Prepar3D Forums

Ph_737NG wrote: ↑Mon Feb 04, 2019 4:57 am Hello,

Are there currently any plans ( or is it even possible) to improve on the current ground friction physics? After some googling I found an adjustment made through FSUIPC (I think) but that looks like it doesn't work with the current version of P3D.

Ground friction is quite high and unrealistic. Single engine taxiing is also practically useless as you will end up using more fuel than saving.

Does anyone know of a way to tweak the ground friction in P3D 4.4?

Ph_737NG wrote: ↑Mon Feb 04, 2019 4:55 pm I probably should have been a little more specific. I’m not having trouble taxiing on two engines. I’m speaking of how realistic the ground friction is. For example in the real world I have never flown a jet that wouldn’t taxi at idle thrust. You obviously need some break away thrust in many cases but, In P3D at idle thrust you will come to a stop. On top of that the real aircraft I have flown ALL taxi on a single engine quite well. In P3D you sometimes need 70%+ N1 just to get moving and after that require much higher than idle thrust to keep moving. Now there are some areas in the real world that could be slightly realistic (e.g. taxiway C in LAX westbound, is slightly uphill and takes a little bit more to get moving on a single engine) but slopes are not simulated in P3D. Additionally while taxiing single engine it requires quite a lot of rudder to stay moving straight. All of which are not issues in real aircraft.

downscc wrote: ↑Tue Feb 05, 2019 4:32 pm The "ground surface friction" modeled in P3D and ESP/FSX/FS9 before it is a very simple example of how to build an arcade game. The problem is not complex for a mechanical engineer and it has little to do with friction unless you are in a skid at which point you broke contact are now sliding and now friction comes into play. What is required is a model that incorporates "rolling resistance," which varies wildly based on tires size, number, inflation, wall stiffness and weight. This kind of simulation is probably beyond the scope of what LM is trying to achieve here and I wouldn't expect it. Ever.

PMDG in their QOTSII B747 product has worked outside of the box and used their own modeling for rolling resistance. The difference is hard to compare to other aircraft simply because of it's size and weight but it's good to see product developers extending the craft.

Ph_737NG wrote: ↑Thu Feb 14, 2019 3:54 am Any idea what specific changes PMDG made to change this on their 747? I'd like to be able to make my own adjustments if need be to make it work at least somewhat more realistically.

funkyaz77 wrote: ↑Wed Apr 24, 2019 8:27 pm @adambreed
@robmcarthy

Please please consider this, even if you alter the PDK to allow developers to better model ground handling!

Blessings

Duncan Odgers

kand wrote: The original default values contained in SIM1.DLL are the issue, I have done a lot of research on friction coefficients and posted my findings on fsdev. The bottom line here is that using the default rolling friction sim value for concrete causes other values within the airfile to become skewed. All of the airfiles available today are made to overcome this incorrect value with side effects.

The high default rolling friction value causes the turbine tables to be skewed to overcome high friction, one must set high Fn (net thrust) in 1506 for the idle/static mach line. This higher unrealistic Fn value causes fuel burn to be higher at idle because FB is proportional to thrust. Some developers then overcome this high FB part by faking the fuel burn gauge to display the correct FB bypassing the sim variables.

I have experimented a lot here and set up an airfiles turbine tables to exactly match the real GE90-94b. I know the thrust values (at ISA and ZFWs) to get the aircraft rolling and also have a thrust to CN1 table. I concluded that the sim1.DLL values are approximately treble the real world values for concrete (0.014).

Its not just the rolling friction coefficients, the sliding is wrong also which is why aircraft skid in turns whilst taxiing.

The problem for LM here is that if they fix sim1.dll, they break every other legacy airfile. Its not a big job to fix the airfile, but I suspect commercial developers would rebel against this !

funkyaz77 wrote: ↑Thu Apr 25, 2019 3:17 pm
SIMBOl - I realise that custom simulation is available but it still does not seem to be a solution for realistic high fidelity flight modelling. Aircraft still fly "on rails" PMDG and FSL do a better job but even those models are not where they could be if LM improved the modelling.

Blessings

Duncan Odgers