Developing a Blade Element Theory

Discussion Hub General Discussion
62

What stuff?

3 Likes
63

I’d say that MSFS has the potential to be better, it depends how much and for how long Asobo keeps improving it.

The downside of x-planes ongoing significant changes is, that developers have to redesign e.g. the flightmodels for all their aircraft after each major version update.
And that’s happing now since almost 30 years.

2 Likes
64

For starters you’d have to actually purchase it.

6 Likes
65

Rocket ships to the moon, I guess.

1 Like
66

You won’t find “drag” as input. There is some logic in that: we’re talking simulation and propellor drag is an output, which results from a bunch of inputs. For instance, these beta_range things I tested, influence the propellor blade angle. This propellor blade angle (ref Nijntje, ref BET) results in the drag effect. How much drag.. the model decides. The inputs for small propellor aircraft are available as beta_range.. and there is some effect (output, drag) at least for DA-62. When I put min_beta_range to 20 or higher the aircraft will land more difficult because of airspeed.

1 Like
67

‘Flight model’ in this context really means two different things. There is the core ‘flight modelling’ algorithmic component built into the application, and then there are ‘flight models’ - specific modelling data - which are processed through the core software to model the behaviour of specific aircraft. And comparing MSFS with XPlane, I’d say that as it currently stands, the core MSFS component still needs some work in specific points (e.g. the prop drag issue being discussed in this thread), but that nothing I’ve seen suggests that it is fundamentally lacking capabilities that XPlane has. And given that it is under active development, it may well have the potential to surpass it, perhaps substantially, simple because Asobo have resources available that the XPlane developers don’t.

When it comes to the second part of ‘flight modelling’ - the ‘data’ part relating to specific aircraft, much of the problem with comparing MSFS with XPlane has been that people aren’t comparing like with like. The default MSFS aircraft were almost all lacking sufficient work on flight model data on initial release, and I’d have to say that a few of them are far from perfect even now. But comparing them , as I’ve seen done, with XPlane payware aircraft that in some cases cost almost as much as the entire MSFS application, clearly isn’t appropriate. The stock aircraft supplied with XPlane are, from what I can recall, similarly lacking - or in a few cases, more or less unflyable.

MSFS needs more work. XPlane is good, but not perfect. Both have the potential to be improved. MSFS is likely to have more resources allocated to it to do so. And improving it is unlikely to be achieved by trying to add ‘blade element theory’ in the naïve way that the OP suggests. That isn’t how software developments works. Software doesn’t consist of ‘theories’. It consists of working algorithms and data, and when it comes to simulations, the appropriate way to compare two different approaches is on their results, rather than on whether they claim to use some over-hyped technique with a fancy name. Anyone with practical experience with XPlane flight modelling will soon realise that the ‘blade element theory’ aspect is only a small part of building an accurate flight model, and that it is only one of the inputs to the final result. Whatever algorithmic approach you use, good flight models require hard work.

6 Likes
68

But I already have it.

70

According to the SDK, the wing dihedral is ignored. Doesn’t this incorrectly make the low-wing airplanes less stable around the roll axis?

71

Where do you see this? When searching docs.flightsimulator.com for ‘dihedral’, I can see it’s defined in the flight model config file, in the tutorial about flight models, and in the section about flight model physics, but nowhere do I see a mention of the dihedral angle being ignored


72

In the Flight Model Physics page:
" Note that the following geometrical parameters can be prescribed but are used only for debug or gauges indications:

  • Wing dihedral
  • Wing position vertical apex"
2 Likes
73

Ahhh thanks for the clarification. I’d assume this does have an impact on the fidelity of the flight model. Not sure how much of a difference that would make though


74

Someone can correct me if I’m wrong but I think that’s just saying those entries in the flight_model.cfg are ignored since that is all being handled by the physics engine. The entries are now just informational and not needed for flight modeling.

1 Like
75

Well the stabilization effect could still be there but it would require from the flight model that the wing sections produce non-vertical lift. And this is not evident from the documentation either.
Therefore this might be (one of) the reason for the balanced low-wing airplanes to bank even in the absence of any aileron input.

1 Like
76

Torque effect?

77

Test it..

Update your SDK to 0.13.0
Download the SDK-DA62 aircraft, using the Help menu of Developer mode
Leave the sim, install SDK and the SDK DA-62 aircraft (run the exe, allow Defender to do that)
Start the sim again, enable Developer mode
Go C:\MSFS SDK\Samples\DA62 and open the XML project for the DA-62
Build your project.

Go to your aircraft selection and load the SDK version of DA-62 (you have two DA-62’s now)
Use Tools/Aircraft Editor to see (or set) above parameters.. I only showed the propellor part of the Engine page, there are many more pages. If all set as you like, use menu Resync.. and fly, see what happens.

btw yesterday I did the same experiment with the TBM-930, that is set min_beta_range to 5 instead of 25. You can’t do that in the SDK, you’ll have to change the aircraft engines.cfg. I found it would take off with this setting.. but land very slow and very unsafely.. :grin: I am not going to do interpretations and speculations this time, experts feel free ! But these parameters ARE active in MSFS, it is not a fake dialog.

1 Like
78

Just an observation I have made (I have kept a complete set of the original “OneStore” folder for my reference). I have frequently examined the “engine.cfg” and the “flight_model.cfg” files, looking for potential data that might make sense in reference to real world aircraft flight and engine dynamics. Just this past few days (following SU4 update), I have noticed three completely new code lines in the engine.cfg file. Has anyone else made this connection yet?

The lines are the last three in the “[PROPELLER]” section, and they are in only the turboprop aircraft, not the GA prop aircraft. I have not yet made sense of them, but they are very curious. These are the three lines:-

prop_effmaxsmooth = 0.0;
prop_effminval = -0.9;
prop_scalepowerabs = 0.8;

They definitely refer to prop configurations in some way, and since Asobo seems to have indicated that they did some prop drag remodeling in the turboprop aircraft, there might be a connection. I have yet to try playing with the numbers to see if there is any correlation to prop drag. Maybe if enough of us give this bit of experimentation a shot, we just might find an explanation for the addition of these three lines.

Just thinking 


2 Likes
79

Try changing these while looking at the Propeller Tables inside Engine / SDK DevMode Enabled. Should become pretty clear what they are doing.

80

BET is no better than Asobo’s take on aerodynamics. Look at some of the aircraft in X-Plane; they will at times have invisible airfoils to get proper handling in th e extreme regimes.

3 Likes
81

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.