Engines starting way too fast

[frankbyte]

Hi there

What i noticed so far: the engine-start-procedure is way too fast. When injecting the fuel it spools up to 22% N1 in about 10 seconds...

Thanks for taking a look at that.

[tkyler]

Yea seeing the same...not sure when this one sneaked in on us, I know we tested it at some point.  Thx for reporting, logged.

-TK 

[tkyler]

So....good example of why this port to XP12 is a pain. ..and this is just one of a lot.

Went over this for a long while yesterday and just had trouble trying to code around it.  Took it directly to Laminar and sure enough, the acceleration ramp between lightoff and idle is simply inaccurate.  The acceleration curve shape between lightoff and idle is hard coded and not tweakable via PM.  Below is the N2 ramp for the default 738 (which has the same problem).  This is Laminar's own data of N2 vs. time (but my overlay comments).   You can see the very rapid rise of N2 after lightoff...getting to within 80% of idle in a few seconds.

So you may ask, "well why don't you custom code it"...and we do to some extent, but there is a limit.  Some parameters, if we decide to override, means its all or nothing.  So if we  "override the engines" for example..then we have to write the entire engine model for all regimes and that is almost a product in and of itself.  Besides, Laminar's model is mature and pretty good...no need to reinvent the wheel.....to me this is simply a "plane-maker design issue"  oversight.  A  'coefficient' should have been provided (and is what I'm lobbying for) to shape this curve...like a lot of curves in PM can be shaped.  

The good news is its official, we're filing a ticket to Austin.  Also good news (but not perfect) is I've coded a workaround for the next patch to get us through.....because I'm sure it'll be some time before the next update to XP comes out with this fix.  The limitation on my patch will be "don't mess with a running engine while the other one is starting".  If you do, then the starting engine will exhibit this "rapid rise" behavior somewhat. ...but many folks may not notice it.

and these kinds of sweeping modeling changes are all over the place, and have really wreaked havoc with the accuracy of our performance.  BUT....we're working on that too now.

TK

 

Edited September 18, 2023 by tkyler

[GusRodrigues]

3 hours ago, tkyler said:

to me this is simply a "plane-maker design issue"  oversight.  A  'coefficient' should have been provided (and is what I'm lobbying for) to shape this curve...like a lot of curves in PM can be shaped.  

A lot of fine tuning or better control of the (jet) engine FM with the current PM as a black box or and undocumented. the latest update in PM documentation is back from 2022.

[Litjan]

Yes, PM is woefully undocumented, but it is not a feature that gets the same exposure as X-Plane itself. It is considered to be a tool for experts that follow along with what is new and changed...not something that a total layman jumps into.

I think I know my way around aircraft pretty good, but it has me scratching my head all the time, too.