CFS1 Curtiss-Wright AT-9 "Jeep" Work In Progress.

Climb test

Hi Ivan,
After doing the necessary corrections, it was of course easier to test climbing on the plane, although it did porpoise a lot above 10000 ft. As expected, power is not at all up to par yet!
I tried to fix the throttle for a 175 mph level flight, which would be about 85% I suppose, but for the moment I settled for 90%, as that was easier to set, and also, 90% is usually possible as a continuous climb, if I´m not mistaken.
So, at 90% throttle, (sea-level horizontal flight equivalent of 180 mph - fast cruise...) I did a climbing test to see what rates of climb could be maintained at a constant 120 mph:
upto ft - fpm
3700 - 850
4800 - 750
5000 - 700
5500 - 600
6000 - 500
7000 - 450
7500 - 400
8000 - 350
8500 - 300
9000 - 250
10000 - 200
11000 - 100
12000 - 13000 ft, 100 fpm could be maintained only at 110 mph.
First I´ll try to reduce Induced Drag and increase Zero Lift Drag to get more speed higher up, keeping the 171 mph max. sea-level speed and then I´ll do another test, although that will only improve things a bit higher up.
After that I´ll try my luck at adjusting the thrust curves to get more power at 90%.

Update:
I´ve just done the Drag adjustments and it seems that as speeds are better at altitude, the RoC also improves noticeably - looks promising!
2nd update: Now it looks like this:
upto ft - fpm
2500 - 1400
3700 - 1200
4800 - 1100
5500 - 1000
6500 - 950
7500 - 900
8500 - 850
10000 - 800
11000 - 700
It was porpoising a lot and I had to stop, but it looks very close, I´d say!
Cheers,
Aleatorylamp
 
Last edited:
Climb test

Hi Ivan,
Actually I had a mistake in the initial climb upto 2400 ft - it was only 1200 fpm, not 1400... but now,
I´ve just ever so slightly tweaked the J = 0.6 values in tables 511 and 512 corresponding to the 120 mph speed, and I got another 200 fpm out! (I´m amazed that it was so easy...).
So, it should be OK!

...IF, of course, a 90% throttle setting for continuous climb is correct!

Now it´s time for another more extensive test!
Update: Here it is:
upto ft: fpm:
2500 - 1600
4500 - 1400
6200 - 1200
8500 - 1000
9700 - 900
10000 - 800
11000 - 700
12000 - 600
13000 - 500
...by now it was bobbing up and down so bably I couldn´t continue.

I think I can leave it here, as it coincides nicely with the calculated expectations, I´d say!

Cheers,
Aleatorylamp
 
Last edited:
Those numbers are not too far off. I hope 90% is the "Climb / 30 minute" power setting.
I believe with full power / military power, it should climb quite a bit faster.

You comment about bobbing and such brings up another point:
There is something about the controllability of this aeroplane that just isn't quite right.
It appears to fly much "heavier" than it should.
I modified the moments of inertia but that didn't completely cure the problem.

- Ivan.
 
Not too far iff indeed!

Hi Ivan!
Thank you for your as always useful input!
The 90% I mention is simply pressing key "9", having 90% on the Beckwith gauge in the "Thrt" square.
This airplane indeed did have WEP, but I have no idea what type it was, so that´s why I haven´t implemented it as yet. If I can´t find out which it was, I´ll use Methanol-water, I suppose, because there is no mention of a supercharger that could have had extra boost for this plane.
Heaviness-stability-manoueverability:
OK, so there does seem to be something strange going on. I tried fiddling the MOI´s, the Stability Factor 2, the Damping... but it´s a mystery. As I´ve said before, an FS98 plane with similar entries performs really fine, but not here! Obviously the CFS1 .exe interprets the FD differently.

I´ll make another .air file then. I had simply copied over the FD from one of my FS98 German twin-engined courier planes into the CFS1 Hurricane FD I think it was, but I think I´ll repeat it, omitting the "funny" parts to do with stability and other "strange" things. ...or maybe the Hurricane FD is not the best one for this. Would the Mustang one perhaps do better?

Thanks again for your feedback!
Cheers,
Aleatorylamp
 
Best AIR File Template

Hello Aleatorylamp,

When I was starting out with AIR files, I chose the P-51D as a template because I believe it actually has more records than some of the others including the Hurricane. I don't remember what was missing so perhaps I am mis remembering. You CAN of course use which ever one you want to start.

With the stock P-51D, there are only two not so obvious gotchas and neither is all that serious:
The Wing Efficiency Factor in Record 1204 is WAY too low to be realistic. Basically it makes the aeroplane not have as much induced drag.
6000 to 7500 seems more reasonable to me.
The second issue I know about is Aileron Control Factors in Table 518. If you look carefully, one of those numbers doesn't fit the sequence.
These are in my notes but I am still adding to those notes as I go.
There are about 3 or 4 other things I have to experiment on.

- Ivan.
 
AT-9 heaviness perhaps norlal

Hi Ivan,

How can one edit the post title? I wanted to say "normal" instead of "norlal"...

So from your reply it seems then that it isn´t so much a matter of which CFS1 .air file to choose from. I suppose it´s more like the difference in damping and stability behavious from FS98 to CFS1.

Probably a "scientific" solution would be to transfer only these factors from one or two stock planes to try them out. As this plane was a trainer for B-25, B-26, A-26 and P-38 bombers, I´ll try the values that are in the stock P-38, and B-25.

Before I do this, however, a second thought: The wings on the AT-9 were purposely made too short, precisely to immitate the behaviour of the larger twin-engined bombers. The tail was too short too, to make it less manoueverable. In the CFS1 FD, wouldn´t this be the cause for the heaviness, which would then be totally normal?

Perhaps the following could confirm this: I found another article on the plane, and this one says they took off at 115 mph, and at 120 started climbing. Then, on a dead stick approach the plane stalled at 75 mph, so the glide angle had to be quite steep. Then, there was an order never to use full flaps because it would drop the plane too fast, but just before touch-down some experienced pilots did use them. Also, it couldn´t fly on one engine and the procedure was a slow spiraling descent to a convenient landing place.

Cheers,
Aleatorylamp

P.S.
The Mustang´s table 518 could be a programmer´s "number jump" mistake. The x entry for y=0 should probably be 1458 like the Hurricane one instead of 1854.
 
Hello Aleatorylamp,

I actually DIDN'T say that the AIR File you use as a template doesn't matter. I know I chose the P-51D for a reason. I believe some of the other files are missing some of the records. I don't remember what why I didn't choose the Hurricane, but I did not and things have pretty much worked out since then. I recommend the P-51D.

You make the assumption that the stock Twin-Engine aircraft have good flight models. I can tell you that this is NOT the case. Try one or two out and see if you agree.

My recollection about the description was that it was "Tricky" to fly and wasn't very stable. That seems to match the physical configuration pretty well. The Wing Area is not particularly small. It has more area than the Mustang but only weighs about 2/3 as much Its empty weight is less than half that of the P-38. Its Lateral and Longitudinal response should actually be pretty good as a consequence of the short Fuselage.
As for mimicking the bigger bombers, I don't think it really did that either because something like the early B-26 was a much much hotter aircraft.
Consider that the Fledgling was fully aerobatic but the B-25 and B-26 were not. Without knowing anything more than its shape an weight, I would guess that the aeroplane was very "twitchy" with very little stability (especially Longitudinally) and difficult to trim or level flight and fly hands off. This is just my uneducated guess and I would rely on pilots notes if you can find them.

As for the speeds you mentioned, 75 MPH Stall is probably a pretty good reflection of the aerodynamics.
The lack of ability to maintain altitude with one engine is a consequence of the very low engine power and probably of the low propeller efficiency as well.
The fairly high take-off speeds are probably because of the requirement to attain minimum control airspeed for single engine operation as soon as possible because below that speed, an engine failure could be fatal.

The number in record 518 are pretty silly anyway:
1500 is about 760 MPH IAS or around Mach 1 at Sea Level
1458 would be 755 MPH IAS or just a bit below Mach 1.
1854 works out to around 850 MPH....

The highest number might be useful for control reversal speeds but should not exceed Mach 1.

I don't know if this helps much because it is mostly based on my opinions.

- Ivan.
 
new AT-9 .air file

Hi Ivan,
I was just wondering if you would like to try out the .air file as I´ve got it now, just to see how it feels - no extensive climb testing or anything like it, because I´m quite satisfied with it thanks to your indications.
It´s attached in a .zip file.
Thanks in advance!

Update:
I just saw your detailed longer post - there is sound logic behind your explanations, statements and opinions. Thanks very much!
Before making a new .air file out of the P51 one, I´ll wait and see if you have an opinion to the attached so that I can see what to try for if it´s impossible with this one.
- The Longitudinal axis seems OK perhaps.
- The up/down instability then would perhaps also be normal with the difficulty of leaving a trim setting put and not being able to make the plane stay with that attitude when flying hands off.

Cheers,
Aleatorylamp
 
Hi Ivan, Hi everyone!
At last! I think I finally found what was wrong: I was just going to make a new .air file from the P51d and it turns out that it was already one, so I revised all the entries, and transferred the P51d ones to do with stability and other "strange" effects.

It turns out that "pitch moment due to AoA rate" was wrong at +3000 instead of -3000. After putting this right, a much lower Pitch Damping now works, as does a much lower Pitch Moment due to Elevator. These were 2 things you´d mentioned, along with the MOI´s, which are now also the lower values from the P51d.

So, now the plane is working like a plane should without wobbling around so much.
I´ve included a new .air file attachment with this post just in case there´s interest, and in case something else should crop up, I´ll try to put it right.

Thanks a lot again for your dedicated help. You definitely deserve a nice cold drink!
Here the heat wave is such that I wait for the tea to cool before drinking it!
Cheers,
Aleatorylamp
 

Attachments

  • Fledgeling.jpg
    Fledgeling.jpg
    42.6 KB · Views: 0
Flight Test Comments

Hello Aleatorylamp,

I just went back over my previous post. I guess the distractions were worse than I thought. There are so many spelling errors it is rather pitiful.
Hopefully the content of the message came through reasonably intact.

Regarding your latest flight model:
1. The Height above ground in Record 301 is still too low. Currently it is around 105,000. It should be around 125,000
2. The Engine will not idle. It slows down to about 400 RPM and then cuts.
3. The inertias are still too high. This masks out the control effectiveness.
4. The aeroplane is distinctly un-controllable. This is most likely because of inertias because although the maximum roll rate is not slow, it is very difficult to start or stop the roll. (I was unable to execute even minimal aerobatics.)

Regarding inertias, I tried changing Record 1001 for Moments of Inertia but that seemed to have very minimal effect.

Edit: I just saw your latest post I will see if it cures the issues. I had started my post well before lunch and completed it only after getting back from Lunch.

- Ivan.
 
Flight test comments

Hi Ivan,
OK, thanks a lot! I know how it goes...
I had cured record 301 but something got in the way and I re-loaded the defective one. I´ll fix that again!
Re. your latest post: I think I´d been able to improve some of the issues just before, when I attached the second .air file. The MOI´s are better, and the pitch issue. The roll ist still a bit too rolly though, and I´ll check the engine idle - thanks!
OK, I have to buzz off into town for the evening now!
Cheers,
Aleatorylamp
 
Still difficult

Hi Ivan,
It´s still not acrobatic, and the rolling as I said before is still off, and then, although the pitching axis is better, it won´t do a looping. Most probably it has to do the more subtle parameters where acting on one axis there are effects on others. Hopefully we can find a solution.
Cheers,
Aleatorylamp
 
Aerobatics

Hello Aleatorylamp,

We all went to see a movie this morning (Minions) so I didn't have a chance to get back to you earlier.
You already know about the Fuselage Height above ground.
The Rolling Axis still seems to have way too much inertia. Perhaps it is something else but it appears that once the aeroplane starts rolling, it doesn't stop. I know the Aileron Control Effect is very high because if you do a complete 360 degree roll, it is VERY fast. The problem is that it is nearly impossible to stop rolling when you want to.

Also the Elevator Control and Pitch is a bit strange. When I pull up into a climb after take off, the aeroplane seems to want to continue pitching up and will go way past vertical without any further control input.

The Directional /Rudder inertia is also still pretty high apparently. In trying to control Torque effect which pulls to the left, the aeroplane will continue rotating way past center and off the runway in the opposite direction.

I do have a silly question: You commented that this file was based on the P-51D AIR file but it seemed that this fact was a surprise to you as well. If so, then where did the AIR file come from if you were not the one to build it? This might give some insight as to what other characteristics are tuned into it that we don't know about.

- Ivan.
 
"New" AT-9 .air file

Hi Ivan,

Originally the airfile is a FS98 one I also included in the work in progress upload, and it comes from my FS98 Siebel FL-104 Hallore twin engined courier, adapting all the parameters for the Fledgeling. This FS98 .air file went fine and I used it while building the model.

When I recently re-took up the model, I adapted the CFS1 Gotha G.4 Hieronymus .air file and started playing with Tables 511 and 512. Initially these were from the Hurricane - that´s why I thought the rest was also from the Hurricane. Later I altered the tables from some generic Cessna charts I saw on the internet, and later still, I tried out 50% and 33% Hurricane tables - along with different combinations of angles of incidence and angles of twist in the Main Wing section.

Then, a couple of days ago, I re-transferd all the AT-9 data with my propeller tables into a new P51d .air file, but at the end I realized it had been a P51d one all along (I saved it just in case). This I noticed when I compared other sections in the .air file with different CFS1 stock aircraft FD.

What I finally did, was correct all the Primary Aerodynamics entries that didn´t come from the AT-9 specs, but from the Gotha, to what the P51d FD had, in order so to speak, to start from a "clean" board.

Then I compared the two .air files, the new one and the corrected one, to make sure they had the same entries everywhere, and tried them out, establishing that both flew the same.

Anyway, that´s the new .air file I posted yesterday. I further corrected another 20000 in the Height above Ground Record 301, and increased the idle speed in Graph 506 - two things which were easy to correct after your kind comments.

The rest is however, not so easy! For the roll axis, I was just now trying the plane out with the data from the P51d, but it isn´t any better, I think. I´d increased Control Moment and Roll Damping trying to make it more acrobatic, but it wasn´t much use. The only thing that´s better now is the pitching axis - one can do a looping - but one can´t fly hands off because it never holds the horizontal trim, always going upwards or downwards. I´m at a bit of a dead end.

Well, we shall see how is goes on... it´s no piece of cake, this one!
One nice thing: I put in your propeller blurrs here too and they are looking really great!!
Cheers,
Aleatorylamp
 
Strange - Fledgeling FS98 .air file definitely better

Hi Ivan,

Just to make sure, I re-did the whole AT-9 .air file once again using a new P51d .air file.
Prior to that I tested the P51d to established its acceptable behaviour - i.e. no unstoppable roll or the strange pitch instability like the AT-9 .air file upto now.

This time, again, I only entered the AT-9´s dimensions, weight, drag and power parameters, including prop and engine tables, leaving in the lift/yaw/roll/pitch and MOI parameters of the P51d.

Unfortunately however, it´s doing the same stupid rolling and pitching things again. Obviously this state of affairs is totally unsatisfactory to upload the plane with a CFS1 .air file, so I think I´ll stop.

The only viable alternative I can see is to upload the plane using the decent FS98 .air file that it already has. I see that quite a lot of CFS1 aircrafts´s .air files come from FS98, so that isn´t too bad either. So, as soon as I finish polishing up the visual model in AF99, I´ll upload the new finished version of the Fledgeling!

Cheers,
Aleatorylamp
 
Hmmm.... That is very odd. At least with the AT-9, a FS98 AIR file won't hurt too much since it does not have a supercharger anyway.

Speaking of FS98, I went through a stash of old software that I hadn't disturbed in a while and found about 6 copies of FS98, Windows 3.1 install disks, MS DOS 6.22 install FLOPPIES, about 4 each of Windows 95 and Windows 98 CDs and at least a dozen Windows NT / NT Server packages new in the wrap along with a LOT of other software including Slackware and Red Hat Linux in a couple different versions and even BSD Unix.....

I am actually fairly certain that a pretty fair twin-engine flight model CAN be built from the P-51D. I have my P-38 and B-25 done that way. I can give it a shot myself if you want to email me some details about the AT-9.

- Ivan.
 
P38 and B25

Hi Ivan,
Well, you have a piont there, that without a supercharger which is the main point in CFS1, there wouldn´t be such a problem using an FS98 .air file. It would be one written for a WW2 Courier plane which was also used as a racer, and goes quite nicely.

I had also actually thought of using your P38 .air file at one point, and a B-25 one - if it´s yours, better... but as they are bigger, heavier and more powerful, I´d gone off the idea for the time being. I´m sure a decent CFS1 .air file from the P51d for the AT-9 can be done, but I can´t do it! Maybe the difficulty is the much lower power that makes it somehow more unstable.

Thanks indeed for your offer for making an .air file for the AT-9.
All the information I have on the AT-9 is actually in the last .air file I posted a couple of days ago - dimensions, weights, power, the lot. I really don´t have any more info on this plane.

Cheers,
Aleatorylamp
 
AT-9 Fledgeling aeroplane

Hi Ivan,
Thank you very much once again for your kind offer to try and put the Fledgeling´s CFS1 flight dynamics right.
Maybe we will not have to make do with only an FS98 one after all!

I have e-mailed you the plane together with the info I could find.

As far as I´ve been able to get it now, the main aspects are fine as follows:
1) Sea-level horizontal flight and power is fine
2) Climbing rates for different altitudes are also fine.

Basically the problems boil down to the rolling and pitching behaviour:
1) It continues rolling too far
2) It won´t stay put on level flight - it always tries to climb or dive.

Autopilot and Beckwith gauge are present on the panel for testing.
I hope this won´t put you out too much!
Thanks and cheers!
Aleatorylamp
 
Hello Aleatorylamp,

I decided to do a bit of research on the AT-9 today.
Turns out there is actually a LOT of data out and some of the data you are working with are simply wrong.
The actual propeller diameter is actually 8 feet 6 inches instead of 8 feet. That should make a serious difference.
Another thing I observed today was that you do not have a DP file. That means that you are working with the DP for the Spitfire Mk.IX.

I haven't gotten very far with it yet but I will let you know when there is something worth testing.

- Ivan.
 
At-9

Hi Ivan,

It´s very good of you to dig around for more info, and wonderful that you found it! My own long searching revealed what I found, although I did have the sensation that there was a lot of garbage around. There were one or two things I couldn´t get at, which were all the At-9 manuals, for maintenance, repairs, flying and spare parts, because they wanted 25 dollars for each one and I couldn´t find a free source.

Hmmm, so the prop was 8 ft 6 inches - great! Yes, that will definitely help a lot.

I did include the Dp file! In the plane I e-mailed to you was the latest model, with prop blurrs and the newest P51d derived .air file. The Dp file had neither bombs nor guns, but with correspondingly adjusted boxes for fuselage, wings, tail and engines. Only the first model I uploaded used the Spitfire´s DP´s.

There´s no hurry on this matter, and I appreciate your efforts and time involved, and look forward to the results, which will no doubt be very interesting!
Cheers,
Aleatorylamp
P.S.
Interesting: I just tried the new 8.5 ft prop diameter and it required reducing Zero Lift Drag from 82 to 77, to maintain speed at 197 mph at 500 ft.
Power stayed as before: 295 Hp at 2289 RPM.
 
You must log in or register to reply here.
Back
Top