Board index FlightGear Development Aircraft

New historically realistic Sopwith Camel JSBSim FDM released

Questions and discussion about creating aircraft. Flight dynamics, 3d models, cockpits, systems, animation, textures.

Re: New historically realistic Sopwith Camel JSBSim FDM rele

Postby Bomber » Wed Apr 24, 2013 4:23 pm

flug wrote in Thu Apr 11, 2013 7:18 pm: - Specifically, the Camel won't even do a slow roll without a whole lot of coaxing, and when it does--23 seconds to complete the roll (according to an article by Brian Lecomber; see however the contrasting description in the Manouvres of Inverted Flight article; both are in the Historical Docs directory. I think they are just using the same term for a slightly different maneuver). This behavior still does not quite match the descriptions but it is a lot closer than it was.


Got a problem with this sentence.... 23 seconds is for a radial engine... Sopwith had a Rotary...

The plane should roll in one direction and be impossible in the other.

Simon
"If anyone ever tells you anything about an aeroplane which is so bloody complicated you can't understand it, take it from me - it's all balls" - R J Mitchel
Bomber
 
Posts: 1934
Joined: Fri Dec 14, 2007 7:06 pm
OS: Windows XP and 10

Re: New historically realistic Sopwith Camel JSBSim FDM rele

Postby LesterBoffo » Wed Apr 24, 2013 6:13 pm

I wouldn't say impossible, but the engine torque did make rolling to the right a bit harder. Otherwise they would have been impossible to fly. And their record for kills sort of belies this.
User avatar
LesterBoffo
 
Posts: 2129
Joined: Sun Oct 02, 2011 4:02 pm
Location: Oregon, USA
Callsign: LesBof
Version: 2018.3.2
OS: Win10 Pro

Re: New historically realistic Sopwith Camel JSBSim FDM rele

Postby Bomber » Wed Apr 24, 2013 10:52 pm

Did you fully comprehend the quoted text ?

This is not a gentle roll to one side.... but a complete barrel roll...

Now if it takes 23 seconds to complete a roll using a radial engine then using a rotary engine it should be impossible in one direction.... the other direction with torque will do it real easy.

Simon
"If anyone ever tells you anything about an aeroplane which is so bloody complicated you can't understand it, take it from me - it's all balls" - R J Mitchel
Bomber
 
Posts: 1934
Joined: Fri Dec 14, 2007 7:06 pm
OS: Windows XP and 10

Re: New historically realistic Sopwith Camel JSBSim FDM rele

Postby Bomber » Sat Apr 27, 2013 2:06 pm

The stated figures in the docs are that the fuel tanks can hold 170 litres of fuel... yet there's only a capacity for 113lbs where it should be [edit]269lbs... that's a big difference..

And whilst I'm at it there doesn't seem to be any lub oil weight taken into account... docs say 30 litres behind the engine which is another [edit]57.9lbs

And having done the calcs on this plane the CoG is too far back and too low... which might explain why there's less fuel weight..

Regards

Simon
Last edited by Bomber on Sun Apr 28, 2013 11:43 am, edited 1 time in total.
"If anyone ever tells you anything about an aeroplane which is so bloody complicated you can't understand it, take it from me - it's all balls" - R J Mitchel
Bomber
 
Posts: 1934
Joined: Fri Dec 14, 2007 7:06 pm
OS: Windows XP and 10

Re: New historically realistic Sopwith Camel JSBSim FDM rele

Postby Bomber » Sat Apr 27, 2013 2:53 pm

'All Up' combat weight should be [edit]1576lbs...

CoG from the prop as a datum (which is used on this fdm) should be...

x = 111.8cm
y = 0cm
z = 8cm (above the thrust line)

Simon
"If anyone ever tells you anything about an aeroplane which is so bloody complicated you can't understand it, take it from me - it's all balls" - R J Mitchel
Bomber
 
Posts: 1934
Joined: Fri Dec 14, 2007 7:06 pm
OS: Windows XP and 10

Re: New historically realistic Sopwith Camel JSBSim FDM rele

Postby flug » Sat May 04, 2013 6:27 pm

Thanks to all, several comments upthread are very helpful and give me a few things to think about. I'll see what I can incorporate into the next version.
flug
 
Posts: 218
Joined: Wed Jun 17, 2009 1:06 am

Re: New historically realistic Sopwith Camel JSBSim FDM rele

Postby Jetman. » Mon May 06, 2013 11:36 pm

Hey flug! I know I was going to find you on here sooner or later! :lol: I shot you down 5 times today with the F-14's missiles. Good to see Bombable working well on multiplayer! Thanks for being the guinea pig! :lol:
Favorite Aircraft: Pretty much all military aircraft.

FG videos on YouTube! Choose the FlightGear playlist!
http://www.youtube.com/user/ParasiteQueen1

SPECS
8GB RAM 1600Mhz DDR3
AMD A8-5500 Quad @ 3.2Ghz Turbo 3.7Ghz
Zotac GTX 750Ti 2GB GDDR5
User avatar
Jetman.
 
Posts: 103
Joined: Thu Apr 29, 2010 8:54 pm
Location: KATL
Callsign: Jetman
Version: 3.2
OS: Windows 8.1

Re: New historically realistic Sopwith Camel JSBSim FDM rele

Postby flug » Tue May 07, 2013 5:00 am

Jetman. wrote in Mon May 06, 2013 11:36 pm:Hey flug! I know I was going to find you on here sooner or later! :lol: I shot you down 5 times today with the F-14's missiles. Good to see Bombable working well on multiplayer! Thanks for being the guinea pig! :lol:


Glad I could be of service! I was, uh, completely oblivious . . .

Oblivious being a fairly normal state for me to be in . . .
Last edited by flug on Tue May 07, 2013 6:46 am, edited 2 times in total.
flug
 
Posts: 218
Joined: Wed Jun 17, 2009 1:06 am

Re: New historically realistic Sopwith Camel JSBSim FDM rele

Postby flug » Tue May 07, 2013 6:16 am

I just released ver. 1.6 of the Sopwith Camel.

http://brenthugh.com/flightgear/sopwith ... Sim-16.zip

I think this is the most historically realistic version yet. I realized in reading through historical documents, that the Camel's controls are really far different from the controls in modern aircraft. The rudder, for instance, was a constant topic of discussion. Was it too small? Should it be increased in size? Was it responsible for some of the training crashes? Etc. etc. etc.

Well, the obvious conclusion is that is WAS too small and relatively less effective--certainly by today's standards. Just for example, the rudder on a Cessna 172 is about twice as powerful as the Camel's rudder (combination of larger surface area and longer moment arm).

Ailerons--same story. The ailerons aren't too small--in fact, they're huge. But they induce a massive amount of adverse yaw and a good deal of drag to boot. These 'side effects' are more noticeable, really, than the primary roll effect.

The elevator is invariably noted as powerful--which it was, in comparison with the rudder and aileron, but maybe not in comparison with the rudder on, say, a modern aerobatic aircraft. Just for the example, the elevator on the Cessna 172 is about 40% more powerful (23% larger surface area and 15% longer moment arm).

When I reduced the effectiveness of the controls, particularly the rudder, and increased the adverse yaw and drag on the ailerons, suddenly a lot of the historical descriptions of flying the craft started to make a lot more sense.

Just for example, one of my objectives in starting this project was to explore the reason why the Camel is reputed to be able to do right turns "three times faster" than left turns. We know it has something to do with the gyroscopic moment of the large rotary engine, maybe it has something to do with the torque, too--who knows for sure?

Here is what we know, and I've discovered or confirmed in experimenting with the JSBSim Camel, about the gyroscopic moment of the Camel's engine and its effect on the flight of the Camel:

- The gyroscopic moment forces the nose down when making right turns, and forces the nose up when making left turns

- The slower the turning rate, the less the effect; the faster the turning rate, the stronger the effect. So, on a very slow, sedate turn--say, a two-minute turn--you won't notice the gyroscopic effect at all. In a fast turn--the Camel is reputed to be able to do a 360 turn in 8-10 seconds--the effect will be very noticeable indeed.

- A quick left turn, in forcing the nose up, can easily induce stall if you're not careful.

- Nose up and nose down is the obvious and noticeable effect. But there is a bit of a chain reaction of other effects. Just for example, to counter the nose-down tendency in the right turn, you'll ease off the rudder or even add left rudder, depending on the turn rate. This is helpful, because it keeps the Camel from over-rolling in the turn, which would require aileron to counter. In the left turn, the effect is reversed: You must add additional left rudder, which causes the Camel to over-roll leftwards, and you must use right aileron to counteract this effect.

- All of this is far more noticeable to the pilot at slower speeds. At higher speeds, general stability is higher and the effectiveness of the control surfaces is higher--meaning that you are more likely to be able to overpower the gyroscopic effect more easily. These lower speed ranges are the ones you easily fall into as you're in the middle of a dogfight.
Image

Now, the gyroscopic effect is noticeable, but it is really not overwhelmingingly strong. One contemporary source compares it in strength to a gust of wind--but one you know to expect and can plan for.

In a modern aircraft, you'd just compensate for the gyroscopic effect with your other controls--which would be designed with plenty of control authority to do so--and you wouldn't notice too much difference between the two directions.

But--what if the Camel's rudder is not strong enough to compensate? At least, not powerful enough to compensate during the quickest turns, when the gyroscopic moment is the most powerful.

That seems to be indicated by historical documents that claim that full or near-full left rudder is used in both left and right turns.

Here is how it works:

- In a right turn, you initiate by rolling, using right rudder and aileron. As the roll deepens, you use elevator to tighten the turn and elevator becomes the primary way you control the turn rate. As the turn rate tightens the nose starts to drop so you ease off rudder to compensate. Now you are at the right roll angle with no rudder, so you need only slight tweaks of the ailerons to maintain this angle. As the turn rate tightens more, the nose drops more and you add more left rudder to compensate. You max turn rate will be achieved with full left rudder and little or no aileron. With practice you just move to full left rudder early on, then use elevator to manipulate the turn rate so as to keep nose level (quicker turn rate drops the nose, slower turn rate raises the nose--when you have max rudder and the nose is level you have reached maximum level turn rate).

Note that in this maneuver you have used the entire control authority of the rudder--you start out with full right rudder; at max right turn rate you are at full left rudder.

- Now contrast this with a left turn. To initiate the turn, use left rudder and left aileron. As you bank, use elevator to tighten up the turn. As the turn tightens, the nose rises. So you must add more left rudder to compensate. More left rudder means more bank to the left, so you now must add right aileron to counteract the roll tendency of the left rudder. As you use the elevator to tighten up the turn, soon you use all of the leftwards movement the rudder has--you started the turn with left rudder--and once you have maxed out your left rudder, that's it. You can't tighten the turn any more or the nose will rise and you will stall.

So that's it--basically, the speed of both your right and left turns is limited by your rudder control authority to counteract the gyroscopic moment of the engine. But in the left turn you run out of rudder control authority very fast, so you can't tighten up the turn much at all.

Now, that is for a level turn about 55-60 knots. If you have some speed--90-100 knots or more--you can power through more easily and the effect of the gyroscopic moment seems less apparent.

Right now with version 1.6 I get left 360 degree turn about half as fast as the right 360 degree turn (conditions as outlined above: level flight, near sea level, about 55-60 knots). I'm not sure if the difference is some level of historical exaggeration (the 3X number is often given in context of "Pilots would rather turn 270 degrees right than 90 degrees left" so perhaps it's not an exact amount of time so much as a dislike of the noticeably more sluggish direction), some combination of factors not explicitly mentioned (perhaps pilots were taught to take the left turn even more slowly than absolutely required, in order to give the stall possibility in the left turn sufficient space, especially at slow speeds which are by nature close to stall speed already and where all controls are relatively less effective than they are at higher speeds), or if there is some more tweaking to be done in the JSBSim model.

Image
flug
 
Posts: 218
Joined: Wed Jun 17, 2009 1:06 am

Re: New historically realistic Sopwith Camel JSBSim FDM rele

Postby flug » Tue May 21, 2013 1:48 am

FYI Oscar chose the historically realistic JSBSim Sopwith Camel FDM that I've been developing in concert with Bombable for his Aircraft of the Month review--check it out here:

http://www.youtube.com/watch?v=C6gRFEEkkFE

There have been a couple of helpful/insightful comments in the youtube comments already.

I made a YouTube channel and added a couple of videos that demo some of the Camel features and shows how I fly the Bombable scenarios with Camel, shows some footage of multi-player dogfights with the Camel, and so on:

http://www.youtube.com/channel/UCtGZ0kqmCRJ9j-jRxRzBr9w
flug
 
Posts: 218
Joined: Wed Jun 17, 2009 1:06 am

Re: New historically realistic Sopwith Camel JSBSim FDM rele

Postby flug » Tue May 21, 2013 2:22 am

daveculp wrote in Sun Mar 31, 2013 10:09 pm:I'm looking through the JSBSim code, chasing the wires to find the gyroscopic force. This may take a while :)


I got looking through the JSBSim code the other day and I think I have it.

The relevant file appears to be this: http://jsbsim.sourceforge.net/JSBSim/FG ... ource.html

If you search through that file for 'sense' you find that the sense is used in for three different purposes:

- To calculate the p factor (which is a kind of a different calculation than the other two)
- To calculate the Torque
- To calculate the gyroscopic force

Interestingly enough, here is the equation for torque:

Code: Select all
vTorque(eX) = -Sense*PowerRequired / (local_RPS*2.0*M_PI);


Notice the negative sign in front of sense - why is that required?

Here is the similar equation involving the gyroscopic force:

Code: Select all
vH(eX) = Ixx*omega*Sense;


Now--why does sense have a negative sign under torque but not here? I believe that a negative sign here would solve the problem of reversed gyroscopic moment from the engine.

Note that both the torque & vH are combined in this equation--again suggesting to me that the rotational direction ('sense') should be treated equivalently in both equations above:
Code: Select all
vMn = in.PQR*(Transform()*vH) + Transform()*vTorque;


Just to re-state: I believe this change in line 276 of FGPropeller.cpp will fix the bug:

Code: Select all
00276   vH(eX) = Ixx*omega*(-Sense);


FYI I filed a bug report here: https://sourceforge.net/p/jsbsim/bugs/110/
flug
 
Posts: 218
Joined: Wed Jun 17, 2009 1:06 am

Ver 1.8 of historically realistic Sopwith Camel released tod

Postby flug » Thu Jan 09, 2014 7:45 am

I released version 1.8 of the historically realistic Sopwith Camel JSBSim FDM project today:

http://brenthugh.com/flightgear/sopwith ... Sim-18.zip

Thanks to much feedback from users of this forum and other places around the internet when the Camel was Aircraft of the Month in May 2013--I was able to incorporate many of the issues and ideas you all raised into this new version.

The Camel is included as part of the Bombable package. I also released a new version of Bombable today, completely compatible with FG 2.12. You can find out more about Bombable or download that package here:

viewtopic.php?f=6&t=5742
flug
 
Posts: 218
Joined: Wed Jun 17, 2009 1:06 am

How to add a/c to Flightgear download page?

Postby flug » Thu Jan 09, 2014 8:00 am

Does anyone know how I could get the Sopwith Camel added to the Flightgear aircraft download page here?

http://www.flightgear.org/download/aircraft-v2-12/
flug
 
Posts: 218
Joined: Wed Jun 17, 2009 1:06 am

Re: New historically realistic Sopwith Camel JSBSim FDM rele

Postby Thorsten » Thu Jan 09, 2014 8:13 am

Not to 2.12 - I don't think that's ever been done, but if you make a merge request to GIT and get it committed to FGData, it will be automatically part of the selection for the upcoming 3.0.

As far as I know the download page is automatically created from all aircraft in the repository and includes only aircraft in the repository.
Thorsten
 
Posts: 11577
Joined: Mon Nov 02, 2009 8:33 am

Re: New historically realistic Sopwith Camel JSBSim FDM rele

Postby Gijs » Thu Jan 09, 2014 9:42 am

Yes, Thorsten's procedure is correct. If you don't want to/can pull fgdata, you may also ask me or anyone else with commit rights to push a zipped package. Merge requests are the preferred way though.

For your aircraft to appear on the download page, make sure to give it a status rating. Aircraft without are hidden by default.

And finally, please do contact Vivian to see if he agrees with your changes (assuming thatthis will overwrite the current Camel's FDM).
Airports: EHAM, EHLE, KSFO
Aircraft: 747-400
User avatar
Gijs
Moderator
 
Posts: 9417
Joined: Tue Jul 03, 2007 2:55 pm
Location: Amsterdam/Delft, the Netherlands
Callsign: PH-GYS
Version: Git
OS: Windows 10

PreviousNext

Return to Aircraft

Who is online

Users browsing this forum: No registered users and 4 guests