Re: [Flightgear-devel] L1011-500
Tiago Gusmão wrote: Michael Stilmant wrote: please update the clock of your computer you send old Email It's probably just my ISP lagging ;) thanks for pointing that out Ampere K. Hardraade wrote: I was using multiple textures on the MD11's fuselage at first but then switch to a single gigantic texture (4096X1024). There is no difference between them performance wise. However, I would advise you to use one single texture for the sake of maintance (such as creating a new livery). This is the very reason I switch to using a single texture. Regards, Ampere On July 6, 2003 12:53 pm, Tiago Gusmão wrote: Is it better to use a singles texture, or separate fuselage texture from other pieces? For now, i'll be sticking to a single texture, easier to manage and to use indeed :) I'm still very confused with the axis used, and the model is appearing behind the actual FDM A/C values used are: AC_CGLOC 921.1 0.0 -18.0 AC_AERORP921.1 0.0 0.0 AC_EYEPTLOC 0.0 -30.0 75.0 AC_VRP 0.0 0.0 0.0 the 3d model has nose at (0,0,0) I understand it this way, please correct what's wrong: AC_CGLOC - this is the center of gravity, from which all other points are related to No, the center location (0,0,0) doesn't need to be specified. The CG is relative to that location, just like any other locations. AC_AERORP - aerodynamic reference, should be near the CG, so it has a similar value I don't think that an offset only on the Z-axis (hight above the ground, negative sign) does actually work well. Maybe you should make them exactly equal, or move the X axis slightly forward. AC_EYEPTLOC - this is where the view is placed when in cockpit view, related to(??) CG (??) No, related to (0,0,0) which is not specified. If you defined it relative to CG right now, then you just have to add the (x,y,z) values of the CG to the (x,y,z) values of the EYEPTLOC. AC_VRP - point related to the CG where the nose of the 3d model should be Not exactly where to nose should be, but rather the difference between (0,0,0) of the JSBSim model and (0,0,0) of your 3D model/ A 3d model with nose made at (0,0,0) in the editor should have a VRP of (0,0,0) and this should be enough to position the model properly This is only true if you put (0,0,0) at the nose in the JSBSim model. Looking at the above data, that might be true. Where did you get these numbers, are they from aeromatic? Erik ___ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel
RE: [Flightgear-devel] L1011-500
I'm still very confused with the axis used, and the model is appearing behind the actual FDM A/C In the aircraft config file, locations are given in structural frame. Structural frame in an aircraft has the X axis pointing backwards, the Z axis up, and the Y axis out the right side. The origin can, technically, be anywhere. Usually, it is near the nose. The exact location does not really matter, because *internal* to the FDM, contact point distances from the CG are computed and used - again: _internal_ to the FDM only the difference between two points is used. values used are: AC_CGLOC 921.1 0.0 -18.0 AC_AERORP921.1 0.0 0.0 AC_EYEPTLOC 0.0 -30.0 75.0 AC_VRP 0.0 0.0 0.0 the 3d model has nose at (0,0,0) I understand it this way, please correct what's wrong: AC_CGLOC - this is the center of gravity, from which all other points are related to Sort of. The CG is the point about which the equations of motion are based - the CG is the point about which the forces and moments act, and about which the aircraft rotates. AC_AERORP - aerodynamic reference, should be near the CG, so it has a similar value Yes. AC_EYEPTLOC - this is where the view is placed when in cockpit view, related to(??) CG (??) This is just the point where the pilot's eyes are located. I do not think this is used in FlightGear, but it is used in JSBSim to also calculate the accelerations that the pilot feels. The pilot's sensors are located near his/her eyes, too. This is given in inches as other measurements are, and in the same coordinate system as CG, reference point, etc. AC_VRP - point related to the CG where the nose of the 3d model should be A 3d model with nose made at (0,0,0) in the editor should have a VRP of (0,0,0) and this should be enough to position the model properly This is the point that the 3D modeler and the aircraft flight model designer agree upon as a reference point. From the FDM point of view, really only the CG is of interest - i.e. the latitude/longitude/altitude that is reported by the FDM refers actually to the specific lat/lon/alt of the CG. After long discussions some months ago, we decided that in order to make things work out better, the FDM would report the position of some common point that the modeler could easily see. That point was agreed upon as the nose of the aircraft. So, the VRP is simply the (x,y,z) coordinate (in inches) of the nose of the aircraft in the same coordinate system being used for reporting the CG, aero reference point, landing gear locations, etc. The modeler would then make sure either that their 3D model had its origin at the nose, or that the offset from the origin of the 3D model to the nose (the agreed upon visual reference point) is given. I forget which file is used to report this 3D model offset, or skew. The end result of using the VRP is that the FDM always reports the position of the nose (considering the orientation and position of the aircraft), and if FlightGear places the 3D model so that the nose of the aircraft model is at the lat/lon/alt reported by the FDM, then everything works out. But, the camera (I believe) looks at the nose of the aircraft, too, so it appears as though the aircraft rotates about the nose, so there is probably some adjusting that needs to be done to the camera angle ... I forget, and that's not my area of expertise. Jon ___ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel
Re: [Flightgear-devel] L1011-500
Jon S Berndt wrote: On Tue, 6 Jul 2004 20:09:38 +0200 Durk Talsma [EMAIL PROTECTED] wrote: Just a few comments below, since I also ran into similar issues creating and updating the MD-11 aero file. I also had problems using the FG_TURBINE and switched to FG_SIMTURBINE, is this a good idea? Yes, this reflects a recent change in JSBSim, which hasn't traversed it's way back into aeromatic yet. Actually, I did make the change - it should be implemented in Aeromatic. I can't test it right now. The file generated by aeromatic had the AC_THRUSTER inside the AC_ENGINE tags and it didn't work until i separated them, aero-matic bug? Not really a bug: This how JSBSim expected things to be, until a couple of weeks ago. It took me a while to figure this one out. ;-) I thought I had made a big announcement. If I did not, my apologies. Jon The announcement is good, i just didn't look in the right place :| I 've tweaked the landing gear to the real locations, assuming (0,0,0) it's the nose. In the 3d model, I have the tip of the nose in (0,0,0) and it's been scaled to match units with meters and rotated to fit the FG's axis. In fgrun, the aircraft rotates around the nose, anything i can do to fix this without messing with the model itself? I also noticed that lighting in fgrun is fixed relative to the aircraft, shouldn't be the aircraft rotating around and light coming from somewhere near eye position? When i tell it to start from a RWY, it starts in air, sinks and ends up in the right orientation, but only the cockpit is above the RWY, the rest is before the RWY. What is the reference point used to align the A/C on the RWYs? With such model positioning, should i need to use any offsets in the the model xml? If someone wants to take a look at it in-game: http://tiagogusmao.home.sapo.pt/l1011/L1011-500.zip the textures are pre-pre-alpha and the model still needs some work ;) Dont forget to put the RB211 in the Engine folder. Regards, Tiago ___ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel
Re: [Flightgear-devel] L1011-500
I was using multiple textures on the MD11's fuselage at first but then switch to a single gigantic texture (4096X1024). There is no difference between them performance wise. However, I would advise you to use one single texture for the sake of maintance (such as creating a new livery). This is the very reason I switch to using a single texture. Regards, Ampere On July 6, 2003 12:53 pm, Tiago Gusmão wrote: Is it better to use a singles texture, or separate fuselage texture from other pieces? ___ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel
Re: [Flightgear-devel] L1011-500
Michael Stilmant wrote: please update the clock of your computer you send old Email It's probably just my ISP lagging ;) thanks for pointing that out Ampere K. Hardraade wrote: I was using multiple textures on the MD11's fuselage at first but then switch to a single gigantic texture (4096X1024). There is no difference between them performance wise. However, I would advise you to use one single texture for the sake of maintance (such as creating a new livery). This is the very reason I switch to using a single texture. Regards, Ampere On July 6, 2003 12:53 pm, Tiago Gusmão wrote: Is it better to use a singles texture, or separate fuselage texture from other pieces? For now, i'll be sticking to a single texture, easier to manage and to use indeed :) I'm still very confused with the axis used, and the model is appearing behind the actual FDM A/C values used are: AC_CGLOC 921.1 0.0 -18.0 AC_AERORP921.1 0.0 0.0 AC_EYEPTLOC 0.0 -30.0 75.0 AC_VRP 0.0 0.0 0.0 the 3d model has nose at (0,0,0) I understand it this way, please correct what's wrong: AC_CGLOC - this is the center of gravity, from which all other points are related to AC_AERORP - aerodynamic reference, should be near the CG, so it has a similar value AC_EYEPTLOC - this is where the view is placed when in cockpit view, related to(??) CG (??) AC_VRP - point related to the CG where the nose of the 3d model should be A 3d model with nose made at (0,0,0) in the editor should have a VRP of (0,0,0) and this should be enough to position the model properly Thanks in advance, Tiago ___ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel
Re: [Flightgear-devel] L1011-500
Jon S Berndt wrote: On Fri, 2 Jul 2004 22:46:14 +0100 [EMAIL PROTECTED] wrote: The L1011-500 uses the Rolls-Royce RB211-524B take a look here for details: http://www.aircraftenginedesign.com/TableB3.html I have committed to JSBSim CVS an RR RB211-524 engine. Jon The RB211-524B has only 5 pounds of thrust, that one is suitable for the H and G models acording to http://www.rolls-royce.com/media/product/rb211524.jsp I have generated a 5 pound version, and I can get TSFC values and stuff , but I don't understand those tables below. I also had problems using the FG_TURBINE and switched to FG_SIMTURBINE, is this a good idea? The file generated by aeromatic had the AC_THRUSTER inside the AC_ENGINE tags and it didn't work until i separated them, aero-matic bug? Now i can get the L1011 working, but if i start from the RWY, it appears around 50 meters AGL and starts rotating backwards until it gets under the floor and freezes. If I start it in flight it behaves nice. What are the conventions for propper positioning of the model? Is it better to use a singles texture, or separate fuselage texture from other pieces? thanks in adavance :) Regards, Tiago ___ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel
Re: [Flightgear-devel] L1011-500
Just a few comments below, since I also ran into similar issues creating and updating the MD-11 aero file. On Sunday 06 July 2003 18:53, Tiago Gusmão wrote: Jon S Berndt wrote: On Fri, 2 Jul 2004 22:46:14 +0100 [EMAIL PROTECTED] wrote: The L1011-500 uses the Rolls-Royce RB211-524B take a look here for details: http://www.aircraftenginedesign.com/TableB3.html I have committed to JSBSim CVS an RR RB211-524 engine. Jon The RB211-524B has only 5 pounds of thrust, that one is suitable for the H and G models acording to http://www.rolls-royce.com/media/product/rb211524.jsp I have generated a 5 pound version, and I can get TSFC values and stuff , but I don't understand those tables below. I also had problems using the FG_TURBINE and switched to FG_SIMTURBINE, is this a good idea? Yes, this reflects a recent change in JSBSim, which hasn't traversed it's way back into aeromatic yet. The file generated by aeromatic had the AC_THRUSTER inside the AC_ENGINE tags and it didn't work until i separated them, aero-matic bug? Not really a bug: This how JSBSim expected things to be, until a couple of weeks ago. It took me a while to figure this one out. ;-) Now i can get the L1011 working, but if i start from the RWY, it appears around 50 meters AGL and starts rotating backwards until it gets under the floor and freezes. If I start it in flight it behaves nice. Aarggg, this sounds exactly like how the MD11 was behaving initially, and which I thought was fixed. Moving the main gears forward and backward a little bit might help. Try have a look at the MD11 and related threads in the FlightGear-Devel mailing archives for some more information (April through Jue of this year IIRC). What are the conventions for propper positioning of the model? Is it better to use a singles texture, or separate fuselage texture from other pieces? thanks in adavance :) Regards, Tiago Cheers, Durk For those of you wondering, I've collected quite a few changes to the MD11, but last weeks were fairly busy, so I didn't get a chance to do the final polising up and send them out for CVS inclusion. Hopefully this weekend I'll get to it. Cheers, Durk ___ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel
Re: [Flightgear-devel] L1011-500
On Tue, 6 Jul 2004 20:09:38 +0200 Durk Talsma [EMAIL PROTECTED] wrote: Just a few comments below, since I also ran into similar issues creating and updating the MD-11 aero file. I also had problems using the FG_TURBINE and switched to FG_SIMTURBINE, is this a good idea? Yes, this reflects a recent change in JSBSim, which hasn't traversed it's way back into aeromatic yet. Actually, I did make the change - it should be implemented in Aeromatic. I can't test it right now. The file generated by aeromatic had the AC_THRUSTER inside the AC_ENGINE tags and it didn't work until i separated them, aero-matic bug? Not really a bug: This how JSBSim expected things to be, until a couple of weeks ago. It took me a while to figure this one out. ;-) I thought I had made a big announcement. If I did not, my apologies. Jon ___ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel
RE: [Flightgear-devel] L1011-500
Thank you all for your help and comments :) I think i can keep the triangles around 8000 for this second version, since the first version of the model had a lot of redundant triangles, and I forgot to take that in acount when i said 1! The editor used was Milkshape, so i will be using .3DS you can take a look at the wireframe: http://tiagogusmao.home.sapo.pt/l1011/vert.jpg You can see almost everything is done by inserting cylinders and doing some vertex manipulation, perhaps not the best approach, but seems to work :P The red vertices are redundant, and will be merged later on. For the texture, I think 1 512x512 will do fine :) For the FDM, for now i'll stick with JSBSim, since it seems to be the easiest. Is it ok to exchange files here, or should I learn to use CVS? Regards, Tiago Gusmão O SAPO já está livre de vírus com a Panda Software, fique você também! Clique em: http://antivirus.sapo.pt ___ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel
Re: [Flightgear-devel] L1011-500
On Friday 02 July 2004 13:54, [EMAIL PROTECTED] wrote: Is it ok to exchange files here, or should I learn to use CVS? Probably not a good idea to send the file to list. I'm sure one of the guys who has CVS access will ask you to email them the files soon, so they can put it into CVS. Look forward to seeing in the air. Cheers, Al ___ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel
Re: [Flightgear-devel] L1011-500
On July 1, 2004 10:14 pm, [EMAIL PROTECTED] wrote: Hi I am almost finishing a L1011-500 TriStar 3D model, and i'd like to contribute it to FlightGear. History is repeating itself: two different tri-jets are introduced at the sametime. ;-) The editor used was Milkshape, so i will be using .3DS Make sure the extension is in lower case: .3ds The model looks good, though the area around the windshield seems to need some improvement. By the way, what type of engines do the tristar uses? Regards, Ampere ___ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel
Re: [Flightgear-devel] L1011-500
On Friday 02 July 2004 21:02, Ampere K. Hardraade wrote: By the way, what type of engines do the tristar uses? This is a useful website for aircraft data. Unfortunately, it only list data for the L1011-100 and not the -500 series. http://www.bh.com/companions/034074152X/appendices/data-a/table-6/table.htm I also refer to this site a lot: http://www.aerospaceweb.org/aircraft/ Unfortunately, it doesn't list the L1011, but has lot's of useful data to build basic Airbus, Boeing, and Douglas Jetliner FDMs using aeromatic. This page actually lists FlightGear as one of it's main referrers. :-) Finally, lots of useful fuel load/tank configuration information can be found here: http://www.simviation.com/rinfofuel.htm Cheers, Durk ___ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel
Re: [Flightgear-devel] L1011-500
Very nice model indeed. Looks like a nice addition to FlightGear's small but continously growing fleet of 3-engined wide-body jetliners. :-) For the FDM, for now i'll stick with JSBSim, since it seems to be the easiest. I have good experiences creating a basic FDM for the MD11 using aeromatic. The earomatic script creates an xml file that you can later edit to tweak. My experience is that you most likely want to change the engine file and fuel tank sections a bit to get the overall performance and range in the right ballpark, but other than that the default aeromatic does a reasonable job (note, I'm neither a pilot nor an aeronautical engineer, so my idea of *reasonable* might actually be quite unreasonble :-)) Is it ok to exchange files here, or should I learn to use CVS? Unless you're sending out small patches for a second opinion, it's best to send your work to one of the developers with CVS write access and ask if they'd be willing to add your contribution to cvs. I'm not sure who actually have write permission these days. Cheers, Durk ___ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel
Re: [Flightgear-devel] L1011-500
History is repeating itself: two different tri-jets are introduced at the sametime. ;-) Indeed! The MD-11 too is a very nice tri-jet, i'm looking forward to see it in action :) The editor used was Milkshape, so i will be using .3DS Make sure the extension is in lower case: .3ds The model looks good, though the area around the windshield seems to need some improvement. Yes, the windshield is not very accurate, the L1011 is a very complicated aircraft IMHO, and my editor doesn't help too much but if after texturing, something doesn't look right, I'll redo that part By the way, what type of engines do the tristar uses? The L1011-500 uses the Rolls-Royce RB211-524B take a look here for details: http://www.aircraftenginedesign.com/TableB3.html I have good experiences creating a basic FDM for the MD11 using aeromatic. The earomatic script creates an xml file that you can later edit to tweak. My experience is that you most likely want to change the engine file and fuel tank sections a bit to get the overall performance and range in the right ballpark, but other than that the default aeromatic does a reasonable job thanks for the links and info, I've used aero-matic today, the process is very straight-forward and it's a good way to get started with a working file, it's a nice tool :) (note, I'm neither a pilot nor an aeronautical engineer, so my idea of *reasonable* might actually be quite unreasonble ) Me neither, but I've met a few online, if you have any trouble, i can ask them to enlighten us :) Regards, Tiago Gusmão O SAPO já está livre de vírus com a Panda Software, fique você também! Clique em: http://antivirus.sapo.pt ___ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel
Re: [Flightgear-devel] L1011-500
On Fri, 2 Jul 2004 22:46:14 +0100 [EMAIL PROTECTED] wrote: The L1011-500 uses the Rolls-Royce RB211-524B take a look here for details: http://www.aircraftenginedesign.com/TableB3.html I have committed to JSBSim CVS an RR RB211-524 engine. Jon ___ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel
[Flightgear-devel] L1011-500
Hi I am almost finishing a L1011-500 TriStar 3D model, and i'd like to contribute it to FlightGear. You can take a look at these screenshots: http://tiagogusmao.home.sapo.pt/l1011/finalv1.jpg this one is my first try, there are missing parts, and the tail is horrible. Current version is this one: http://tiagogusmao.home.sapo.pt/l1011/tail-hstab.jpg It's just a matter of finishing the tail, and add small details It has around 1 tris and 7000 vertices, is that ok for FG? What are the texture requirements/limits/recommendations for FG? What technical data do we need to make it fly like the real one? Any good tips or advice are welcome :) This is my first time using a mailing list, so if i mess up something, please warn me :P Thanks, Tiago Gusmão O SAPO já está livre de vírus com a Panda Software, fique você também! Clique em: http://antivirus.sapo.pt ___ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel
Re: [Flightgear-devel] L1011-500
tiagogusmao said: snip You can take a look at these screenshots: http://tiagogusmao.home.sapo.pt/l1011/finalv1.jpg this one is my first try, there are missing parts, and the tail is horrible. Current version is this one: http://tiagogusmao.home.sapo.pt/l1011/tail-hstab.jpg It's just a matter of finishing the tail, and add small details It has around 1 tris and 7000 vertices, is that ok for FG? Looks like they (the vertices) have been wisely used. Nice job, good detail. If you have a doubts compare to some of the other models. That many vertices is a little heavy, but first impression from the screenshots is the usage is not unreasonable. Just make sure you aren't wasting any. Best, Jim ___ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel
RE: [Flightgear-devel] L1011-500
Hi Tiago From: tiagogusmao writes Hi I am almost finishing a L1011-500 TriStar 3D model, and i'd like to contribute it to FlightGear. Looks nice what 3D program are you using. If it is AC3D have you put the model through the optimise vertices/surface routine to trim some of the fat.One of the places to start is with the fuselage.Do you have the center section divided up onto anymore than one section ifso you can cut it down to one long section.Also with the wings and tail surfaces you can quite easely add surfaces that dont realy make a great difference in the sim It has around 1 tris and 7000 vertices, is that ok for FG? Getting a bit on the big side. What are the texture requirements/limits/recommendations for FG? One 1024x1024 texture would be getting on the big side What technical data do we need to make it fly like the real one? Any good tips or advice are welcome :) Just the data that is with the aircraft types on airliners.net should be enough to get you close using the AEROMATIC program on JSBSIM.It is the one I find easiest to use. Thanks, Tiago Gusmão Cheers Innis _ Smart Saving with ING Direct earn 5.25% p.a. variable rate: http://ad.au.doubleclick.net/clk;7249209;8842331;n?http://www.ingdirect.com.au/burst6offer.asp?id=8 ___ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel