Glider PIK20 mod...

[jcomm 6,095]

Sometime ago I participated in a couple threads at the .Org regarding soaring flight in XP12. 

Back then we were discussing, among other aspects, the fact that X-Plane's flight dynamics doesn't cope with some features common to most higher performance gliders, like th euse of flaps with negative settings.

Some authors tried to overcome those limitations, but all of the models I tried didn't convince me 😕

Recently a payware HpH 304 S Shark HPG, which also uses some tweak to model negative flaps, but I haven't purchased it (yet), so, I was tempted to try this PK20 mod, released yesterday, and the initial tests I ran revealed very positive outcome!

It's free, and you need to download the original version, also from the .Org ( link in the MOD description text ). A link to an important PDF is also included:

PIK-20B Flight Model Mod (negative flaps) - General Aviation XP12 - X-Plane.Org Forum

[blingthinger 672]

Each flap setting gets its own airfoil lift and drag coefficients. Doesn't get much more realistic than that.

Seems to be hitting the glide numbers too.

[jcomm 6,095]

And btw, thx @blingthinger for the link you sent me yesterday !

[blingthinger 672]

No problem. It appeared to be something worth digging into. According to Dan (the base acf author), it's a first for XP gliders. And to think those datarefs have been in XP for "decades"?? The capability has been in XP all along.

 

10 hours ago, jcomm said:

HpH 304 S Shark HPG

Now I'm very much wondering what this dev did to achieve the negative flap behavior... and how well it performs. Maybe that Pik20 mod is actually second to achieve the feat?

Edited June 15, 2023 by blingthinger

[jcomm 6,095]

Further to this great release, and the rather ingenious approach followed by @cmemory it's developer, I would suggest that you read across a thread that "started" this idea and ended up in what I consider to be the most sophisticated, and for the very first time rather "realistic" glider for X-Plane I ever used.

I am aware that a payware glider has recently been released, but since I am only marginally using XP12 these days and don't really want to spend money in addons for it, I had to pass on that one. Supposedly the author of that other glider also tried to circumvent XP's FM limitations regarding the modelling of flap aerodynamics, but I don't know the outcome... 

Glider flaps - Flight Model - X-Plane.Org Forum

[blingthinger 672]

5 hours ago, jcomm said:

Supposedly the author of that other glider also tried to circumvent XP's FM limitations regarding the modelling of flap aerodynamics

The org store description says they modified the airfoils. Probably static adjustments like the original Pik20 author tried to do. I tried the original Pik20 and it's a bit crazy. Like completely unstable at times. This mod is night and day difference.

 

5 hours ago, jcomm said:

Glider flaps - Flight Model - X-Plane.Org Forum

Wow... last public record of that dataref being mentioned is 10 years ago! For a boat simulation. Ha!

You know, now that I think on it didn't you say some time ago that @Murmur had also suggested some idea like this for negative flaps to Austin? Something about using different polars? Of course this can't be the first time anyones ever thought about the differences generated by flaps.

But mixing the ideas of updating the forces on the fly and finding that "dinosaur bone" dataref to actually make it work.... it's a really cool combo. Seems to work really well.

This could maybe help most any airfraft model. XP flaps are already very good and 90% of the time and they're only used at one point in the flight: low speed near runway. But if someone could run the models for different flap settings (e.g. with CFD), it would be the near-ultimate realism. Though the mod author says also that it didn't work well for braking flaps. So a full Fowler configuration maybe won't work as well as expected. Possibilities...

Edited June 16, 2023 by blingthinger

[blingthinger 672]

5 hours ago, jcomm said:

"realistic" glider for X-Plane I ever used.

How does it compare to Condor Soaring so far? That is obviously the benchmark glider model.

Edited June 16, 2023 by blingthinger

[jcomm 6,095]

53 minutes ago, blingthinger said:

How does it compare to Condor Soaring so far? That is obviously the benchmark glider model.

There's no PIK20 in Condor, and, I haven't yet had the time to thoroughly test it against table data for the original PIK20 ... 

But from the handling I would risk to say that the only aspects not so well reproduced now are the "post-stall" region (although mushing stall might well be realistic and is characteristic of many gliders), as well as the instruments that would still need some additional care & filtering in order to better reproduce the indications their real world counterparts give.

But it is very good, the best so far in XP for me, and surely a night & day from the original model...

For testing I'll refer to:

Eiri Pik-20 / Siren Pik-20 (all-aero.com)

as well as an available PDF Manual.

Edited June 16, 2023 by jcomm

[cmemory 6]

@jcomm

Whoa, I totally forgot I had created this account. I'll beg pardon if one forum is already enough to fill this brain 🙂. Thanks much for the review on Org and kind words above. Didn't have a clue this would work but was encouraged after the first conversations with you and flightwusel on the Org. And it works! Here's hoping it will get better with future tuning.

The datarefs I used are such a perfect match for gliders with negative flaps. And then Dan+Pik20 approved the experiment? It's a great 3d model plus in the .zip he includes some articles with flight test data that helped so much over the past few weeks since I started the digging.

And I don't hesitate to say that most of the credit here goes to XP's flight model. All PlaneMaker authors have to create or use airfoil coefficient polars, right? I just squished a whoooole bunch of them into a C++ array and then that into an interpolation loop to dance between them all. I had to tweak a few minor things here e.g. conversion from XFOIL to .afl, but   was "emergent behavior" from there (Austin's F-4 article words). Put some real forces on the glider airframe and it just worked.

Oh and for any curious who might have read some of my statements on org earlier this week: LR confirmed the dataref bugs that flightwusel and I saw (20 yr old dref, so rusty) and it got a bug tracker ID. Various LR also asked follow up questions that made me confident that it will get fixed. Dunno when with 12.06 incoming though, which I'm pretty excited for...

[flightwusel 10]

8 hours ago, cmemory said:

LR confirmed the dataref bugs that flightwusel and I saw (20 yr old dref, so rusty) and it got a bug tracker ID. Various LR also asked follow up questions that made me confident that it will get fixed.

Wow, that's great. I am amazed how well that all turnt out!

[jcomm 6,095]

Not wanting to sound pushy but...

Now that you've mastered the negative flaps performances for XP!2, could you find some of those "dynamic datarefs" allowing for a modified lateral fuselage drag so that we can have realistic sideslip performance?

This - sideslip performance - is one of the aspects where XP still performs rather bad. The PIK20 is no exception - you can fully deflect the rudder, either side, keep wings level or even a bit down upwind and... the descent rate is marginally affected - you can actually even climb happily 😕

This is not a problem with your mod but rather something that is related to poor fuselage aerodynamics modelling in XP. But maybe there are datarefs that could be used to fine tune the fuselage drag, or to create drag from other hidden wings / surfaces that could be put to play? Or simply by, again, using a new set of airfoils that can be dynamically enabled when beta values due to sideslip are detected?

[cmemory 6]

On 6/17/2023 at 2:31 AM, flightwusel said:

Wow, that's great. I am amazed how well that all turnt out!

That makes 2 of us!! 🙂

 

17 hours ago, jcomm said:

could you find some of those "dynamic datarefs" allowing for a modified lateral fuselage drag so that we can have realistic sideslip performance?

Not pushy at all. As far as I can tell Zibo is doing exactly this. And that's one of the reasons I filed a bug report so early on. Even though I didn't have an idea if this demo concept would work at all, I knew that if it did that we would see a lot of untapped potential in these datarefs. Fuselages in that list too. The datarefs work just enough to model the wings of this Pik pretty well. Controlling more of the geometry like this will require LR's fixes. Fingers are definitely crossed.