AI landing distance advice...

[Trevytt2004]

Does anyone know if there are any specific ways to get AI to land and not come to a stop in unrealistic short distances.

I have so far tried tweaking the brake scalar in some models to as low as 0 or in some cases, tried -0.1.
This gave me aircraft landing and rolling for a much more realistic feel, but some failed to stop.......and just disappeared into the distance.

I have also changed the reverser available to 0, this seems to have some good results also.

The main reason I'm doing this is to prevent go arounds without having to use any of the ai smooth type of programs, I know modern day fighters can stop very quickly, but in FS9 or FSX, they almost stop where they touch down....then taxi along the runway at 20kts till the next taxiway exit.

Anyone have any ideas?

Regards from Tobago

[Firebird]

It really depends on the specific model that you are talking about. There are several ways that you can alter the landing roll of an aircraft and you have picked on one of them.
The problem is that most parms will affect something else. The braking scalar is one of those that shouldn't affect anything else.

One that you could also use is under GeneralEngineData, min_throttle_limit. Quite often its set to 0, or sometimes to -0.1 if you want reverse thrust. You could always set the minimum thrust to something positive, like 0.05 or 0.1, so that the engine will always deliver 5 or 10% or thrust, respectively.
HOWEVER, this could affect the way makes an approach. If the particular model has to go to 0% thrust to enable a slowing down of airspeed to enable an approach it will not slow down quick enough and possibly balloon when the flaps go down.

I would suggest that if you wanted to experiment try only 0.05 to start with to see the effect.

Any specific model that you think is wrong?

[Trevytt2004]

Steve,

Thanks for the reply.

I was sat there watching AI at Grantley Adams in Babados when an AirCanada 767 appeared on the approach, this aircraft is by AIA as far as i'm aware.
The settings i changed are below.

[gear_warning_system]
gear_warning_available = 1 //Normal
pct_throttle_limit = 0.1 //Percent throttle that activates warning
flap_limit_idle = 5.0 //Flap angle that activates warning at idle
flap_limit_power = 25.5 //Flap angle that activates warning at above idle
[brakes]
parking_brake = 1 //Parking brake available
toe_brakes_scale = -0.05 //Brake scalar
auto_brakes=4
[hydraulic_system]
normal_pressure = 3000.0 //PSI
electric_pumps = 0 //no electric pumps
engine_map = 1,1,0,0 //pumps on Eng1,2
[Views]
eyepoint=43.7, -1.31, 3.7 //(feet) longitudinal, lateral, vertical distance from reference datum

The toe brakes scale was originally much higher....cant remember exactly what, but it used to stop very efficiently.Changed it to -0.05, seems to roll and slow down nicely then reach the predetermined taxi speed then vacate the runway, only prob is.....reverser doesnt seem to work now.

Any idea what the auto brakes = 4 does?

Eventually i think i will start tweaking the c-130's and other mil aircraft to try and get them to roll for longer, rather than hit the brakes so hard within what seems like 100feet then slowly taxi at 20kts to the next available exit point.

I tried this with the project ai dash 8s, but they now seem to all push back from terminals and keep on pushing..............til disappearing. Looks like i need to revisit those cfg files again.

Trev in Tobago

P.S.
where around Heathrow are you.......I used to live just outside Cargo

[Firebird]

The toe brakes scale was originally much higher....cant remember exactly what, but it used to stop very efficiently.Changed it to -0.05, seems to roll and slow down nicely then reach the predetermined taxi speed then vacate the runway, only prob is.....reverser doesnt seem to work now.

Although you don't show it here you mentioned it earlier that you altered reverser available to 0. This would stop that appearing, I believe. Whether the min_throttle_setting set to a positive number will also have this affect I don't know as I have never tried it.
I would not have a negative number for toe_brakes_scale as 0 is no brakes. At the very least it could mean that you wiould not be able to stop yourself from hitting the aircaft in front whilst taxying. I would have it set to 0.05 at the very least to ensure some breaking.

Any idea what the auto brakes = 4 does?

Nope. I checked the FS2004 SDK and it doesn't mention it. It might be a holdover from FS2002, but I don't know whether it actually does anything or not. You could try lowering it to autobrake =2 and seeing the effect or commenting it out altogether and seeing what happens.

where around Heathrow are you.......I used to live just outside Cargo

Harlington

[mikewmac]

Trev,

The reason you see these short landing rolls terminated by hard braking is that many add on AI models use direct derivatives if not duplicates of the FS9 default AI FDE's which seem to have been created to minimize landing roll distances and allow AI to exit the runways as soon as possible in order to minimize air traffic congestion at busy airports.

With the AI FDE's that I have created for MAIW I have focused on creating as realistic as possible takeoff and landing rolls unique to the particular military aircraft model in question. This leads to significant variations in takeoff and landing roll distances between the various MAIW AI models that I created the AI FDE's for and consequently significance differences in how quickly they are able to exit the runway after landing. Some seem to like this, others would prefer uniformly short landing rolls with quick runway exits and others don't seem to even notice the differences.

Mike

[Trevytt2004]

Steve & Mike,

Thanks for your explanations.


Its trial and error most of the way for me......was mainly playing around with some civies visiting Grantley Adams that prompted me to delve further into it.

I shall continue to plug away at a few files and see what happens.

Trev

[kungfuman]

Some seem to like this, others would prefer uniformly short landing rolls with quick runway exits

I think it's a mistake for people to want unrealistically short landing rolls, as the time it takes for an aircraft to clear a runway has much more to do with how near it is to a runway exit. The more runway exits there are, then the higher the probability that the aircraft will come to a stop near one of these exits, and the quicker it will clear the runway.

But runways with many exits tend not to be so problematic for smooth AI operation. The real problem is with the runways that don't have lots of exits. Which means that the chances are it's a long way to the nearest exit. Therefore it is disingenious to program AI to have these short landing rolls, as this actually increases the time the AI aircraft will be clogging the runway at these more vulnerable runway's with few exits. The measure actually makes things worse at the very runways that need it most!

[Trevytt2004]

Valid point made Kungfuman.

Ar Grantley Adams, there is one exist right down by the terminals......and when a little island hopping dash 8 lands, it is so annoying to watch it slooooooowwwwwllllly taxi to the exit.

This is why i started experimenting......also the tonkas at ACG's Marham kind of get to me a little also.........long slow runway departure.

life is fun

[Firebird]

Yeah, in the case of Marham and Coningsby etc the problem could be overcome by allowing the military 'one on, well down' landing clearance which FS just doesn't allow.

[fishlips]

It a bugger of a thing as each airport is different, some have exits some distance down the runway while others have more exits, which by the way, does not always work well with ground movements of other AI due to a lack of taxiways to devert ground traffic.
While you may get your AI to work well at one airport, it can be nightmere at another airport that they fly too.
I played with the brake_scalar once on the PAI-C-17 and ended up with a Kangaroo bounding down the runway after it stopped, I certainly got a good laugh from that.

[kungfuman]

Longer (more realistic) landing rolls wins out for me, as it increases the probability that the landing aircraft will leave the runway further away from the threshold, which helps decrease the chances of conflicting with AI traffic enroute to the threshold for takeoff.

When tinkering with things like landing rolls, I think people forget that you can't just tailor the flight dynamics to suit a couple or so fixed airfield scenarios - unless you only intend that AI model to be used at the airfields that you've tailored it too. You need to consider a continuous spectrum of possibilities, and plan according to the more fluid world of probability.

[MIKE JG]

To get them to the end quicker you can try increasing their taxi speed, but then that looks a bit odd when they're hauling ass on the ramp areas....

[fishlips]

Generally in non-militarised zones there is no need to jump on the brakes and exit the runway quicker than normal. I go for realism in this event.

[Firebird]

The only reasons that I can see for very quick braking is operating from short runways. I can see a reason for reasonably quick stopping distance, if the aircraft is doing rollers. You may need to stop quite quickly to ensure that the aircraft lifts off again.

[fishlips]

That's very true Steve but once again adds a another demention to the issue concerning the use of the same aircraft on normal length runways. I guess a second FDE for some AI would come in handy (Normal Charictaristic and Militarised Charictaristic for short distance). The flight planss could then be made to suit each AI. Much like a dry hornet and a wet hornet.

[Firebird]

That's a fair point. I am wondering if it would unnecessarily complicate things for some packages. For example, F-18s. If you provided a wet fde, for carriers, a dry fde, for normal ops, and then a TNG fde.
A Package Manager would then either have to split up the schemes amongst the three versions and make sure that he used the right ones in the right flightplans, or double or triple up on schemes and increase the size of the package.

Would the extra complication, of this admittedly extreme example, would be worth the increased realism?

[kungfuman]

If you design the package with that outcome in mind, it wouldn't be a problem imo.

[fishlips]

I agree, the packages may need to be made more manageable if the developer feels they have bitten off more than he or she can handle without basic mistakes being made.
The larger the package the more difficult it becomes for beta testers as the time required to view each AI and the scenery if included is increased and more prone to something being overlooked.

In saying that, different FDE's in packages are meant to be left alone, if the user of the package starts to fiddle with it, then they need to be mindful.

I don't think the FDE issue is more difficult to handle than the loadouts on AI as each model is clearly labelled. Many of MAIW sceneries include carrier landing practice strips so it would add an extra demension which I think is something that we all strive to achieve in the sim as developers.
It's like the debate about welded models, when I first saw the F14 Tomcat four ships shoot across Mirramar, I was gob smacked and even today its one of favourite models to watch, as is going back to the MAIW airshow, its something different and innovative.
If we don't try and add new things, we'll simply go off like a bag of prawns in the sun and quite frankly the MAIW development team have always been right up keeping the sim interesting and fun to use.

[Firebird]

In which case, maybe we should bear it mind.

[BadPvtDan]

Maybe, but that's a heckuva lot of work for very few people.