Military AI Works

The FSX Macdill scenery is all done as is the RB-47H for both FS9 and FSX.

I'm glad John B suggested this variant. It's more than just the stub nose that I was expecting. It's nearly 3ft longer and the bomb bay has been converted into a new pod. There's an asymmetrical pod under the right fuselage in addition to the one under the tail plus one near each wing tip with cameras in the rear fuselage. I've added 4 to Macdill as visiting aircraft:





Next job is to have some more fun trying to see if I can get JATO to work.

John

John, are you working "just" on the smoke effect for the JATO, or also on modified FDE?

Giorgio

I'm assuming a smoke effect with an aircraft.cfg light entry (like afterburners) will be the best option as it provides the trigger only at take-off. The thing is JATO generates a great deal of smoke, so I need to experiment with some existing effect files, to see what's possible. I don't think normal smoke files are powerful enough. The FS9 volcano effect is quite spectacular though!

I'm not really into designing effect files from scratch and that could be a limitation, but I might be able to modify something. I'm only 30% optimistic, but I should know by tomorrow evening if it's doable. I'll post some screen shots of the tests for a bit of fun.

John

Great aircraft, John. I love it!!!

Thanks John.
Yesterday I've tried some test on a C-130 using multiple light line with 5 effect close one to each other, using the fx_BMK_smoke_white, changed in fx_BMK_smoke_grey.
If I get something good, I'll send you a screenshot.

Giorgio

Martin sent me some C-130 JATO effect files Giorgio, but the result is nothing like those seen on a B-47 - they were much too tame.

I've been experimenting for 4 hours this morning with the following results:

Starting point was the default DDFire effect file:



Next came SmokeW:



Smokestack2 looked hopeful if I could turn it on it's side:



That was possible, but it's still too tame and undefined, even when an afterburner flame is added:



There's also a secondary problem at night:



The delay value in the effect file needs adjusting so the effect triggers after the aircraft is accelerating down the runway or the smoke tends to drift upwards like an engine fire. Blacklight entries are also needed to prevent Pure Virtual Function Call crashes.

The result isn't convincing enough. I think that's because the B-47 JATO smoke comes from 15 separate bottles on each side of the fuselage. They all merge together in a maelstrom of violent activity that's hard to achieve. A Google Search of "B-47 JATO" will show what I mean. Having 30 separate light entries with slightly different positions might produce something better, but would be frame rate prohibitive.

I think I'm going to abandon the JATO option. It would need a separate aircraft folder and I'm not sure there is much lasting interest to justify continuing the effort.

I'll spend some time on flight planning instead.

John

When I said I was going to spend some time on flight planning, I wasn't anticipating the problems and frustrations of the past week.

I started first by setting up an FS9 waypoint 300 miles south of Macdill just below the Florida Keys. That had an AFCAD at 20000ft with a tiny runway. I worked out the timings of a single IFR TNG flight plan by following the aircraft in slew mode. It did the TNG on time and climbed out, only to disappear about 10 minutes later. I'm told that's common practice.

I could also view the inbound aircraft from a viewpoint at Macdill, having located the IFR spawned aircraft about 15 minutes before arrival. That landed fine. However, the aircraft would only approach from the wrong end of the runway against the way it departed.

I took all the scenery files out of the folder except the AFCAD and tested that, but with the same result. I then changed the waypoint so it was less than 100 miles away. That too made no difference.

Next I tried a virtual waypoint in the airfields.txt file only with no AFCAD and a simple VFR flight plan with no TNG. That worked fine following the aircraft to the waypoint, where as expected, it disappeared on the approach. However watching it from Macdill on the return leg after spawning, it decided to toss the engine smoke into the air as it crossed the threshold. It does the same on an IFR approach, but again, the aircraft would only use the "wrong" end of the runway:



I've never seen the smoke problem before, but I'm told it does happen.

Note that IFR flights with the aircraft flying circuits of Macdill have no problems. Both runway ends are used and the smoke stays put.

So, two questions:

1. What is forcing IFR approaches from a spawned aircraft to use the "wrong" end of the runway? Could it be the close proximity of Tampa International airport on the other side of the Bay? I'm pretty sure it's not the AFCAD.

2. Is there a fix to prevent the engine smoke detaching when crossing the threshold?

I should have had this finished by now, but I'm at a loss as to what to try next, short of providing circuit only traffic.

Any ideas would be appreciated.

John

John

I have no idea why the smoke does what it does.

As far as the approach to the airport is concerned, I have seen this before. Loading at an airport and then changing the time of day sometimes changes the direction of traffic.

Try to put a wind in the weather options to "force" a particular runway. With no wind and no conflicting traffic, real aircraft land straight in despite the designated runway in use if cleared by ATC.

Hope this is of some help. I am assuming the spawning waypoint would be in the direction of approach.

John

It seems that there are no solutions for the smoke. I've highlighted the same when the first pack with the B-52 was published and later, with the smoke used by the aerobatic teams in my Pratica di Mare Airshow package. These use a different smoke effect file, with the same problem.


Giorgio

John, that was a really useful suggestion, thank you. I've never bothered with a wind value over my years of designing and provided the heading was correct, AI would land the right way round, at least for non-waypoint traffic as far as I can remember, unless it was pure luck. However, adding a value of 8 knots solved the problem with Macdill and waypoint traffic now lands appropriately, VFR or IFR.

The detached, elevated engine smoke I think as Giorgio suggests, can't be fixed. Mike Mac suggested inserting a TNG at the end of the return leg to force the aircraft into a circuit, but alas, although the circuit was performed in my test, the smoke still detached.

There's only 3 smoke options really - live with it, remove the smoke completely or program local circuit only flights where the smoke doesn't detach. I can't program flights to other airfields because I have none of that era. If the smoke stays, it then really boils down to having the aircraft disappear for 60 minutes (waypoint) with detached smoke on return, or an absence of 15 minutes (circuit) with smoke intact on return.

John

My vote goes for the first option.
Live with it.
As stated before, the problem is present in other MAIW packages. Didn't notice complains about it.

Giorgio

The engine smoke thing drives me nuts. I could swear that when adding engine smoke to AI was first a thing, there was a solution for the detaching smoke problem... but scouring the internet can't seem to locate it. I think part of the problem there is that we are talking about forum posts that would ten years old now and have long vanished.

I vote to leave the smoke and just live with it... after all, no one has yet to complain about all the C-130s and P-3 I've put out that all occasionally suffer the same bug.

It's been a frustrating week, but a valuable one. I've learned more about waypoint flight planning, both with an AFCAD and virtually. I've come to realise that I need to put some wind speed in the FS weather engine in order to guarantee the right landing runway direction and I'm now aware of the problem of detaching engine smoke.

I don't personally like the detached smoke, particularly after adding all the animation refinements for landing - flaps, and brake chute. Macdill has some very long taxi routes that will display the detached smoke for long periods on the way back to the parking spot. I could though kill the ground smoke (Type 1 beacon light) and just leave the Type 2 strobe light) to confine the problem to the landing run.

Either way, to satisfy everyone, I'll provide two FS9 traffic files and users can choose between circuit only (no detached smoke) and a mix of circuit and waypoint (some detached smoke). I'm glad to say that the smoke does not detach in FSX so I'll only provide a mixed circuit and waypoint file for that version.

Thanks to everyone who has helped me with the problems.

John

The B-47s and Macdill retro scenery are all done and zips, including a paint kit, were sent for inclusion in the Download Hangar this morning. The files are: JYAI_B-47s_FS9_1.zip and JYAI_B-47s_FSX_1.zip.

John

And released.....

Thank you, John - looking forward to stop coding and use them in the sim!

Bravo sir, bravo

Just been watching it do circuits on my Retro setup. Yet another wonderful addition to our AI skies. Thanks John

Thanks John for another great aircraft.
I have a problem with scenery. Anyone else have this?

[url=https://postimg.org/image/7yyicacjf/]

Giorgio

If that's FS9 Giorgio, I haven't seen that problem before and I'm sure the 3 other users who tested would have flagged it. I haven't put a ground polygon over that area so it's not a missing texture from the package.

Do you have the elevation stub file in FS9\scenery\world\scenery and have you switched off any modern day scenery of Macdill?

John

Not seen in my Retro setup either !