Aircraft Suspension Physics Adjustment Question

[Bt103504 30]

I know Seb mentioned in the most recent developer stream that they are working on a much more comprehensive ground physics and airframe modeling for aircraft to help them feel a bit more weighty. But one thing that has bothered me since the sim has launched has been the almost RC plane suspension physics when it comes to landing, takeoff, or any sort of bush trip. I'm seeking a little bit of assistance on identifying where the suspension parameters live within the flight model cfg? 

I know I'm going to get flamed for this, but I bought and tested this product; https://secure.simmarket.com/illuminators-improved-physics-special-edition-msfs.phtml Yes, I know the dubious nature of the "publisher", which is why this is the only product I've ever purchased from them. I do have to admit, despite the physics being a little overdone, it does provide a lot more weight to the aircraft. 

I've searched high and low for how they seemed to accomplish the ground physics. All I've been able to find is criticism from folks saying "how dare they publish a product that is just flight model cfg edits that everyone already knows how to do" but then never elaborating on the how end-users might be able to make these changes themselves. I did use notepad and did the compare plugin to compare to the current flight model cfg's and needless to say there have been so many changes to MSFS since the product was released the effort seems to be futile in terms of identify the individual changes that were made. 

Would any kind simmer with knowledge of the flight model cfg files care to provide some insight? Are these adjustments to the contact points section? Or should I be looking elsewhere in the flight model cfg?

Thanks all!    

Edited April 5, 2022 by Bt103504

[ErichB 3,638]

Excellent point. Hope you gain traction with this.

[Mets737 193]

That is one major thing about this sim that is driving me nuts, every airplane I fly has stiff as rocks suspension, A couple examples are the L39 and F18's trailing link landing gear, they DONT move and the sim doesn't think they have any weight on them. I really hope this gets fixed in the future....

I don't like to be that guy that compares different sims but from what I seen on X-Plane (I dont use X-Plane) the gear suspension on that platform looks fantastic. 

-Mets747

[robert young 2,897]

2 hours ago, Bt103504 said:

I know Seb mentioned in the most recent developer stream that they are working on a much more comprehensive ground physics and airframe modeling for aircraft to help them feel a bit more weighty. But one thing that has bothered me since the sim has launched has been the almost RC plane suspension physics when it comes to landing, takeoff, or any sort of bush trip. I'm seeking a little bit of assistance on identifying where the suspension parameters live within the flight model cfg? 

I know I'm going to get flamed for this, but I bought and tested this product; https://secure.simmarket.com/illuminators-improved-physics-special-edition-msfs.phtml Yes, I know the dubious nature of the "publisher", which is why this is the only product I've ever purchased from them. I do have to admit, despite the physics being a little overdone, it does provide a lot more weight to the aircraft. 

I've searched high and low for how they seemed to accomplish the ground physics. All I've been able to find is criticism from folks saying "how dare they publish a product that is just flight model cfg edits that everyone already knows how to do" but then never elaborating on the how end-users might be able to make these changes themselves. I did use notepad and did the compare plugin to compare to the current flight model cfg's and needless to say there have been so many changes to MSFS since the product was released the effort seems to be futile in terms of identify the individual changes that were made. 

Would any kind simmer with knowledge of the flight model cfg files care to provide some insight? Are these adjustments to the contact points section? Or should I be looking elsewhere in the flight model cfg?

Thanks all!    

The contact points in MSFS are largely a copy and paste of FSX and P3d points. On a three point/wheeled average aircraft there is a nosewheel (point.0) and two other, unsurprisingly, below wing main gear points (point.1 and point.2) The contact point section has these three plus other contact points which are for want of a better word "scrape" points along the fuselage, ie: tips of wings, top of the vertical fin and ends of the horizontal stabiliser, plus the nose and tail scrape points.

The suspension is arranged in an array of parameters going left to right for each point, that defines the position of the contact points (wheels/tires for example) in relation to the model center in length (z=+/-), width (x+/-) and height (y+/-). Following that are stress/impact values that define when a "crash" would be detected if crash detect is switched on.

Next is the radius of each wheel of the three (a collection of multi wheels or main undercarriage is treated usually as one unit), so two for each side and one on the nose).

Next we have one single value which is the notional potential steering angle of the nosewheel, (or sometimes the tailwheel in tail draggers). Usually only the nose or tail is referenced, so the values of the right and left main gear is valued zero.

Next we have the "stiffness" of the suspension, then the extent to which the suspension travels and lastly what I would describe as the inertia by which the suspension moves (the official definition is different).

The last set of params on the right concern the time taken for each unit of wheels to retract and deploy.

ABOVE these are general params which are broadly new to MSFS compared with FSX/P3d. These define how quickly or slowly the nose or tail wheel begin to truncate their ability to steer, either by differential braking or by a link to the rudder, in feet per second, as the aircraft accelerates down the runway. It is these specific points which must be tuned with infinite care, as they are extremely sensitive to changes and largely dictate how twitchy the ground steering is at various speeds, although their response is also very linked to the inertia values set in the weight and balance section.

I've just given you a small snapshot of how it works. I can happily tell you that the above is just the start. I have been tweaking these values for well over 25 years and probably know as much as anyone about them, but a word of caution. They take a very long time to adjust, partly because you have to keep reloading the aircraft many times to see the effectiveness of even very slight changes. If you've got several days/weeks or even in extreme cases months (!) you might be able to improve what you wish. The contact points represent a small fraction of the rest of the flight_mode.cfg file. I hope you find this useful 🙂

Edited April 5, 2022 by robert young

[Glenn Fitzpatrick 3,798]

Interesting that pod strikes are not modelled.

[somiller 694]

Guess I won't be trying to fix this myself!

[Baber20 1,438]

This has been my biggest issue with the SIM. Suspension modeling is mediocre right now. Every plane feels like an RC plane. You can especially feel this on landings when the aircraft seems to be completely missing suspension modeling.

[Bt103504 30]

6 hours ago, robert young said:

The contact points in MSFS are largely a copy and paste of FSX and P3d points. On a three point/wheeled average aircraft there is a nosewheel (point.0) and two other, unsurprisingly, below wing main gear points (point.1 and point.2) The contact point section has these three plus other contact points which are for want of a better word "scrape" points along the fuselage, ie: tips of wings, top of the vertical fin and ends of the horizontal stabiliser, plus the nose and tail scrape points.

The suspension is arranged in an array of parameters going left to right for each point, that defines the position of the contact points (wheels/tires for example) in relation to the model center in length (z=+/-), width (x+/-) and height (y+/-). Following that are stress/impact values that define when a "crash" would be detected if crash detect is switched on.

Next is the radius of each wheel of the three (a collection of multi wheels or main undercarriage is treated usually as one unit), so two for each side and one on the nose).

Next we have one single value which is the notional potential steering angle of the nosewheel, (or sometimes the tailwheel in tail draggers). Usually only the nose or tail is referenced, so the values of the right and left main gear is valued zero.

Next we have the "stiffness" of the suspension, then the extent to which the suspension travels and lastly what I would describe as the inertia by which the suspension moves (the official definition is different).

The last set of params on the right concern the time taken for each unit of wheels to retract and deploy.

ABOVE these are general params which are broadly new to MSFS compared with FSX/P3d. These define how quickly or slowly the nose or tail wheel begin to truncate their ability to steer, either by differential braking or by a link to the rudder, in feet per second, as the aircraft accelerates down the runway. It is these specific points which must be tuned with infinite care, as they are extremely sensitive to changes and largely dictate how twitchy the ground steering is at various speeds, although their response is also very linked to the inertia values set in the weight and balance section.

I've just given you a small snapshot of how it works. I can happily tell you that the above is just the start. I have been tweaking these values for well over 25 years and probably know as much as anyone about them, but a word of caution. They take a very long time to adjust, partly because you have to keep reloading the aircraft many times to see the effectiveness of even very slight changes. If you've got several days/weeks or even in extreme cases months (!) you might be able to improve what you wish. The contact points represent a small fraction of the rest of the flight_mode.cfg file. I hope you find this useful 🙂

Robert thank you so much for this extremely thoughtful and enlightening response. This is really helpful. Just to see if I am understanding the formula you are describing can u see if this formula is correct. Let's use one of the main landing gears with the following config line;

point.1 = 1, -15.8, -6, -4.14, 2000, 1, 0.5, 0, 0.33, 2.8.0, 0.05, 0, 0, 2, 0, 0

If my understanding of what you discussed is correct then each of the entries (seperated by a comma) for this specific contact point would be described as below;

point.1 = (type of contact point 0 = nosewheel OR contact point 1 or 2), (length relation to model center), (width relation to model center), (height relation to model center), (stress/impact crash detection value), (radius of wheel), (potential steering angle of nosewheel/tailwheel),  (suspension stiffness), (extent of suspension traversal), (suspension inertia), (remainder of values are retract and deploy values of contact point)

So in the example point above .33, 2.8.0, and 0.05 are the specific suspension settings that would need adjusted? Do I have this correct? Is there a way of doing this real-time in the sim? Perhaps in Dev Mode?

[robert young 2,897]

1 hour ago, Bt103504 said:

Robert thank you so much for this extremely thoughtful and enlightening response. This is really helpful. Just to see if I am understanding the formula you are describing can u see if this formula is correct. Let's use one of the main landing gears with the following config line;

point.1 = 1, -15.8, -6, -4.14, 2000, 1, 0.5, 0, 0.33, 2.8.0, 0.05, 0, 0, 2, 0, 0

If my understanding of what you discussed is correct then each of the entries (seperated by a comma) for this specific contact point would be described as below;

point.1 = (type of contact point 0 = nosewheel OR contact point 1 or 2), (length relation to model center), (width relation to model center), (height relation to model center), (stress/impact crash detection value), (radius of wheel), (potential steering angle of nosewheel/tailwheel),  (suspension stiffness), (extent of suspension traversal), (suspension inertia), (remainder of values are retract and deploy values of contact point)

So in the example point above .33, 2.8.0, and 0.05 are the specific suspension settings that would need adjusted? Do I have this correct? Is there a way of doing this real-time in the sim? Perhaps in Dev Mode?

Hi BT,

Almost right. Best thing is to give you an example from a default aircraft:

point.0 = 1, -2.7, 0, -3.45,  750,   0, 0.523, 15, 0.245, 1.33, 0.5, 0, 0, 0, 165, 165, 4
point.1 = 1, -7.9, -4, -3.30, 1500, 1, 0.523,  0, 0.162,    2,    0.5, 0, 0, 2, 165, 165, 4
point.2 = 1, -7.9, 4, -3.30, 1500,  2, 0.523,  0, 0.162,    2,    0.5, 0, 0, 3, 165, 165, 4

Going left to right on point.0 

=1 (is a wheel, not a scrape point - can be any value between 0 and 5 thus:

0 = None, 1 = Wheel, 2 = Scrape, 3 = Skid, 4 = Float, 5 = Water Rudder )

-2.7 (is -2.7 feet from the notional model centre)

0 (is dead centre laterally

-3.45 (is feet below the model centre, ie: the lowest part of the wheel/tire)

750 (Is the value of collision/impact strength at which a "crash" is detected by the sim if crash detect is on)

0 (wheel brakes 0= no brake, 1 = left brake 2= right brake)

0.523 (radius of wheel/tire)

15 (max steering angle of nosewheel - note: optional value of 180 means it doesn't "steer" but is free castoring -useful for tail draggers)

0.245 (amount of stiffness/resistance/rcompression of wheel when contacting the tarmac - low values stop tire from descending below surface when braking)

1.33 (amount of travel in suspension or ratio of static compression)

0.5 (inertia of suspension travel or put another way the amount of damping in suspension movement/compression)

0 (extension time in seconds)

0 (retraction time in seconds)

0 (type sound triggered on contact with ground thus:

0 = Center Gear,

  1 = Auxiliary Gear,

  2 = Left Gear,

  3 = Right Gear,

  4 = Fuselage Scrape,

  5 = Left Wing Scrape,

  6 = Right Wing Scrape,

  7 = Aux1 Scrape,

  8 = Aux2 Scrape,

  9 = Tail Scrape

165, 165 (airspeed in knots that respectively stops gear from retracting due to damage and airspeed at which gear is damage)

Not sure about "4" - could be something introduced by Asobo or perhaps I can't remember.

This should answer your pm also!

 

Edited April 5, 2022 by robert young

[robert young 2,897]

I should add:

There is probably a way to see the contact points by using the one of the pages in the SDK, running them while in the sim. There should be a wireframe schematic of these values/points.

Traditionally there was a hack to see the points in relation to the ground by assigning lights to the same positions of each contact point so you could see exactly where the contacts are made with the tarmac.

Also, answering your dev mode query: You can reload the aircraft after a flight_model edit by doing this:

Press escape with the aircraft loaded and sim running. Edit your flight model cfg file using notepad or similar. Save the file after each edit. Then resume the sim and do the following:

With dev mode on, find in the SDK menu and at the top left - click on "create new project", then name it by filling in the fields. If you did it correctly you should see a black strip with the name of your project (you can recall this project at any time even after you've shut down the sim).

If the project is valid you can now open the "aircraft editor" at the top SDK menu strip (top of the sim window also on the left). Click on "aircraft editor" and another black strip will appear. Now click "RESYNC" and the aircraft will reload with your edits. Sometimes the sim crashes, other times it reloads successfully. DO NOT CLOSE the black strip menus as that WILL crash the sim. Just press the white arrow so they are dormant but wait for the aircraft to reload successfully otherwise the sim might crash!

Edited April 5, 2022 by robert young

[overspeed3 410]

As stated above:

"The contact points in MSFS are largely a copy and paste of FSX and P3d points"

If this actually is, and seems to be the case, realistic landings have long been a source of disappointment in many of the aircraft over the more than 35 years of flight simming .  Was not the Asobo team tasked to create new flight dynamics in their version of MSFS simulated aircraft?  So can we expect any real innovation in this area any time soon? 

[jcomm 5,896]

I am saving this thread in my favourites because not only of the questions it brings but also because of the excellent contributions by Robert Young ( as usual... ).

ASOBO / Sebastien has mentioned in a few Q&A sessions that they are well aware of the limitations in ground physics modelling and willing to update it in the future, so, I believe we can have some hope about that.

 

Edited April 6, 2022 by jcomm

[cavaricooper 2,613]

Jose- RY is one of those “few” precious members of our Community that are gifted with brilliance, determination and a touch of altruism.

I use “few” with the same reverence as Winston. We are indeed fortunate and indebted.

C

[robert young 2,897]

7 hours ago, overspeed3 said:

As stated above:

"The contact points in MSFS are largely a copy and paste of FSX and P3d points"

If this actually is, and seems to be the case, realistic landings have long been a source of disappointment in many of the aircraft over the more than 35 years of flight simming .  Was not the Asobo team tasked to create new flight dynamics in their version of MSFS simulated aircraft?  So can we expect any real innovation in this area any time soon? 

Actually there isn't too much wrong with the legacy contact points. But they do need a lot of tweaking to be convincing. It is the underlying friction modelling that needs an overhaul. One of the problems is that when you steer the nosewheel through the rudder most pedals/controls have very little resistance, so you can command an almost instant steering input. In a real aircraft you could not do this - the nosewheel takes a certain time to move left or right. More sophisticated real aircraft have a clever system whereby as you accelerate the wheel steering angle drastically reduces, lessening the chance of squirreling down the runway.

In X plane there is also an excessive reaction to steering inputs which is why many of the smaller (and sometimes larger) aircraft are permanently skidding left and right.

The other issue in MSFS is that crosswinds have a grossly exaggerated affect on weather vaning into wind, with ridiculously large amounts of lateral instability on the ground when the slightest, tiniest crosswind is encountered. That is not a flight model issue or even a friction issue but a weather/environmental algorithm problem which is well overdue for addressing by Asobo.

[Bt103504 30]

Needless to say I'm thrilled that there has been such a positive response and great discussion regarding this topic. Hopefully someone at Asobo might bring these issue to a roundtable at some point in the future. I agree with @robert young that the friction model is simply atrocious. Luckily to some degree we had things like the latter versions of EZDOK in P3D that allowed you to tweak some of the ground friction instantly. Also in terms of with the Advent of realturb + EZDOK in P3D you also were able to get much better turbulence in sim as well.

I know the criticism for this sim comes from a place of deep love and passion for this franchise (at least for most of us), but its important to keep in mind that this iteration of MSFS is still in its infancy. Personally I've never encountered a Dev that has been so transparent with the community about discussing internal projects and improvements that are being made. There is MUCH work to be done even Asobo has been clear about this, the good news is that they are actively working on it and are committed to long-term development.

Back to the original topic. I'll do some testing today with one of the default aircraft tweaking. From what I gather Robert is that I should focus tweaks to following areas of a contact point if I'm wanting to fiddle with suspension (per the example you provided);

15 (max steering angle of nosewheel - note: optional value of 180 means it doesn't "steer" but is free castoring -useful for tail draggers) = entry #9

0.245 (amount of stiffness/resistance/rcompression of wheel when contacting the tarmac - low values stop tire from descending below surface when braking) = entry #10

1.33 (amount of travel in suspension or ratio of static compression) = entry #11

 

Edited April 6, 2022 by Bt103504