way to solve this differential.

 

Update:

 

(i) changing 'empty_weight_CG_position= 0.000, 0.000, 0.000' to 'empty_weight_CG_position= -1.00, 0.000, 0.000' in the aircraft.cfg file has 'cured' the Halb's great nose-heaviness - it's now in neutral trim (for pitch) at about 80% throttle, slightly tail-heavy at full throttle, which I will stick with, as I think this not only makes flying easier wihtout resort to trim (simulating the plane being rigged to fly level) but could be more realistic (slight tail-heaviness compared to great nose-heaviness).

 

(ii) the player flies the QC1 plane in quick combat but definitely flies the .sqd plane in campaigns, not QC1 - making the above edit in the QC1 file had no effect in campaign, till I made it to the .sqd file as well.

 

(iii) applying the centre of gravity change to all planes did not stop the AI-led flight hitting the trees on the slope at the end of Bertincourt aerodrome.

 

(iv) the player's flight's AI planes, which avoid the trees, may well do so because it uses the same .sqd plane files as the player (and hence ignore the AI empty weight mod's extra weight). But if (as seems likely) that is indeed so, it's all the more extraordinary that the player-led AI flight still adopts a very long takeoff run with a very shallow climb out, which means they they only just clear the trees, by a very small margin. Watching them after takeoff, they seem well able to climb quite strongly, when they 'choose' to do so.

 

(v) 'engine boost' - all the Halb's already have 'turbocharged=1' so the lack of this setting can't be what's causing the problem.

 

My preferred solution would therefore be to change any or all of AI behaviour on takeoff, leader-following or terrain avoidance, or to find some kludge (eg takeoff speed boost) which would compensate, but the chances of that seem small, so there may be no alternative to reducing the extra weight, for those planes whose AI cannot cope with it.

 

What I'd like to try is reducing rolling resistance (ie on the ground) so planes can build up speed quickly when taking off. Camera angles and the rate at which film is shot can be deceptive, and there's always the effect of the speed of the wind into which they're taking off, but I've always noticed how quickly WW1 planes (originals and replicas) seem able to get their tails up and leave the ground, and how steeply they can climb away. If I could get all OFF planes to unstick more quickly by reducing rolling resistance, I'd be quite happy; reducing takeoff crashes due to the AI having to fly at realistic weights would almost be a bonus. In First Eagles there are co-efficients associated with the landing gear which can be edited to this end, not sure about OFF but will do some digging and report back. Still think the best solution is to try to isolate and tackle any underlying issues and cut back on weight only as a last resort.

Edited December 26, 2011 by 33LIMA

[33LIMA 972]

OK I've had a look at the weight values reported by AirWrench and it looks like the thing to do is go with HPW's original suggestion to bring the AI weight up to 'ramp weight minus some fuel', not as I've done 'loaded [max] weight minus some fuel'.

 

This is because (i) Pol has confirmed FMs were designed to be right for the PLAYER'S plane (and then tested to check AI could cope) and (ii) confusingly but significantly, the weights for the loads carried by the player's plane, from the .air files as reported by AirWrench, are consistently LOWER than the correct figures. It doesn't matter whether these too-low figures are by deliberate design or not; the point is, right or wrong in themselves, they're the ones for which the FM is designed to be right, when added onto the 'empty' weight. At least, that's the current hypothesis!

 

To give an example, the stock OFF .cfg file for the Nieuport 17 has near-enough correct figures for empty and loaded ('max') weights, 825 lbs and 1280 lbs. The real plane's loads were 143 lbs for fuel & oil and 264 lbs for pilot & military load. But the .air file gives 114 lbs for fuel and 180 lbs for other loads. Added to the empty weight, this gives what it calls a 'ramp weight' of 1019 lbs.

 

I suspect it's this 'ramp weight', and not the 'max' weight, that the FM is designed around - a weight that's about 260 lbs lighter than an N17 with pilot, and a full load of fuel, oil and ammo! And about 180 lbs lighter than the AI Empty Weight mod uses, now.

 

From what I've seen, the difference this makes will vary from plane to plane; for the Halb DII, it looks like taking off c.300 lbs, partly because the Halb starts from a very low empty weight of 1144 lbs (c.100 lbs to low, per my references) and only adds 93.5 lbs for fuel and 190 for other load.

 

It's pointless speculating why (if the above hypothesis is correct) the FM was designed to be correct with weights so far below the normal loaded weights, tho I wonder if it was this factor which enabled the AI to take off successfully in P3, whereas in P2 they more often crashed and burned, like the AI-led Halbs are doing now with the empty weight mod.

 

To test this I'm going to re-work the Halb DII's AI empty weight mod to use the 'ramp weight' figure instead of the loaded ('max') one. Will report back.

Edited December 26, 2011 by 33LIMA

[Herr Prop-Wasche 7]

Thanks for all your work, Lima. I might be able to save you some time by reporting on some testing that I have also managed to do today. In short, I have found that the empty weight does matter, at least for the Halb. Planes with the a heavier empty weight of 1713 were more likely to crash into hills or trees at the end of the runway than planes that had an empty weight of 1141 or even 1373.

 

I have also worked out which "type" of folder controls who flies what type of plane:

QC planes are for Quick Battles only and do not seem to have a role in most campaign missions.

Sqd planes are the planes the human player has in the campaign.

AC planes, of course, are for any Aces in your squadron.

SQ planes are for the non-aces in your squadron.

 

I did some testing, varying the weight of QC, Sqd, AC, and SQ planes in campaign missions involving Jasta 1 Halbs. Briefly, here are my results:

 

Sqd, AC, and SQ folders all set to "light empty weight:" 0 out of 3 crashes in Ace flight, 0 out of 5 crashes in player flight.

 

Sqd, AC and SQ folders all set to "heavy weight:" 6 out of 6 crases in Ace flight, 3 out of 3 crashes in player flight. Note: my plane seemed excessively sluggish, but did not crash.

 

Sqd folder set to "light", all others set to heavy: 2 out of 5 crashed in Ace flight, 2 out of 2 crashed in player flight.

 

Sqd folder set to heavy, all others set to light: 3 out of 5 crashed in Ace flight, 0 out of 2 in player flight. Again, my plane was sluggish, but I managed (barely) to avoid crashing.

 

In each case that the Sqd folder was set to "light" I did not have trouble flying, but each time Sqd was set to heavy, my plane had a very difficult time getting altitude.

 

So, it seems that empty weight does indeed make a difference, especially for low-powered craft such as the Halb. However, I don't think the empty weight variable is entirely responsible for the difficulty the Halb has in climbing. As Lima has noted, the Halb is nose heavy. I looked at the .cfg files and found that the difference between the CoG and CoL on the Halb is -1.13 feet--double that of the next closest aircraft, the Fokker EIII and the Nieuport 16--which have a difference between the CoG and the CoL of -.57. Having the CoL that far back from the CoG does tend to make the Halb more nose heavy. On the other hand, in my FM tinkering experience, I have found that making an aircraft too tail-heavy can make the plane too prone to stalls, so the balance between CoG and CoL has to be carefully watched so as to not go too far in the other direction.

 

At this point, my proposed solution is also to change the basis for the empty weight setting to the Ramp weight rather than on the Maximum weight, at least for some aircraft. This will probably have to be combined with some slight adjustment to the CoG and CoL to make the Halb more tail heavy and thus somewhat more responsive, as well.

[Herr Prop-Wasche 7]

HPW, I had come to the same 'who flies what' conclusion altho had come to doubt that your flight-mates flew the same, lighter plane as the player, since (with the AI empty weight mod enabled) my flightmates seem to take a lot longer to catch me up after settling on course post takeoff and (ii) at least one player flying with an AI flight leader reported being able to catch up and keep up better. anyway, I think that some difference does exist between the AI-led flight and the player-led AI, and that understanding this factor is what I should concentrate on first, before concluding that adjusting relative weight is the solution.

 

You are right, your flight mates do not fly the same weight planes--see my above post.

 

Two other pieces of evidence from yesterday, partly contradictory:

 

(i) flew a campaign mission with Pups - another very low-powered plane - and saw a similar differential behaviour - the AI-led flight just about made it over the trees at the airfields end, my flight's AI had no probs; and

 

(ii) then i remembered some time back, flying a Strutter 3 Naval Wing mission in OFF's spectacular hilly terrain - this was long BEFORE the AI empty weight mod - we came across Fokkers just above which proved hard to catch. However, at one point, their leader flew into a hill - he may or may not have been damaged, can't recall - but his wingmen then flew into the same hill, seemingly quite gradually and deliberately, crashing one after the other. They had plenty of time to pul up, but 'follow your leader' seemed to take priority, at least until it was too late.

 

Putting all this together, I think an underlying 'Lemming factor' of some sort is at least part of the problem here - the AI-led flight mates are crashing, or staying low, because that's what their leader does too.

 

Yes, the AI planes do tend to follow the flight-leader, as if they were in a WWII squadron. Another carryover from CFS3, I think, is that the AI planes all tend to fly at 70-80% throttle virtually all of the time. It's as if running the engine at 100% is like "emergency power," something that WWII birds only did in combat or for short bursts, to avoid engine overheating. I bet this is another factor in the AI's poor low speed or low altitude performance.

 

Next line of investigation is this 'engine boost' thing - which planes or plane models have it, which don't, and what difference that makes. I recall an OFF post or FAQ mentioning this was not the same as WEP but was parcelled out selectively to simulate engine variations etc. The OFF Halb's all already seem to have 'turbocharged=1' and 'emergency_boost_type=0'; from a quick look, at the .cfg files only, the pattern seems to be all planes have the latter setting but some have turbocharged at 0 (Camel, Bristol Scout), others at 1 (Pup, Halb). I want to check if it's applied selectively within a different aircraft (so far i think it isn't) and whether or not the planes with the same setting are the same ones with problems (unlikely but need to check).

I don't believe that boost has much to do with this, but I could be wrong.

 

Also want to check it there is any difference that applies to Ace-flown planes - on the hypothesis that it's these planes that are having some kind of FM or weight-related terrain-avoidance problem and are leading their wingies to their doom.

In general, I think the Aces tend to avoid crashing into hills and trees better than the non-aces.

 

(ii) the player flies the QC1 plane in quick combat but definitely flies the .sqd plane in campaigns, not QC1 - making the above edit in the QC1 file had no effect in campaign, till I made it to the .sqd file as well.

Correct.

 

(iii) applying the centre of gravity change to all planes did not stop the AI-led flight hitting the trees on the slope at the end of Bertincourt aerodrome.

Did you notice any improvement, however? I think maybe a combination of reduced weight and CoG adjustment might be the key.

 

 

OK I've had a look at the weight values reported by AirWrench and it looks like the thing to do is go with HPW's original suggestion to bring the AI weight up to 'ramp weight minus some fuel', not as I've done 'loaded [max] weight minus some fuel'.

 

This is because (i) Pol has confirmed FMs were designed to be right for the PLAYER'S plane (and then tested to check AI could cope) and (ii) confusingly but significantly, the weights for the loads carried by the player's plane, from the .air files as reported by AirWrench, are consistently LOWER than the correct figures. It doesn't matter whether these too-low figures are by deliberate design or not; the point is, right or wrong in themselves, they're the ones for which the FM is designed to be right, when added onto the 'empty' weight. At least, that's the current hypothesis!

 

To give an example, the stock OFF .cfg file for the Nieuport 17 has near-enough correct figures for empty and loaded ('max') weights, 825 lbs and 1280 lbs. The real plane's loads were 143 lbs for fuel & oil and 264 lbs for pilot & military load. But the .air file gives 114 lbs for fuel and 180 lbs for other loads. Added to the empty weight, this gives what it calls a 'ramp weight' of 1019 lbs.

 

I suspect it's this 'ramp weight', and not the 'max' weight, that the FM is designed around - a weight that's about 260 lbs lighter than an N17 with pilot, and a full load of fuel, oil and ammo! And about 180 lbs lighter than the AI Empty Weight mod uses, now.

 

From what I've seen, the difference this makes will vary from plane to plane; for the Halb DII, it looks like taking off c.300 lbs, partly because the Halb starts from a very low empty weight of 1144 lbs (c.100 lbs to low, per my references) and only adds 93.5 lbs for fuel and 190 for other load.

 

It's pointless speculating why (if the above hypothesis is correct) the FM was designed to be correct with weights so far below the normal loaded weights, tho I wonder if it was this factor which enabled the AI to take off successfully in P3, whereas in P2 they more often crashed and burned, like the AI-led Halbs are doing now with the empty weight mod.

 

To test this I'm going to re-work the Halb DII's AI empty weight mod to use the 'ramp weight' figure instead of the loaded ('max') one. Will report back.

 

I think the main reason the empty weights in OFF may be lighter than the historically reported empty weights may have something to do with getting the climb rates correct. My feeling is that when the empty weights were set to historical figures, the devs found that the planes were too "heavy" to get good--or at least historically accurate--climb rates. (This may be why the Fokker DVII's, as you know, have such low empty and max weights in OFF. Otherwise, they would not be good enough climbers.

Edited December 27, 2011 by Herr Prop-Wasche

[Herr Prop-Wasche 7]

All right, everybody, I have put together an add-on for the FM and AI Empty Weight mod that contains fixes for the Halberstadt D II. I'll put the add-on up in the downloads section.

 

The add-on contains an FM update for the Halb DII along with revised empty weight settings. CoG and CoL settings are adjusted along the lines that Lima suggested so that the Halb is no longer nose heavy. Although the empty weight for the QC and Sqd versions was raised 50 pounds, this had no negative impact on the AI aircraft, even with their empty weight settings being based on Ramp Weight minus 75 pounds. I tested the mod in a campaign mission for Jasta 1 at Bertincourt field--11 aircraft sent on a patrol mission. All 11 aircraft took off and avoided the trees (they still like to brush the top of the trees, but none crashed) and hill at the end of the field and went on to successfully complete the mission.

 

Be sure to give me your comments and let me know of any other aircraft that may be having trouble with the empty weight mod.

[33LIMA 972]

Great work HPW - this sort of thing could be the approach needed for any other planes that have trouble coping with the extra weight.

 

With the CoG moved back to minus 1 and the empty weigth set to ramp minus 80, I'd found the same behaviour with the AI-led Halbs at Bertincourt - brushing the treetops and only just getting off. Even at ramp weight they still crashed, altho they got off the ground before flopping back down.

 

I'm shocked the AI fly at less than full power - talk about asking for trouble! I recall reading about an incident towards the end of WW2 when a bunch of senior officers were killed because their RAF pilot had this habit of deliberately taking off at less than full power and didn't get away with it that day because of some icing.