|
Post by George Carty - HJG on Dec 19, 2013 6:53:46 GMT
I tried to use the autopilot for an ILS approach with a 90o crosswind of about 24 knots. The autopilot gave me a reasonable crab angle, but it headed not for the runway, but about 100 metres downwind of it -- is this something which needs fixing by modifying the aircraft.cfg or AIR file?
Hope you can help here...
George
|
|
|
Post by Tony Madge - HJG on Dec 19, 2013 11:46:41 GMT
Interesting George, would the high X-wind have anything to do with limitations of the aircraft, I would have thought 24 kts was tight but workable on the L1011, however it depends on the FDE set up?? I look forward to hearing the result of this one. In fact I may when I get chance recreate this one
|
|
|
Post by Mike Monce - HJG on Dec 19, 2013 13:51:09 GMT
George,
Which L1011? As I recall they probably share a common air file, but the cfgs are different. I'd like to try a test today.
Mike
|
|
|
Post by George Carty - HJG on Dec 19, 2013 13:58:23 GMT
I was using the L-1011-100 model (it may have been a beta model, btw).
George
|
|
|
Post by Mike Monce - HJG on Dec 19, 2013 18:00:40 GMT
Yep... I did a test and the plane behaved exactly as you said. Here's a short term fix: replace the following lines in the autopilot section with these: nav_proportional_control= 12.00 nav_integrator_control= 0.25 nav_derivative_control= 0.01 nav_integrator_boundary= 2.50 nav_derivative_boundary= 0.75 With these adjustments I was able to get very close and took over manually at about 400 AGL with a good landing. Mike
|
|
|
Post by aerofoto - HJG Admin on Dec 19, 2013 18:32:44 GMT
During pre-release testing (2009) .... and since also .... I've not encountered this particular problem with the L1011 during "AP controlled ILS coupled auto-approaches to landing" and even during an acute 90* (degree) intercept from around 18 DME .... BUT .... this was also during, what might best be described as, "normal/default FS meteorological conditions" only. L1011 AP controlled ILS coupled auto-approach to landing perforfamce was't tested during such extremeparamaters as George descibes. PLEASE NOTE .... I'm not implying there's no such problem to be reconciled .... but rather .... I'm only saying that a 90* (degree) and 24 KT crosswind is approaching "a fairly extreme" component considering most large civil aircraft are probably limited to around/have maximum crosswind landing restriction of around 30 KTS or so It (what's being described here in terms of the L1011 siulations performance with these sorts of parameters applied) could, possibly, be an FS thing .... since I've often "felt" that FS does tend over-exagerate the effect of crosswind component. In the event that you feel you can improve it/performamnce George (assuming there's something there that does need to be addressed) .... then please .... feel free do do so .... and either email, or PM, the new data to me for testing .... and I'll gladly look at it when I can. The only issue I've, ever experienced, personally, using any of the current L1011 simulation/s .... and in partucular during AP controlled ILS coupled auto-approaches to landing .... is a tendancy to (sometimes, but, not all the time) loose ILS/GS allignment during "the last several hundred feet" (only) of an approach to landing .... thus triggering the "GIDESLOPE" callout .... BUT .... I soon diagosed this observation to possibly be the result of landing at too high a weight and with possibly a liitle more than desirable airspeed too. What happens George/Mike if you try this type of approach to landing with a full payload .... but .... with total fuel quantity reduced to no more than around 20% ?
A number of factors can often influence the integrety/fidelity of such landings and this "MAY" .... possibly .... make a difference.Once again .... I'll be glad to have a look at this with you George .... and with also Mike .... as I can and between other impositions/commitments at the moment. Mark C AKL/NZ
|
|
|
Post by aerofoto - HJG Admin on Dec 19, 2013 18:36:03 GMT
OH .... sorry Mike .... I see (only after my posting) you replied whilst I was still trying to Mark C AKL/NZ
|
|
|
Post by aerofoto - HJG Admin on Dec 21, 2013 1:08:57 GMT
OK folks .... I had a bit of time, so, I've had a closer look at this report .... and .... here's my own analysis of this situation based on what I've observed, today, flying the HJG/VL L1011 TRISTAR on 2 seperate AP controlled ILS coupled auto-approaches to landing using different weather scenarios and using the current FDE and panel options for these simulations .... For my 1st flight .... I performed a standard/typical AP controlled ILS coupled auto-approach to landing on RWY 34R at the default FS9 KSEA airport .... using the "default FS9 meteorology" (no wind or other weather parameters applied .... as is loaded with any "default FS9 flight scenario"). The simulation was flown with a full payload .... but .... its fuel load was reduced to around 20% in each of the 3 tanks in order to approximate aircraft weight by the conclusion of a typical flight and ensure I wasn't overweight for landing. The simulation flew this entire approach to landing "PERFECTLY/B-E-A-U-T-I-F-U-L-L-Y" .... all the way down .... save for what I mentioned yesterday in regard to it slipping slightly below the GS, triggering the automatic "GLIDESLOPE" callout, within the last several hundred feet of the approach prior to landing .... BUT OTHERWISE .... the simulation never strayed from the ILS beam/RWY centerline, at all, and a near-perfect landing resulted. This satisfies me sufficiently to say the current AP data (within each L1011 CFG file) is fine/OK In respect of the simulation slipping slightly below the GS during the last several hundred feet of the approach prior to landing .... I also hasten to add .... I generally disconnect the AT at around 1000 FT, and then disconnect the AP also at around 500 FT, in order to be able to hand fly this final stage of any approach to landing fully manually (my own personal procedure/preference), so, these actions .... combined with the fact that I've not flown any of the HJG/VL L1011 simulations for "quite a while" (I'm probably a little "out-of-sync" with these particular simulations by now) might account for this particular discrepency. For my 2nd flight .... I performed the very same AP controlled ILS coupled auto-approach to landing on RWY 34R at the default FS9 KSEA airport .... BUT THIS TIME .... I edited the FS weather parameters as follows .... WIND = 24KTS
WIND DIRECTION = 248* degrees (represent a roughly 90* degree crosswind) My payload and fuel loadings remained the same "as previous" though (full payload and a fuel loading reduced to 20%). The simulation intercepted the ILS just as perfectly .... but this time .... the applied weather parameters did influence it, slightly, so as to result in the simulation flying this particular approach to landing with a slightly "crabbed" attitude (nose very slightly to the left) .... precisely as reported by both George and Mike .... and which is about what I'd expect to see anyway during any such approach to landing during these sorts of conditions .... SO .... "no surprises there at all". During most of this approach to landing the simulation still flew the ILS beam perfectly .... although it had, by the last 1-2 DME, definitely drifted slightly to the right of the RWY centerline. I once again disconnected the AT at 1000 FT .... BUT .... when I disconnected the AP, at around 700 FT "a slight" roll to the left" was experienced .... BUT .... was quickly corrected using bit of aileron and rudder influence in order to tidy-up the approach and which then resulted in an otherwise perfectly normal landing. Another detail I obseeved at the conclusion of this 2nd flight (and this's also what prompted me to comment, yesterday, that FS "does", IMHO, tend to over-exaggerate the effects of crosswind influence) is .... with a 90* degree and 24 KT crosswind applied, the simulation couldn't/wouldn't rollout after landing or taxi in a straight line. In fact it wanted to constantly, but very slowly, drift across the RWY as a result of the applied crosswind component. MY oWN CONCLUSION .... I don't think there's any problem/s here. I think what's being reported is simply a manifestation of how FS weather (crosswind component in this particular case) influences not only the HJG/VL L1011 TRISTAR simulations, but also, other simulations too .... because I've made very similar observations in the past with these sorts weather conditions applied and using other simulations. Mike .... I've not had time today to fly a 3rd test using your recomended AP data edit .... and probably won't be able to do so until sometime after Christmas, but, I "will" do so, when I've time, and then report back to you later on .... of course Mark C AKL/NZ
|
|
|
Post by walterleo on Dec 21, 2013 10:33:38 GMT
|
|