Jump to content


Photo
* * * * * 4 votes

The Fokker Dr1 FM Revisited


  • Please log in to reply
197 replies to this topic

#1 Chill31

Chill31
  • Posts: 1891

Posted 20 March 2015 - 03:48

So when ROF was in its infancy, NeoQB created a Fokker Dr1 model for MS Flight Sim X.  It looks very much like the ROF Dr1, but it has been so long since I flew it, I cannot yet comment on the FM.  I recently found out that MSFS X has a SDK where users can edit the FM parameters of the aircraft in the game. 

 

I am going to use this thread to document my FM editing journey for the NeoQB Dr1.  I think their License Agreement prevents me from distributing a modified version of their Dr1, but if I come up with something really nice, I will ask them (777 now?) if I can share it with the community or give it to them to share. For now, we will see how closely I can get it to resemble the Dr1 I have flown.

 

More to follow as it develops...


  • 1

#2 1PL-Husar

1PL-Husar
  • Posts: 557

Posted 20 March 2015 - 06:01

Good luck and good idea to make reference for others who would like to implement realistic DR.1 in their games.
  • 0

34zg6jd.jpg


#3 ZachariasX

ZachariasX
  • Posts: 794

Posted 20 March 2015 - 09:46

Good luck! It would be great if you manage to tweak it. It requires a lot of tweaking. It is reasonably stable longitudinally, laterally it has little adverse yaw and it hates to do a full roll to the left. The gyro effect feels weird. It's nothing like the Dr.I we have in RoF. It's not even particularly quitky, it just feels like you have to pull it by the hair in the direction you want to go. A Piper Cub feels like an F-22 compared to it.


  • 0

#4 Chill31

Chill31
  • Posts: 1891

Posted 20 March 2015 - 19:47

Ok, so I tried out the NeoQB Dr1 for FSX.  I can see there are some definite similarities as well as some significant differences.  There are a few things that the ROF Dr1 does that really disappoint me.

 

1) It requires a lot of down elevator to fly level.  Much more than I have seen in any Dr1 flying today.

2) It is far too stable in yaw

3) It has insufficient adverse yaw

 

After looking at the aircraft parameters and aerodynamics files in FSX, I have a lot of room to tweak the FM.  If ROF uses something similar, I think I will be able to provide useable information for making the changes relatively easy if they want to make them in ROF.

 

In just a few minutes, I managed to change the horizontal stabilizer angle of incidence to reduce the amount of nose down elevator deflection required.  In FSX, the horizontal AOI was set at -1 degree.  On the actual plane, it is at +4.7 degrees (causes a nose down pitching moment).  I didn't get it to settle in on a particular angle yet, since I am learning to edit the files and how to save them (setting file permissions and such). 

 

My first goal will be to get the plane to fly in a slip without my input.  That would show that I can get the yaw stability down to realistic levels.  Second, I will try to increase the adverse yaw.  If I can get those two things to work properly, it would warrant continued work on the FM to see if I can get the performance parameters to match historical data.

 

For anyone interested in modifying FSX files, here are some tools I have found helpful and free to use:

You need this one to edit .AIR files: https://ootb.wordpress.com/aam/ (this program has nice definitions of the many coefficients found in the files)

This website also has ok definitions: https://ootb.wordpress.com/aam/

Here is a list of things in the AIR file from MS: https://msdn.microso...y/cc707067.aspx

 

Finally, there is a .CFG file for each aircraft which stores physical dimensions and scalars (for aerodynamics in the .AIR file) for the aircraft.  FSX uses a otto cycle model to give engine power/rpm.  It accounts for engine bore and stroke, mechanical efficiency, etc.  If the ROF guys are using a model similar to this, I can see how it would be a very time consuming process to tweak each plane to fly just as it should.  For me, I guess that will be part of the fun!


  • 1

#5 =HillBilly=

=HillBilly=
  • Posts: 5605
  • LocationSouthern Ozark Mountains

Posted 20 March 2015 - 20:00

Chill I do have some questions on the 3 points.

1 no questions, and I agree

2 would a rotary make a differences compared to a inline engine?

3 again would a rotary make a differences compared to a inline engine, you know gyroscopic effect keeping adverse yaw to a minimum?     .


  • 0

     So Long, and Thanks for All the Fish

 
 


#6 piecost

piecost
  • Posts: 1318

Posted 20 March 2015 - 20:13

Chill,

 

I look forward to your future posts and will be intrigued so see how you get on. I wonder if you will come up against limitations in modelling in FSX which are not present in RoF? For example; should adverse yaw become larger with increased incidence? Do you notice this on your real DRI? Can FSX model this?

 

As far as engine modelling, in the SE5a FM Revision thread A.P. used a curve of power (thrust & torque) versus rpm. This being more accurate than to estimate from parameters using equations. You might end up getting similar graphs for the engine and fidling the FSX parameters to match them.

 

Are you going to model the original DRI with rotary or your DRI?


  • 0

#7 FourSpeed

FourSpeed
  • Posts: 1746

Posted 20 March 2015 - 20:23

Sounds like an interesting little project, Chill.  I'll be curious to see how close you can get with it.

 

I did some dabbling with "custom" planes for FSX a few years ago, and you can get some pretty
radical behaviors by fiddling around with those files (esp. if you have no idea what some of those
parameters actually control - which was my case at the time).  :icon_eek:

 

Have Fun!

 

Regards,

4 :icon_e_salute:


  • 0

#8 ZachariasX

ZachariasX
  • Posts: 794

Posted 20 March 2015 - 21:28

Hi Chill

 

I'm glad you got the things working. With the Neoqb Dr.I it appears to me as well that basically everything is written in the Dr.I.AIR file. So there are not further things to take into account as how FSX interprets the .AIR file. So that's just one area to play. And also just one thing to replace for updated FM. The license explicitly forbids a distribution of an altered version of this Dr.I. Independent distribution an  .AIR file however has not much to do with that.

 

So then... I'd be very interessted for your new .AIR file, once you came up with something convincing!

 

Tumbs up!

 

:icon_e_salute:


  • 0

#9 Chill31

Chill31
  • Posts: 1891

Posted 21 March 2015 - 01:43

Chill I do have some questions on the 3 points.

1 no questions, and I agree

2 would a rotary make a differences compared to a inline engine?

3 again would a rotary make a differences compared to a inline engine, you know gyroscopic effect keeping adverse yaw to a minimum?     .

1) is fixed.  I have it where it no longer requires so much down elevator to fly level

2) the airplane's reaction to yaw would be different.  Pushing the rudder would yield a pitch up/down depending on the direction you push the rudder.

3) I would say the same thing.  Adverse yaw will be present, but it may be felt more in the pitch axis than in yaw with a rotary.

 

Chill,

 

I look forward to your future posts and will be intrigued so see how you get on. I wonder if you will come up against limitations in modelling in FSX which are not present in RoF? For example; should adverse yaw become larger with increased incidence? Do you notice this on your real DRI? Can FSX model this?

 

As far as engine modelling, in the SE5a FM Revision thread A.P. used a curve of power (thrust & torque) versus rpm. This being more accurate than to estimate from parameters using equations. You might end up getting similar graphs for the engine and fidling the FSX parameters to match them.

 

Are you going to model the original DRI with rotary or your DRI?

I think ROF is similar enough to FSX that I will not run into any limitations.  The core behavior of the NeoQB Dr1 is the very similar to FSX.  I have done a few experiments and can make the ROF Dr1 using FSX.  Adverse yaw does increase with increasing AOA and I do notice very much in my own Dr1.  I haven't gotten into the adverse yaw portion yet, but it looks like the parameters are there for the modding...

 

I remember AP with an engine power curve on his desktop in a screen shot one time.  If ROF uses a similar engine model as FSX, then the only purpose of the graph would be to come up with tab data for the tables in the FM files.  I have Oberursel power charts which might yield torque and power coefficients, but I'll have to check them later.  There are so many variables in an engine that, without actual testing of an engine, I don't think I will be able to get a truly accurate model of the Oberursel.  My alternative would be to tweak engine/aerodynamics until the performance matches what data I can find.

 

I am going to model this after my Dr1.  There are functions in the code for gyroscopic moments, but I just want to see if I can get it to behave like something I know before I try to make it something I don't know.


  • 0

#10 Chill31

Chill31
  • Posts: 1891

Posted 21 March 2015 - 01:49

I have been experimenting with yaw stability and rudder authority this evening, and I have made major progress.  I think that I can say that I will be successful in modeling the yaw characteristics of the Dr1.  If anyone can fill me in on how to make a video, I will try to make a before and after shot so you can see the difference on youtube.  For now, just know that it is possible to make it like the video of my Dr1 doing the flat turn on youtube, here at 3:20: https://www.youtube....h?v=9_USshb6BE0


  • 0

#11 unreasonable

unreasonable
  • Posts: 1452
  • LocationBangkok

Posted 21 March 2015 - 03:34

Very interesting progress. If you can get the model flying like your Dr1 first with a virtual stationary engine, that is a great test of the accuracy of your inputs. The rotary effects can be estimated later.


  • 0

"Democracy is the theory that the common people know what they want, and deserve to get it good and hard." H. L. Mencken


#12 =HillBilly=

=HillBilly=
  • Posts: 5605
  • LocationSouthern Ozark Mountains

Posted 21 March 2015 - 03:37

Thank you Chill for your answers, and please keep up the work.


  • 0

     So Long, and Thanks for All the Fish

 
 


#13 ZachariasX

ZachariasX
  • Posts: 794

Posted 21 March 2015 - 07:43

1) is fixed.  I have it where it no longer requires so much down elevator to fly level

2) the airplane's reaction to yaw would be different.  Pushing the rudder would yield a pitch up/down depending on the direction you push the rudder.

3) I would say the same thing.  Adverse yaw will be present, but it may be felt more in the pitch axis than in yaw with a rotary.

 

I think ROF is similar enough to FSX that I will not run into any limitations.  The core behavior of the NeoQB Dr1 is the very similar to FSX.  I have done a few experiments and can make the ROF Dr1 using FSX.  Adverse yaw does increase with increasing AOA and I do notice very much in my own Dr1.  I haven't gotten into the adverse yaw portion yet, but it looks like the parameters are there for the modding...

 

I remember AP with an engine power curve on his desktop in a screen shot one time.  If ROF uses a similar engine model as FSX, then the only purpose of the graph would be to come up with tab data for the tables in the FM files.  I have Oberursel power charts which might yield torque and power coefficients, but I'll have to check them later.  There are so many variables in an engine that, without actual testing of an engine, I don't think I will be able to get a truly accurate model of the Oberursel.  My alternative would be to tweak engine/aerodynamics until the performance matches what data I can find.

 

I am going to model this after my Dr1.  There are functions in the code for gyroscopic moments, but I just want to see if I can get it to behave like something I know before I try to make it something I don't know.

 

I think it is really the way to go to first make the Dr.I fly like the one you have. Especially if you gets hands on a rotary powered type later on. Like this you can make a clean profile for the airframe itself. Basically you have one big variable less to design propperly. Once you have the airframe modelled correctly, all you have to add are the parameters for the added torque and gyro.

 

To get individual variables right, separating those for the analysis is always the way to go.

 

I'd be thrilled getting the new .AIR file. And you can make two. One for an inline powered variant, such as yours, plus then one for a rotary powered type. This would also be vry instructive to demonstrate a direct comparison of rotary vs. inline flying propperties. To give some rationale to a sometimes "very colorful discussion" on what rotaries do vs. inline engines.

 

Cheers,

Z


  • 0

#14 Chill31

Chill31
  • Posts: 1891

Posted 21 March 2015 - 16:49

There are parameters for gyroscope precession in both pt ch and yaw. For the moment, I have them zeroed out.

Right now I am trying to eliminate the extreme roll moment at high fuselage side slip angles. I am getting close, but there are several tables of data to edit so that it behaves appropriately.

After that, I will be able to completely model the low yaw stability.
  • 0

#15 Chill31

Chill31
  • Posts: 1891

Posted 25 March 2015 - 17:16

Ok, so I found a program called AirWrench which helped me create a new Dr1 (not a new 3D model though).

 

I now have the Dr1 where the yaw is believable, though I would not yet go so far as to say this is how the plane actually flies.  I made a lot of progress, but this is a slow, tedious process of making an edit and observing the results.  It is not so simple as putting in real world figures and out pops this perfect FM.

 

For starters, and I suspect ROF is this way as well, the FM doesn't actually model 3 wings!  It models one wing only, so I have to adjust it to fly like 3 wings with relation to stall, roll, yaw, CoG etc.

 

Currently I am focused on getting an accurate yaw model.  The Dr1 is very unstable in yaw and can achieve some incredible sideslip angles and keep flying.  In the FM, as side slip increases, stall margin decreases which results in wild stall gyrations at the moment.  I am trying to solve those.

 

Currently, I have successfully modeled the wandering nose and adverse yaw found on the real aircraft.  If you fly strait and don't use the rudder, she just wont go where you want her to go.  It really is a rudder airplane now as you start and stop a turn takes a significant amount of rudder which approximates the real aircraft.

 

Each part I adjust causes me to have to adjust another 2 or 3 parts to get everything back in balance, so this is truly a time consuming process.  I do think I will be able to deliver a realistic portrayal of the Dr1 using FSX.


  • 0

#16 ZachariasX

ZachariasX
  • Posts: 794

Posted 25 March 2015 - 21:03

Ok, so I found a program called AirWrench which helped me create a new Dr1 (not a new 3D model though).

 

I now have the Dr1 where the yaw is believable, though I would not yet go so far as to say this is how the plane actually flies.  I made a lot of progress, but this is a slow, tedious process of making an edit and observing the results.  It is not so simple as putting in real world figures and out pops this perfect FM.

 

For starters, and I suspect ROF is this way as well, the FM doesn't actually model 3 wings!  It models one wing only, so I have to adjust it to fly like 3 wings with relation to stall, roll, yaw, CoG etc.

 

Currently I am focused on getting an accurate yaw model.  The Dr1 is very unstable in yaw and can achieve some incredible sideslip angles and keep flying.  In the FM, as side slip increases, stall margin decreases which results in wild stall gyrations at the moment.  I am trying to solve those.

 

Currently, I have successfully modeled the wandering nose and adverse yaw found on the real aircraft.  If you fly strait and don't use the rudder, she just wont go where you want her to go.  It really is a rudder airplane now as you start and stop a turn takes a significant amount of rudder which approximates the real aircraft.

 

Each part I adjust causes me to have to adjust another 2 or 3 parts to get everything back in balance, so this is truly a time consuming process.  I do think I will be able to deliver a realistic portrayal of the Dr1 using FSX.

 

I'm really looking forward to what you achieve Chill! Regarding the inner workings of the sim, I always suspected the engine modelling monoplanes only. Adding bi- or triplanes makes things a hell lot more complicated with little to gain, as stall behaviour is scripted anyway. 

 

:icon_e_salute: 

 

Z


  • 0

#17 Chill31

Chill31
  • Posts: 1891

Posted 25 March 2015 - 22:11

FWIW I don't think stall behavior is scripted in FSX. I have control of all of the aerodynamics affecting stall behavior. I think it just takes a lot of tuning to get post stall behavior correct. A Petrovich did a nice job with ROF. It is a long process to get everything just right
  • 0

#18 Chill31

Chill31
  • Posts: 1891

Posted 26 March 2015 - 02:53

Regarding the Gyro effects implementation...

 

Within FSX, and I suspect it is the same within ROF, I have the ability to edit the gyro interaction with pitch, roll, and yaw.  I simply adjust the value of a scalar and it has a corresponding effect on the aircraft.  I have no way to calculate the magnitude for FSX.  I am only able to say "that looks right" when I have adjusted it to my liking.

 

With the Dr1 FM I am creating, I will make one with and one without Gyro effects, but the gyro effects will only be as good as I can recreate from watching Dr1s on youtube until I get to fly one.


  • 0

#19 Bando

Bando
  • Posts: 146

Posted 26 March 2015 - 08:21

A monks job it is (as we say in Holland). 

Thanks for putting the effort in this.

I hope it will result in some improvements to flightsims in general and ROF in particular.


  • 0

#20 Chill31

Chill31
  • Posts: 1891

Posted 27 March 2015 - 02:32

Ok, so the instructions I was using to learn how to modify the FM was misleading as it is translated from another language.  I have figured out what I wanted though and, after going back to square one, made some solid progress in getting the yaw model to be representative of the actual airplane.

 

If I can the yaw squared away, I think the rest of the FM will come together quickly since those parameters are much more straight forward.

 

FWIW, I have had to work through some of the problems I see with the ROF Dr1...so if I can get this one to work in FSX, I think 777 could use my data to improve their Dr1. 


  • 0

#21 hq_Reflected

hq_Reflected
  • Posts: 4711

Posted 27 March 2015 - 06:34

Good luck Chili, and keep us posted!  :icon_e_salute:


  • 0

#22 Chill31

Chill31
  • Posts: 1891

Posted 29 March 2015 - 03:02

Ok, so I think I know why the ROF Dr1 is too stable in yaw...its because it cant be done with FSX code.

 

I need a parameter to relate "weathervane" effect to side slip angle.  With very little side slip, the Dr1 should want to yaw on its own until it gets enough side slip to cause it to weathervane back into the on coming air.  FSX flight models done have such a parameter (at least not that I can find.  If anyone knows, I'm all ears!) 

 

The result is that I have either a completely unstable aircraft in yaw or a completely stable aircraft in yaw.  This is also the same phenomenon we see in ROF.

 

My debate is whether or not to continue making pretty good FSX Dr1 or abandon it since I don't think FSX has the capability to be perfect. 


  • 0

#23 =HillBilly=

=HillBilly=
  • Posts: 5605
  • LocationSouthern Ozark Mountains

Posted 29 March 2015 - 03:11

Chill just a little experiment, remove the centering spring out of your rudder peddles and see if the DR1 wants to drift about the yaw axis.  :icon_e_salute:


  • 0

     So Long, and Thanks for All the Fish

 
 


#24 Chill31

Chill31
  • Posts: 1891

Posted 29 March 2015 - 03:28

Chill just a little experiment, remove the centering spring out of your rudder peddles and see if the DR1 wants to drift about the yaw axis.  :icon_e_salute:

 

It may very well do that.  The real Dr1 will yaw uncommanded with the rudder locked in position though.  You can see here at 2:50 what I am talking about: https://www.youtube....h?v=9_USshb6BE0

 

I think I can still make a nice Dr1 model, but it wont have all of the yaw characteristics of the real plane.


  • 0

#25 =HillBilly=

=HillBilly=
  • Posts: 5605
  • LocationSouthern Ozark Mountains

Posted 29 March 2015 - 12:21

 

 

The real Dr1 will yaw uncommanded with the rudder locked in position though

What do you mean locked, is that wise practice? 


  • 0

     So Long, and Thanks for All the Fish

 
 


#26 ZachariasX

ZachariasX
  • Posts: 794

Posted 29 March 2015 - 12:32

What do you mean locked, is that wise practice? 

Kept centered. If you keep the rudder straight, the plane will start to yaw on it's own. Thus, you constantly have to use the rudder to actively keep the nose where you want it.

 

In the Dr.I, it is consequently NOT a whise practise. In the Spitfire (or many well behaved aircraft) you can let the rudder straight and a well trimmed plane will fly straight. There it is wise practise, as it might keep you from unintentional yaw. I say, keep it straight, in regular flight. LOCKED, you keep it tied down on the dground to avoid wind damage ;)


  • 0

#27 =HillBilly=

=HillBilly=
  • Posts: 5605
  • LocationSouthern Ozark Mountains

Posted 29 March 2015 - 13:01

I do enjoy discussion but I think misunderstood my question, how is it locked? Is it locked  by a mechanical means, or is locked by your legs and feet?   


  • 0

     So Long, and Thanks for All the Fish

 
 


#28 ZachariasX

ZachariasX
  • Posts: 794

Posted 29 March 2015 - 17:25

I do enjoy discussion but I think misunderstood my question, how is it locked? Is it locked  by a mechanical means, or is locked by your legs and feet?   

 

Legs and feet.


  • 0

#29 =HillBilly=

=HillBilly=
  • Posts: 5605
  • LocationSouthern Ozark Mountains

Posted 29 March 2015 - 18:41

Legs and feet.

Ok do you have rudder peddles? If so try this go into mission pick a plane, open responses select yaw then try to hold 30 degrees, how long before it starts to drift? Do not fly just sit on the ground.   


  • 0

     So Long, and Thanks for All the Fish

 
 


#30 Chill31

Chill31
  • Posts: 1891

Posted 29 March 2015 - 19:46

When I "lock" the rudder in place in the plane, I put my heels on the floor and then press the ball of my foot on the rudder bar.  Doing that, I can keep the rudder in place pretty well.  Certainly better than  the yaw which comes in as a result of not moving them.  You can also see in my dual view video that I am pretty decent with the rudders in the Dr1 and the ball stays pretty well centered during my maneuvers.  I am very sure that the yaw in the video is not due to pilot inability to fly the airplane.


  • 0

#31 ZachariasX

ZachariasX
  • Posts: 794

Posted 29 March 2015 - 20:53

Ok do you have rudder peddles? If so try this go into mission pick a plane, open responses select yaw then try to hold 30 degrees, how long before it starts to drift? Do not fly just sit on the ground.   

 

Just taking the foot from the pedals will keep it centered with the saitek. In modern real planes, you can also take the foot off the rudder and it will keep it straight, provided it is reasonably trimmed. It is only on flying tails, where controls are so light that it is not advisable to do so, especially on the elevator. This is why Piper provides an anti-servo tab for the Cherokee and Archers etc. The Morane Monoplane is a nasty example, as it had almost no stick force on the elevator (balanced flying tail) and if you let it go, it would fall all the way forward out of your reach. In full flight. But that is as shitty as it goes. Bottom line is, you can hold the rudder reasonably well straight without fatige, especially when you do it like Chill does. Keeping it in a deflected state is a different sport.


  • 0

#32 Chill31

Chill31
  • Posts: 1891

Posted 30 March 2015 - 03:10

Ok, so I finally got the yaw down to something resembling the Dr1, though it is very stable.

I also got the roll rate and adverse yaw to something resembling the Dr1.

Yay for those 2.

I also modeled my Lycoming engine, and it seems to be a pretty decent match. Though it climbs a little too fast at the moment : 1400 fpm


I am now trying to tackle pitch, to include stall behavior. This is a pretty big task, so it will take a while to get it all worked out.
  • 1

#33 hq_Reflected

hq_Reflected
  • Posts: 4711

Posted 30 March 2015 - 05:43

Fantastic work, Chili! And thanks for posting that video, it's very insightful especially with all the parameters displayed in the top left corner. The way you turned 90 degrees without banking is astonishing. I still can't do that in RoF. Somehow it looks much less stable on the yaw axis than in RoF.

 

Anyway, good luck and thanks again!  :icon_e_salute:


  • 0

#34 Chill31

Chill31
  • Posts: 1891

Posted 02 April 2015 - 17:50

Ok, so I have learned a lot about engine/propeller modeling from working with FSX.  I thought I would share what I've learned as I think it applies to ROF.

 

First, the engine model itself.  The basic model is an ideal otto cycle engine.  This means you input all of the engine parameters, bore, stroke, max rpm, max hp.  The model produces heating values, torque/power values.  Now, if I were to put this ideal engine into an airplane, it would have way too much horsepower.  Engines lose a lot of efficiency due to imperfect seals and friction.  So I then go into the model and adjust for mechanical efficiency of the engine which for most engines is 50-70% depending on RPM.  In FSX, I have one final adjustment for power losses due to friction related to RPM. 

 

With that complete, I now have an engine that may or may not be exactly right, since I have no way to know if the output is realistic or not.

 

So I start making my propeller.  In my case, I used some generic propeller efficiency and power charts as a function of advance ratio to make my propeller.  I also made an excel sheet to calculate the coefficient of power based on my actual engine and prop using the actual engine power charts.  The result of using real data is that I now have a known point upon which I can adjust the generic propeller efficiency and power curves. 

 

I tuned the engine power to provide actual RPM observed in my Dr1.  The result is that the FSX Dr1 now has the same topspeed and climb rate as my Dr1. 

 

I thought that was pretty cool.  I am still working on high aoa/stall behavior which requires more learning on my part as to the effect of modifying each variable in the FM.


  • 2

#35 ZachariasX

ZachariasX
  • Posts: 794

Posted 02 April 2015 - 19:14

Ok, so I have learned a lot about engine/propeller modeling from working with FSX.  I thought I would share what I've learned as I think it applies to ROF.

 

First, the engine model itself.  The basic model is an ideal otto cycle engine.  This means you input all of the engine parameters, bore, stroke, max rpm, max hp.  The model produces heating values, torque/power values.  Now, if I were to put this ideal engine into an airplane, it would have way too much horsepower.  Engines lose a lot of efficiency due to imperfect seals and friction.  So I then go into the model and adjust for mechanical efficiency of the engine which for most engines is 50-70% depending on RPM.  In FSX, I have one final adjustment for power losses due to friction related to RPM. 

 

With that complete, I now have an engine that may or may not be exactly right, since I have no way to know if the output is realistic or not.

 

So I start making my propeller.  In my case, I used some generic propeller efficiency and power charts as a function of advance ratio to make my propeller.  I also made an excel sheet to calculate the coefficient of power based on my actual engine and prop using the actual engine power charts.  The result of using real data is that I now have a known point upon which I can adjust the generic propeller efficiency and power curves. 

 

I tuned the engine power to provide actual RPM observed in my Dr1.  The result is that the FSX Dr1 now has the same topspeed and climb rate as my Dr1. 

 

I thought that was pretty cool.  I am still working on high aoa/stall behavior which requires more learning on my part as to the effect of modifying each variable in the FM.

 

That's really cool! Thumbs up. ....and I cant wait getting a hold of your .air file...

 

:icon_e_salute:

Z


  • 0

#36 Chill31

Chill31
  • Posts: 1891

Posted 02 April 2015 - 22:53

I worked on it a bunch today and learned a lot about using real world data in an imperfect FM.  In editing lift coefficients and pitching moments, I found that if I used the actual data from an airfoil section (Goe 298 for the Dr1), the plane did not fly in the virtual world as it does in the real world.  In the virtual world, guess what!  It flew EXACTLY like the ROF Dr1.  So, I guess there is something to be said for the ROF guys using real data, but there is a major disconnect between this artificial virtual world and what really happens.  I think that is where people like me could really help them with their FMs.

 

What I had with inputting real data was a Dr1 that required a LOT of nose down elevator, and my elevator looked exactly like it does in ROF.  I adjusted the pitching moment from the horizontal stab (there are a lot of pitching moments to adjust which could affect the overall pitching moment of the Dr1. So for the ROF Dr1 to be pitching up as aggressively as it is...is an easily correctible error) to fix the elevator position and forward stick pressure.  I am using a TM warthog with a stick extension, and the feel of the controls reminds me of my Dr1.

 

I modified the max lift coefficient to give me the desired stall speed and stall characteristics I have in my Dr1.  So now it stalls at about 45 mph and mushes into a level descending attitude with about 1,500 fpm descent rate. 

 

Where I am right now over all...Max level speed 107 mph, climb rate 1000 fpm at SL at 75 mph, it loops in 10 seconds with .5 Gs on top, rolls in 6 seconds and does a 360 degree max rate turn in 8 seconds, it does 90 degrees of flat rudder turn.  Those numbers match my Dr1 perfectly.  For me, this validates my zero lift drag coefficient (which is a major factor in determining top speed, regardless of engine). I should be able to create a rotary powered version using the power charts I have for the Oberursel URII.

 

So far, I am really pleased with how it is progressing, and I really hope the ROF guys can use the finished product to refine their Dr1.

 

If you have questions about working on a FM or how things work in the FM, let me know!


  • 2

#37 piecost

piecost
  • Posts: 1318

Posted 02 April 2015 - 23:31

Chill, it is great to hear of your progress.

 

I am really impressed with your FM performance matching to your real plane.

 

I would be interested to see a few points from your drag polar to compare to those derived from comtempory British flight tests (see attached). I recall a pilot account of the DRI being low drag relative to conventional wire braced planes of the time.

I also attach a wind tunnel test result of a model Fokker DVII. (The wind tunnel data does not nessesarily represent reality, but is interesting).

 

I also attach a report on gyro couples. You might consider scaling the inertia data for the 80HP engine and propeller to the DRI rotary engine/prop.

 

One question on the FM; is the FM limited to linear aero coefficients linear with yaw angle? Or can you add "non-linearities" at extreme angles (similar to the non-linearity of an aerofoil stalling). I get the impression that the RoF FM feels so rich due to the non-linearities at high angles of attack (things such as the reduction in aileron control near the stall)

Attached Files


  • 0

#38 Chill31

Chill31
  • Posts: 1891

Posted 03 April 2015 - 01:13

One question on the FM; is the FM limited to linear aero coefficients linear with yaw angle? Or can you add "non-linearities" at extreme angles (similar to the non-linearity of an aerofoil stalling). I get the impression that the RoF FM feels so rich due to the non-linearities at high angles of attack (things such as the reduction in aileron control near the stall)

 

The short answer, is no it is not limited to linear responses.  It is limited to some degree by the fact that it is designed for stable aircraft.  (I did just have an idea as I write this to try a yaw dampener function to help with modeling low yaw stability...have to see if it works)  I am unable to accurately model the low yaw stability of the Dr1 because I have to define it either stable in yaw or unstable.  I can't transition the stability with increasing yaw angles.

 

However, I can reduce aileron, rudder, and elevator effectiveness at different AOAs.  For example, as AOA increases, the ailerons on the Dr1 become almost useless except to produce large amounts of adverse yaw.  In FSX, I can model this to a large extent, though I think the adverse yaw is scaled to the aileron lift coefficient instead of to the deflection angle which results in decreased adverse yaw along with decreased aileron authority.

 

I can also reduce effectiveness of the controls based upon dynamic pressure (airspeed)!  In other words, I can model the controls becoming "heavy" and more difficult to use at high airspeeds.

 

ROF and FSX seem to share a very high level of similarity in the FMs, so I think I will be able to produce a FM that is equivalent or (IMO) superior to what we experience in ROF.

 

I will check on the profiles and gyroscopics.  I will also share my .air file and .cfg file for anyone who has FSX and wants to try it out when I am done. 


  • 0

#39 unreasonable

unreasonable
  • Posts: 1452
  • LocationBangkok

Posted 03 April 2015 - 03:46

Very interesting, Holmes!  This is indeed fascinating.


  • 0

"Democracy is the theory that the common people know what they want, and deserve to get it good and hard." H. L. Mencken


#40 hq_Reflected

hq_Reflected
  • Posts: 4711

Posted 03 April 2015 - 05:57

I worked on it a bunch today and learned a lot about using real world data in an imperfect FM.  In editing lift coefficients and pitching moments, I found that if I used the actual data from an airfoil section (Goe 298 for the Dr1), the plane did not fly in the virtual world as it does in the real world.  In the virtual world, guess what!  It flew EXACTLY like the ROF Dr1.  So, I guess there is something to be said for the ROF guys using real data, but there is a major disconnect between this artificial virtual world and what really happens.  I think that is where people like me could really help them with their FMs.

 

 

Does this mean that 777 is using real world data, but somehow the equation is wrong, so using accurate data results in inaccurate FM-s? Because they always keep saying this is the data they have, therefore the FM-s are correct. Looks like it is not the case, then?

 

Great work by the way, I can't wait to try your Dr1, and I really hope 777 will decide to review the RoF Dr1 accordingly.


  • 0




1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users