AI Flightplanning

[hobby]

Many years ago I produced quite a number of successful AI flightplans using the excellent FS9 program Ttools.

Having very belatedly moved over to FSX I am now trying to use AIFP v3.

The plan I am using to familiarise myself with AIFP involves a Canberra starting the day with T&Gs at base and when those are completed fifty minutes later flying to another airfield to approach , but not land away, and then return to base.
In plain language the Canberra moves off at 07:30 Hrs GMT and T&G at base until 08:20 hrs. I then programmed the acft to leave base at 08:24 hrs and fly to RNAS Yeovilton, make an approach and depart Yeovilton at 09:16 Hrs GMT and return directly to base.

To test by observation I remained at base and the Canberra moved off at 07:31 hrs taxied to runway and t/o at 07:39 hrs; three T&Gs were completed by 08:21hrs but the expected fourth T&G did not take place. Am I correct in assuming that after the third T&G the Canberra then flew to Yeovilton without contacting base tower?

I nipped down to RNAS Yeovilton altered the clock in my acft to two minutes before the Canberra's ETA at base and warp speeded back to base. Sadly the Canberra appeared to have arrived earlier than expected so I missed the return to base.

Constructive advice would be welcomed - perhaps it would have been a better test of my flightplan if I had just sat at base at the mid point of the duty runway and seen the whole plan through from 07:30 hrs GMT to 09:59 hrs plus.

[Firebird]

The problem is most probably the time difference between the two legs.

I think the actual limit between legs is 5 mins but years of experience have shown that it you have more than 3 minutes between a TNG leg and the next leg that the aircraft could go ballistic and disappear eventually.

Always program at least 2 mins but preferably 3, to allow for compilation inaccuracies, so set to leg for Yeovilton to 08:23. It could be that simple.
Experience has also shown that if you program two, or more, TNG legs consecutively that you can get weird circumstances that you didn't expect. So be prepared for that.

On the TNG timings the TNG completed at 08:20 so once the approach finished at 08:21 it turns to the next leg but the next leg start is too far from the end of the TNG leg end time and you get what you saw.

Now in anticipation of your question, if the TNG leg actually finished at 08:21 and the next leg start time is 08:24 that is only 3 mins so why did the problem happen?
Simply put the actual completion time is irrelevant only the programmed completion and start times with regards to TNG legs.

So to fix your issue I would amend the 08:24 time to 08:23.

Finally, no this is not an AIFP issue it happens with Ttools as well. The difference is that Ttools was not that accurate when compiling times whereas AIFP is far more accurate. You may find that if you recompiled some of your old Ttools plans, from the txt files, with AIFP you would get far more of the same.

Now what I do to pickup potential issues to correct is I set two parms in AIFP.
Minimum Time Arr->Dep (min.) to 3
ETD Preset/Standard sit-Time (min.) to 20

This means that completed flights must have at least 20 mins before the next fleg and the second one means that you must have 3 mins between legs. That will highlight any TNG legs that either have too much or too little time before the next leg.
If the times fall outside the limits you will get an error message for each leg on compilation.
You would be amazed at how many of the old MAIW flight plans fall foul of this as well.

Hope this helps.

[hobby]

Thanks for your help Firebird. I have been experimenting and I found that it is not always possible to add AI traffic to third party scenery which already has AI traffic - this could be me making an error while trying to learn how to operate AIFPv3.


I think that I have successfully produced a flightplan in which a Canberra takes off from home base, does one IFR TNG and then is sent off to another airfield which it departs from one hour after landing to return to its original airfield. By observation I have checked and noted the time taken to complete the IFR TNG immediately after (3 minutes) the TNG I programmed the AI Canberra to fly to the second airfield. I have observed that the Canberra does leave the second airfield at the programmed time of departure to fly home.

I surmise that it is not necessary to make the Canberra do a second circuit and land at base prior to setting off to the second airfield 3 minutes after landing. You did write that two consecutive approaches to the same airfield might produce unexpected results.

[Firebird]

We might be a bit out on terminology here, what I thought that I was trying to get over was that two TNG legs had issues not that you got two TNGs at the same place.
If you think about it you can set the time to do circuits for an hour and it will do an hours worth, i.e multiple TNGs, with no problems. The issue is when it moves to the next flightplan leg. If you have two TNG legs you can get unpredictable results.

I shall try really hard to explain.
As you know a flightplan must have at least two legs. In simple terms an outbound and an inbound leg. That is if you stick to the convention that you should always end up where you started from in a flightplan.

What you can do is a direct leg from A to B and then a direct leg from B to A with no issue at all - providing that you have a suitable full stop landing in the middle.

It gets more troublesome when you start throwing in TNG legs. Now after a TNG leg you should, as mentioned before have a 3 minute gap before the next leg.
What sometimes throws things out is if you have two TNG legs without a direct flight full stop landing in between.

Remember that for the best and most consistent results you should not fullstop after a TNG leg. You should follow it with a direct full stop leg.
So in practical terms you should TNG from A to B for a period of time, followed by a direct B to B (with the obligatory 3 min space) and the aircraft lands cleanly.

What you have is a TNG leg A to A followed by a TNG leg A to B followed by a direct leg B to A - if I understood everything correctly.
What you have is back to back TNG legs and this can cause problems.
What we have seen over the years is that in this case the results are never the same for everybody and at least some get issues.

What you should do is eliminate the first leg. So just TNG from A to B followed by a direct leg from B to A.

Hopefully you didn't nod off in the middle of that!

What I am intrigued by is your comment of not always being able to add traffic to a third party that always has scenery.
I have honestly never heard that. The only reason that you couldn't add AI traffic is if the afcad is not AI friendly, i.e no parking spots, taxiways not linked. Things like that.
If you think of it in the truest sense of the word all of our scenery is third party, to MS/GD, and if you mean third party as non-MAIW then again my suspicions fall on a non-AI compatible afcad.

If you wish me to investigate for you then pm me and we can arrange for me to test the plans/afcads.

[hobby]

I've just written a reply to Firebird to be informed that I am not allowed to reply on this forum! I'll have another go now!

[hobby]

We have a slight misunderstanding regarding my current Canberra flightplan.

My intention is that a Canberra takes off from base, completes one TNG at base and 3 minutes after TNG flies directly to airfield 'B', lands and parks. No TNG whatsoever at airfield 'B'. After 60 minutes the Canberra takes off from 'B' and returns directly to base at which it lands (no TNG) and taxies to parking awaiting next flight.

[Firebird]

That would be within the rules, the trouble would be in getting it only to do one TNG. The time would depend on whether it was VFR or IFR and if its VFR then weather can come into play as well.

It may be that there is a minimum leg time, although I confess that I have not heard of such a thing, but it might be worth extending your TNG leg time by a couple of minutes if you get issues - even if you end up with two TNGs.

No what about this unable to add to third party airfields thing.

[hobby]

Firebird, I tried to send you a long PM a few minutes ago but being unused to the MAIW PM system this may have failed to reach you. My PM covered AI traffic and third party scenery as well as the history behind my current plans for a series of Canberra flights. Please let me know if you received that PM. If not I'll spend another 30 minutes retyping it - I am definitely not a trained typist!

[Firebird]

I have just checked and there is definitely no pm.
What I suggest is that if you have a long and/or intricate pm then before you post it copy it to notepad so that if anything happens it is very easy to resend it.

I have been caught out a few times like that and it is always on long posts. I think that the system allows you a certain length of time before the link to the database is lost and therefore no post appears.