Military AI Works

I apologize in advance for asking questions about (ugh) civilian aircraft, but the issues are... well, odd!

Issue #1:

When sitting beside the runway threshold and watching my AI aircraft taxiing up and taking off, sometimes an aircraft seems to want to overtake another, they "collide", and then everything in the queue stops, nothing moves further and on to the runway to take off. This typically happens when the aircraft are turning off the taxiways and onto the runway, at one time one aircraft was on the runway another taxied into it and everything got jammed.

I guess you could say that some aircraft apparently do not maintain proper spacing when taxiing, if they did then this would not happen. Is there a parameter in the aircraft.cfg file that controls this spacing, or how does this actually happen?

I posted this some days ago on AIG, Johan replied that it might have something to do with the brake values on the culprit aircraft, but unfortunately I haven't been able to spot the culprit yet.

Issue #2:

This time I have a definite culprit, it is the DWAI DHC-7. What happens is that it taxies along, stops for aircraft ahead of it in the queue, moves forward and stops again as appropriate, looking like any other AI aircraft until it turns onto the runway and lines up, then stops. Then it just sits there, for over 10 minutes while traffic is piling up behind it. This is also what it did "quite a while ago", and I have not seen this AI aircraft since then, but this time it appeared and performed exactly like it did back then.

I also posted this on AIG, no replies yet.

Any ideas and hints are very welcome -

Jorgen

It might be a problem with the AFCAD design Jorgen. I've seen many instances of aircraft running into each other at the hold and piling up until they disappear. It is usually solved by repositioning the nodes on the entry/exit to the runway, especially the red one. You could also try listening to the radio chatter to see if that gives a clue - are all communications working as they should?

John

The thing is (in issue #2), several aircraft taxi onto the runway hold and then take off normally, and then the DHC-7 comes taxiing normally down the taxiway, keeping its place in the queue, turns 90 degrees and holds before the runway, taxies onto the runway and holds before takeoff like the others, and then nothing happens with it...

Based on the fact that only this aircraft behaves this way, where others behave normally, leads me to believe that the problem lies with this particular aircraft.

The location where this happens (both issues) is at DD's KJFK, RWY 04L, where I sit and observe this on the ILS pad. I have cleared up some traffic mess there before, because there were 2 entries to the runway there, causing confusion, but this issue remains. I can try playing around with the red hold and see what happens.

Thanks for the tip, John -

Jorgen

I now have the culprit from issue #1 - it is the SBAI Airbus 220-300. It was in line behind an UTT Airbus A350, and when the 350 turned 90 deg off the taxiway to hold short off the runway, the 220 tried to overtake it and do the 90 deg turn at the same time, leaving them both jammed together at the hold short node.

I edited the aircraft.cfg of the 220 and changed the brake values to emulate those of the FAIB Airbus A318, roughly the same weight class and the 318 behaves like it should in the line.

Off to test some more -

Jorgen

I have also gone over the KJFK AFCAD, and the (expletive deleted)s at DD had actually placed 2 (two) Hold Shorts on RWY 04L!

Maybe that's OK in real life, but I somehow don't think so in P3D... there was actually also one on RWY 22R.

Jorgen

I may have solved the issue with the A220 as decribed above, at least it now taxies along and flies off the runway without issues. I'll have to see it in a couple more situations, though, before I am totally satisfied.

But modifying the AFCAD as described has not solved the issue with the DHC-7, I'll have to tinker a bit with that one.

Jorgen

Speaking of odd behavior, my SBAI Airbus 220's like to cartwheels at their respective gates and I to do not know the problem.

md11driver44 wrote: 14 May 2026, 19:05 Speaking of odd behavior, my SBAI Airbus 220's like to cartwheels at their respective gates and I to do not know the problem.

That's a contact points in the aircraft.cfg issue usually...

Jorgen wrote: 14 May 2026, 13:21 I may have solved the issue with the A220 as decribed above, at least it now taxies along and flies off the runway without issues. I'll have to see it in a couple more situations, though, before I am totally satisfied.

But modifying the AFCAD as described has not solved the issue with the DHC-7, I'll have to tinker a bit with that one.

Jorgen

Hey Jorgen, regarding the Dash, that may be an issue with the FP... every so often it happens when I am testing some of my new aircraft... not the model as other of the same will depart, but can be related to FP timing...

jetpilot1980 wrote: 14 May 2026, 19:11

md11driver44 wrote: 14 May 2026, 19:05 Speaking of odd behavior, my SBAI Airbus 220's like to cartwheels at their respective gates and I to do not know the problem.

That's a contact points in the aircraft.cfg issue usually...

These are the contact points for both aircraft models.

[contact_points]
static_pitch = 0
static_cg_height = 6.0
tailwheel_lock = 0
gear_system_type = 1
max_number_of_points = 7

point.0 = 1, 40.187, 0, -7.5, 1000, 0, 1.2, 55.92, 1.5, 1, 0.6, 4, 4, 0, 220, 250
point.1 = 1, -6.5733, -10, -7.5, 1000, 1, 1.6, 0, 1.5, 1, 0.6, 6.9, 6.9, 2, 220, 250
point.2 = 1, -6.5734, 10, -7.5, 1000, 2, 1.6, 0, 1.5, 1, 0.6, 7.1, 7.1, 3, 220, 250
point.3 = 2, -14.17, -47.33, 4, 787.4, 0, 0, 0, 0, 0, 0, 0, 0, 5, 0, 0
point.4 = 2, -14.17, 47.33, 4, 787.4, 0, 0, 0, 0, 0, 0, 0, 0, 6, 0, 0
point.5 = 2, -57.1645, 0, 4.3495, 787.4, 0, 0, 0, 0, 0, 0, 0, 0, 9, 0, 0
point.6 = 2, 48.1729, 0, 1.9119, 787.4, 0, 0, 0, 0, 0, 0, 0, 0, 4, 0, 0

I don't know the first thing about models in FS so I will need some help.

I'll readily admit that my knowledge of flight plans and their inner workings amount to less than nothing, but I would assume that when an AI aircraft taxies, it's because that a flight plan has called for it to take off and go somewhere, so the flight plan making it just sit on the threshold makes no sense - to me, that is. But again, it is an assumption on my part...

I'm still experimenting with that one.

Jorgen

For a cartwheeling aircraft, that might be due to the CG position as suggested, or the MOI values. The calculations for the latter are beyond me, but I have found that the following set (taken from my P8 Poseidon) have worked for me in a number of AI aircraft.

;Moments of Inertia
empty_weight_pitch_MOI = 1750000
empty_weight_roll_MOI = 2262183
empty_weight_yaw_MOI = 3337024
empty_weight_coupled_MOI = 0

See if they help if you can't solve the problem with CG position.

John

So some insights I would give you as I have dealt with many contact point related headaches over my time doing this...

1. I always put the legend at the top and then put a comment line numbering everything so I can tell what I am messing with...

//0 Class
//1 Longitudinal Position (feet)
//2 Lateral Position (feet)
//3 Vertical Position (feet)
//4 Impact Damage Threshold (Feet Per Minute)
//5 Brake Map (0=None, 1=Left, 2=Right)
//6 Wheel Radius (feet)
//7 Steer Angle (degrees)
//8 Static Compression (feet) (0 if rigid)
//9 Max/Static Compression Ratio
//10 Damping Ratio (0=Undamped, 1=Critically Damped)
//11 Extension Time (seconds)
//12 Retraction Time (seconds)
//13 Sound Type

//0, 1, 2, 3 4, 5, 6, 7, 8, 9, 10, 11, 12,
point.0 = 1, 40.187, 0, - 7.5, 1000, 0, 1.2, 55.92, 1.5, 1, 0.6, 4, 4, 0, 220, 250
point.1 = 1, -6.5733, -10, -7.5, 1000, 1, 1.6, 0, 1.5, 1, 0.6, 6.9, 6.9, 2, 220, 250
point.2 = 1, -6.5734, 10, -7.5, 1000, 2, 1.6, 0, 1.5, 1, 0.6, 7.1, 7.1, 3, 220, 250
point.3 = 2, -14.17, -47.33, 4, 787.4, 0, 0, 0, 0, 0, 0, 0, 0, 5, 0, 0
point.4 = 2, -14.17, 47.33, 4, 787.4, 0, 0, 0, 0, 0, 0, 0, 0, 6, 0, 0
point.5 = 2, -57.1645, 0, 4.3495, 787.4, 0, 0, 0, 0, 0, 0, 0, 0, 9, 0, 0
point.6 = 2, 48.1729, 0, 1.9119, 787.4, 0, 0, 0, 0, 0, 0, 0, 0, 4, 0, 0

2. I always ensure that the static_cg_height abs value is 1< than the vertical positions of the contact points abs value so in this case static_cg_height=6.5

3. Generally speaking, it helps if the compression and damping values on the nose are less than those values on the mains so for point zero values 9 and 10 you may want to try setting to .5 and .3 respectively like this:


//0, 1, 2, 3 4, 5, 6, 7, 8, 9, 10, 11, 12,
point.0 = 1, 40.187, 0, - 7.5, 1000, 0, 1.2, 55.92, 1.5, .5, 0.3, 4, 4, 0, 220, 250


4. A couple of other things to look at in sim settings:

https://prepar3d.com/forum/viewtopic.php?t=112692

a. Are your frame rates and or sim rate set too high? For example if I set sim rate to 4x (do this sometimes when testing AI aircraft) sometimes may AI aircraft will start to bounce especially if my contact points are off...

b. Look at tessellation and or terrain detail settings if those are too high it can cause AI aircraft bouncing...

c. look at any addons / addon utilities such as sloped runway utilities and see if those are conflicting...


Hope all that helps...

The more and more I go blind reading the aircraft.cfg for the DWAI DHC-7, the more and more I tend to agree with Mason that the reason it does not commence its takeoff run might be flight plan related...

Oh well, time to read something different then!

Jorgen