Is anybody else excited about the new 20 KM CFD in SU11?

[mrueedi 702]

The essay section about coefficients which you posted is not discussing the coefficients for the normalization.

3 hours ago, blingthinger said:

Are you referring to this text?

For me this text is one of many sources which hints the answer. So, with what do you think these coefficients will be multiplied at runtime?

[blingthinger 671]

2 hours ago, mrueedi said:

The essay section about coefficients which you posted is not discussing the coefficients for the normalization.

So, with what do you think these coefficients will be multiplied at runtime?

Ah, I thought you wanted to address why I'm claiming Asobo is misleading everyone by saying they are solving forces on the elements when the SDK says they are actually not. No worries. We can come back to that if you want. My end curiosity is to know why the CFD (weather or FM) necessitates so much rework of the underlying element model

 

It appears that there's 2 types of coefficients:

- forces and moments (example: lift or CL)

- normalization (we can call N)

Normalization coefficients appear to be used to account for flap and spoiler deflections for the entire aircraft, that the FSX tables don't account for:

"20 configurations involving flaps and spoiler deflections have been chosen. For each of these 20 aircraft control configurations, 6 normalization coefficients are computed and stored."

These are computed at sim startup and are then multiplied to the lift/drag/moment coefficients: (example: N*CL), each time the FM is computed.

Did I misunderstand?

 

https://docs.flightsimulator.com/html/Samples_And_Tutorials/Primers/Flight_Model_Physics.htm

Edited November 11, 2022 by blingthinger

[joemiller 269]

Useless

[blingthinger 671]

1 hour ago, joemiller said:

Useless

Funny how anti-science comments keep coming up in this forum.

Almost as if you're discouraging asking questions, no?

Thus far I haven't heard from anyone who can quote the essay/SDK to prove me wrong. Mrueedi keeps talking like they can, but I have yet to see any text in these links that proves that out.

[mrueedi 702]

 

11 hours ago, blingthinger said:

Ah, I thought you wanted to address why I'm claiming Asobo is misleading everyone by saying they are solving forces on the elements when the SDK says they are actually not.

Asobo is not misleading everone, you are the only one who apparently still has not understood the normalization concept. Otherwise you would have answered my question what is multiplied with these normalization coefficients at runtime. So don't evade. Answer my question. A sentence with a couple of words answers my question.

 

11 hours ago, blingthinger said:

Did I misunderstand?

Yes, mostly.

A hint: How would you interpret this from the essay:

This interpolated set of normalization coefficients is then used to uniformly weight the lift and drag forces and the moments computed on each surface element

Now you only need to find out, what math operation is meant with the term "weight".

Edited November 12, 2022 by mrueedi

[jcomm 6,081]

I don't know if all of the new aircraft added in MFS SU 11 + 40th are CFD-enabled.

 

 

Edited November 12, 2022 by jcomm

[tweekz 2,165]

As far as flight dynamics, I am really impressed how well they've modeled the slipping. When you apply rudder and correct with aileron you really drop like a stone.

[blingthinger 671]

6 hours ago, mrueedi said:

Yes, mostly.

And you explain nothing. I get the impression you really don't have a clue here. But...we'll continue:

6 hours ago, mrueedi said:

This interpolated set of normalization coefficients is then used to uniformly weight the lift and drag forces and the moments computed on each surface element

Ahhh, you ARE reading the essay/SDK. You found it! Here's my confusion with that statement. I've highlighted your quote above in green as well. A few lines later, they say they are using coefficients. Not forces:

https://docs.flightsimulator.com/html/Samples_And_Tutorials/Primers/Flight_Model_Physics.htm

So, which one is it? Forces? Or coefficients? Another reason I think it's coefficients and NOT forces is a bit further down, we have this image:

See inside those boxes in the FM solution loop? Lots of upper case Cs in there. Those are all coefficients. Not forces.

 

 

Another question. How do you distribute moments? Again, you cite a line that says the normalization values are multiplied to the forces and moments, correct? That really breaks the brain. Can you explain how you distribute moments? Moment is explained well here:

https://web.mit.edu/4.441/1_lectures/1_lecture5/1_lecture5.html

Moment = force * distance

Distance in the sim in this case would be from the center of airframe rotation out to the model element. If the model is already distributing forces...what does it mean to distribute moments...which already have those same forces inside? Does that mean they are doing 2x the forces? Why??

 

Edited November 12, 2022 by blingthinger

[Sceadu 117]

Quote

Again, you cite a line that says the normalization values are multiplied to the forces and moments

Its clear to me that the word "Moments" in this case has been used as short term for "Moment Arm", which would simply replace the word  "distance" as shown in your illustration

[blingthinger 671]

3 minutes ago, Sceadu said:

Its clear to me that the word "Moments" in this case has been used as short term for "Moment Arm", which would simply replace the word  "distance" as shown in your illustration

Now that's an interesting perspective...so the moment arms (distance from the airframe center to each element square) are being scaled to be bigger or smaller? Per the essay/SDK, the scaling is applied uniformly across all elements.

This would mean they are effectively squishing and growing the airframe model, no? In other words, the designer inputs a wing span or a fuselage length, and it gets scaled by the normalization coefficients...which are in turn, based on flap and spoiler deflections.

So every time the pilot drops a notch of flaps, the entire airframe shrinks or grows by some factor?

[blingthinger 671]

7 hours ago, mrueedi said:

what math operation is meant with the term "weight"

And, I forgot to address your question: "weight" in this scenario means multiply. Am I wrong?

[mrueedi 702]

4 hours ago, blingthinger said:

Here's my confusion with that statement. I've highlighted your quote above in green as well. A few lines later, they say they are using coefficients.

The first describes what happens at runtime to calculate the forces per surface. The second describes what happens once at initialization to obtain the coefficients.

[blingthinger 671]

2 hours ago, mrueedi said:

The first describes what happens at runtime to calculate the forces per surface.

It contributes in the calculation, yes. The first states that the normalization coefficients are multiplied to something during the FM calculations.

"This interpolated set of normalization coefficients is then used to uniformly weight the lift and drag forces and the moments computed on each surface element"

https://docs.flightsimulator.com/html/Samples_And_Tutorials/Primers/Flight_Model_Physics.htm

"uniformly weight" = "multiply by", correct?

 

The model is initialized with force/moment coefficients. You're saying that those values then get turned into forces? So going from a number with no physical units (lift coefficient), to a number with Newtons as units (lift force)?

You're saying that it starts as:

(Lift force coefficient x normal coeff)

and then later on turns into

(Lift force x normal coeff)

??

 

 

For an example then, let's say they are forces. If we assume the aircraft is sitting on the runway w/ zero wind, there is no force. Then the aircraft starts moving forward. Air starts to put force on the element. What laws of physics are they using to calculate a force on each element?

Also, why are the moments there too, at each element? Why do they need to distribute moments if they already have forces?

Edited November 12, 2022 by blingthinger

[SAS443 1,749]

been tinkering with gliders and helicopters.

They may not be perfect (but enough to laud praise from both communities). Looking back at FSX/P3D and how it never really materialized into proper heli/soaring sim. One must admire the versatility of the physics in MSFS (when airborne. Ground is still meh). And it's only been on the market for 2,5 years. Insane rate of evolving. Unmatched even.

 

[Sceadu 117]

Quote

why are the moments there too, at each element? Why do they need to distribute moments if they already have forces?

Seriously?, you think a Force's resultant isnt dependent on the Moment?