Board index FlightGear Development Aircraft

Sukhoi 26M aerobatic

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

Sukhoi 26M aerobatic

Postby Flying toaster » Fri Dec 01, 2006 5:23 pm

Here is a *very* slightly updated version of the Su-26 (not quite finished yet).
Image
Sukoi 26M (export)
Just thought it could be an interesting post in this forum (just doing a little advocacy here)

Once the model reaches production status, I'd like it to be integrated in CVS (if there is room left ;) )
Flying toaster
 
Posts: 348
Joined: Wed Nov 29, 2006 6:25 am
Location: Toulouse France

Re: Sukhoi 26M aerobatic

Postby curt » Fri Dec 01, 2006 5:36 pm

Flying toaster wrote:Here is a *very* slightly updated version of the Su-26 (not quite finished yet).
Just thought it could be an interesting post in this forum (just doing a little advocacy here)

Once the model reaches production status, I'd like it to be integrated in CVS (if there is room left ;) )


You guys that seem to prefer forums need to let me know specifically what you think of this particular flightgear forum and the available set of topic categories. Does this look like it will be a worthwhile and valuable resource to you? Are there new topics or categories we should create to make things work better?

There are always going to be vocal people on the dev list that dislike web based forums. But the most important factor in my view is to determine if there is a large enough group of people that feel they can communicate more effectively and more comfortably in this specific web based flightgear forum.

If that answer to that is yes, then I think it makes a lot of sense to maintain a forum. If the answer is no, then we should probably let this die on the vine and cut it off once it's all shriveled up.

If it grows and flourishes and sees a lot of use, then perhaps that will be the answer to our question.

I do want to at least attempt to minimize duplication of topics. I think it makes sense to try to limit the forum discussions to end user support and content creation and maybe other things that would be fun and interesting and aviation related, but outside the scope of the hard core code devleopment effort.
curt
Administrator
 
Posts: 1167
Joined: Thu Jan 01, 1970 12:00 am
Location: Minneapolis, MN

Re: Sukhoi 26M aerobatic

Postby Flying toaster » Fri Dec 01, 2006 5:42 pm

curt wrote:I do want to at least attempt to minimize duplication of topics. I think it makes sense to try to limit the forum discussions to end user support and content creation and maybe other things that would be fun and interesting and aviation related, but outside the scope of the hard core code development effort.


My opinion is that it is most suited for content creation due to the possibility to get precise feedback on a specific development, instead of clobbering the mailing list.
Flying toaster
 
Posts: 348
Joined: Wed Nov 29, 2006 6:25 am
Location: Toulouse France

Postby Flying toaster » Thu Dec 07, 2006 7:54 pm

This is why I love the beast ;)
http://www.wideo.fr/video/iLyROoaftY2R.html

Now I can try this at home ! LESU is a nice backdrop for that
Flying toaster
 
Posts: 348
Joined: Wed Nov 29, 2006 6:25 am
Location: Toulouse France

Postby Flying toaster » Tue Dec 12, 2006 7:47 pm

There are two files to update the flight dynamics.
flight dynamics to be placed on the root Su-26M directory
propeller to be placed in the Su-26M/Engines directory

The filenames should be changed to Su-26Mprop.xml and Su-26M.xml.

Improvements are :
- Better handling when braking hard (reduction of tendency to nose over)
- Propeller residual thrust at idle greatly decreased. The aircraft should be able to land without shutting down the engine
- Contact points for the wing tips.

There is a strange behavior with the structure contact points for the tips... When colliding hard (well not, so hard) with the ground, the aircraft will bounce to the stratosphere ...

Any flight impressions or suggestions for improvement are welcome ...

Enjoy

Enrique
Flying toaster
 
Posts: 348
Joined: Wed Nov 29, 2006 6:25 am
Location: Toulouse France

Postby Quantum » Wed Dec 13, 2006 7:53 pm

Your Su-26M is really fun to fly!

Image

Amazing how much aerobatics is possible. I've read your interesting article in the jsbsim newsletter ...

I'm really impressed!

Chris
Quantum
 

Postby Flying toaster » Mon Dec 18, 2006 6:18 pm

Yet another update ;)

Corrected sideslip indicator :

Models/Su-26M-cockpit.xml

Selection of the livery :

Models/Su-26M.xml

Keyboard shortcut for livery selection ( lowercase L ) :

Su-26M-set.xml

New view offset 10cm (4 inches) to the left to have a better view during approach

Su-26M-views.xml
Flying toaster
 
Posts: 348
Joined: Wed Nov 29, 2006 6:25 am
Location: Toulouse France

Postby grtux » Mon Dec 18, 2006 10:30 pm

hello, Flying toaster

Many thanks for that wonderful aircraft,
it is not only a perfect 3D model, it is a nice FDM, and a very good example on how to write xml animations file.

I enjoy to fly it
g.robin
LFMO
User avatar
grtux
 
Posts: 432
Joined: Thu Dec 14, 2006 4:19 pm
Location: Provence France

Postby Joacim Persson » Tue Dec 19, 2006 3:58 pm

A tip:

There's a quirk with how propellers are defined in jsbsim that isn't documented (is anything in jsbsim?) but which can be seen in the source code.

Variable pitch propellers are defined by setting minpitch and maxpitch paremeters. But if you also set the minrpm/maxrpm (to be more precise: set them to different values) you will instead get a constant speed propeller (or "fixed speed propeller" as it's called in the source code).

I don't know if the SU-26M (the real arcraft) has a variable pitch prop or a constant speed prop, but thought it's worth pointing out.

There are several models in FG that are equipped with constant speed props by mistake. This probably because aeromatic does it wrong. (Unless that bug has been fixed now. It's been reported.)


The source code lines in question are in the method "double FGPropeller::GetPowerRequired(void)" in src/FDM/JSBSim/models/propulsion/FGPropeller.cpp at line 230 and on (FG cvs version):

Code: Select all
   if (MaxRPM != MinRPM) {   // fixed-speed prop
Joacim Persson
 
Posts: 5
Joined: Tue Dec 19, 2006 3:08 am

Postby grtux » Wed Dec 20, 2006 2:18 pm

hello Joacim,

I am not sure to understand your remark.

I told you on the IRC channel, that the existing JSBSim code regarding propeller suit to most of the requests,

and the result depends mainly of the <table name="C_THRUST" type="internal"> and the <table name="C_POWER" type="internal">
content.

Jon Berndt could answer , may be you must ask again your question on the JSBSim-devel list , which is the right place for it.
g.robin
LFMO
User avatar
grtux
 
Posts: 432
Joined: Thu Dec 14, 2006 4:19 pm
Location: Provence France

Postby Joacim Persson » Wed Dec 20, 2006 10:20 pm

grtux wrote:hello Joacim,

I am not sure to understand your remark.


You are familiar with the difference between a fixed-speed (constant speed) propeller and a variable-speed (variable pitch) propeller?

From the pilot's perspective, it makes a difference regarding how the propeller rpm varies with the throttle setting.

A constant-speed prop is self-regulating: varies the pitch to achive the propeller rpm the pilot selects with the handle. If the pilot increses the throttle, the engine and propeller stays on the same rpm, it only increases the blade pitch. When going from cruise to climb, he won't have to adjust the propeller (assuming the ideal rpm is the same regardless, and a normal fixed ratio gear box of course)

For a variable pitch propeller, however, the rpm will vary with throttle settings. So when going from cruise to climb or vice versa, the pilot will have to adjust the propeller pitch manually. (again to maintain the same rpm)
Joacim Persson
 
Posts: 5
Joined: Tue Dec 19, 2006 3:08 am

Postby grtux » Thu Dec 21, 2006 12:39 am

hello Joacim,

Sure you could say it on the JSBSim-devel.

However, with the existing code,
if you build the right C_THRUST and C_POWER data table,
and
if in propeller.xml file, you comment out, the following
<minrpm>nnnn</minrpm>
<maxrpm>nnnn</maxrpm>

It is possible to adjust manually the propeller pitch, in order to maintain the same RPM.

If you don't comment out, you will get the constant speed propeller function.

My conclusion is, JSBSim answer the two following requests:

Constant-speed propeller

Variable pitch propeller

If my understanding regarding your remarks is right.
g.robin
LFMO
User avatar
grtux
 
Posts: 432
Joined: Thu Dec 14, 2006 4:19 pm
Location: Provence France

Postby Joacim Persson » Sat Dec 23, 2006 1:44 pm

grtux wrote:My conclusion is, JSBSim answer the two following requests:

Constant-speed propeller

Variable pitch propeller

If my understanding regarding your remarks is right.


Yes JSBsim can define etiher, so it seems. But you have to wade through an ever-changing source code to find out how to do it.

There is a (reported, possibly corrected by now?) bug in Aeromatic though: if one asks for a variable pitch propeller, but sets minrpm/maxrpm (which may happen as no hints are given what this data will be used for), one gets a constant speed propeller instead of the wanted variable pitch propeller, and then perhaps wonders why the propeller pitch magically adjusts itself all the time.

The thrust/pitch/torque/rpm curves of the propeller is of course not affected by this. The piloting of the aircraft is.

The fundamental problem with JSBsim is that after 8--9 years of development there is still no documentation on the configuration files. And no documentation of the design either. Only source code: doumentation of implementation. Deviations between design (what was intended) and implementation (what was actually built) is what is also known as "bugs" (unintended function). Without a design documentation, or at least a user interface documentation, there is by principle no telling what is bugs and what is features. Even if you figure out how it works right now from studying the source code, you can't be sure it will remain that way in the future.

Here someone perhaps would inflict: "Ok. You document it then, smartass." I cannot do that because I am, like everoyne else, incapable of mind-reading. Only the designer can document his/hers own thoughts. Relating thoughts to others is what we have languages for, written or spoken. After eight years the thought comes creeping up: the reason why lions can't speak is simply that they have nothing at all to say. Is there a design at all? Is there a plan? Or is there only a prototype?
Joacim Persson
 
Posts: 5
Joined: Tue Dec 19, 2006 3:08 am

Postby grtux » Sat Dec 23, 2006 3:40 pm

Because i do not pretend to defend FDM JSBSim, i will not try to say if the documentation is right or wrong, if the Aeromatic is right or wrong.

I am only an old user, (very old) who started with JSBSim and Aeromatic, to give life to model of Aircrafts, (i have a lot of them) years ago when we could access and download these oldies FlighGear versions, which tooks to us several hours to download through a low speed connection (DSL was not yet there)


Aeromatic is to me, a very useful tool with wich you can start to create the basic FDM xml file,
And BECAUSE ----it is not the "full absolute" tool you could whish--
that xml file must be modified according to the real Aircraft specifications, the parameters which are there, are understandable, when we know the aerodynamic and "any others know how" which makes that an Aircraft to fly or not.

I cannot say if the documentation is good or not to a new model developper, who start for the first time on JSBSim.

However, i can say , it could be some misunderstanding on the JSBSim developpement schedule regarding the functionnality.

I, unfortunately, experienced it last year, when i started the F-8E Crusader, conviced that a JSBSim will include a new functionnality carrier landing and catapult (was said on JSBsim news letter, and a patch did exists).
When latter on, --too late--- (the model was yet commited to CVS) i was said that function will not be (never) included in the coming FG version, my work became suddenly wrong, and to answer to friends request i made a specific FlightGear version which can fly that achieved Aircraft.

I asked to whitdraw the model from CVS which was an unuseful corpse within Flightgear.
Some member on that forum , may be, remember the story.

:) Since i have been said that some new coming function are under developpemen, which could answer my request :)

To conclude,
Aeromatic is right if the user takes the results not fully right, but only a working base.
JSBSim , to an "old" user is understandable, and practising is the best teacher
g.robin
LFMO
User avatar
grtux
 
Posts: 432
Joined: Thu Dec 14, 2006 4:19 pm
Location: Provence France

Postby hvengel » Sun Dec 24, 2006 5:32 am

Tried this model today and found some problems. Installing the initial version from the first post at the top of this thread works but has overly sensitive brakes. I then followed the instructions in the 5th post to install the updated Su-26M.xml and Engines/Su-26M-prop files. At that point the aircraft will not taxi or move at all. Backing out those changes and only installing the updated propeller file works. Then adding the Su-26M.xml file causes the problem. So there appears to be something wrong with the updated Su-26M.xml file.

I am running FG 0.9.10. Does this model need a differenet version of FG (perhaps CVS)?
hvengel
Retired
 
Posts: 1128
Joined: Sun Dec 24, 2006 4:35 am
Location: Minden Nevada

Next

Return to Aircraft

Who is online

Users browsing this forum: No registered users and 1 guest