When are they going to fix ground handling in windy weather?

[Captain Nuts 93]

15 hours ago, jcomm said:

XP has lost my hopes again and again, version after version... I'm unfortunately not moved by nice-looking sceneries, PBR, and that sort of stuff I see most simmers worried about instead of actually placing their attention in the primary aspects, IMO, of FLIGHT simulation - as sound / accurate as possible flight dynamics modeling - as well as systems modelling.

Well it's maybe a bit early to lose all hope.  Default planes had and still have a quite limited accuracy in all sims I know, actually most of the payware stuff is not that much better (IMO) in all sims, particularly at the edges of the envelope.  Only very few models really stick out of the masses of planes with shiny visuals that roughly hit the POH numbers but fail to depict the characteristics and idiosyncrasies of a specific aircraft.  Maybe you have heard of Rob Young, he was one of the very, very few who knew how to hack/abuse a FDM to teach a plane details in behavior that were not deemed possible with the FSX/ESP technology before.  Likewise, the amount of XP planes and developers being that good is very limited, but the good news is that we may be getting there. 

For example, the REP for the 172 is the only thing I know (in all sims) that simulates the ground handling of a 172 with its spring loaded steering rods pretty dead on, and practically it doesn't suffer from the ground friction issues at all (because it models that in its own plug-in).  It still lacks a bit in the rudder authority department IMO (with a big improvement possibly coming) but the current beta already improved things a lot, it really nails stalls now, with some luck you may or may not end up in a spin and it's got a much more realistic general "crossed controls" behavior now. On top of an improved FDM you get better system simulations and a "persistent", "living" aircraft and all that for 20 additional bucks.  I've spend a lot more money for aircraft that deliver much less. [/REP commercial]

On 4/28/2019 at 5:47 PM, strider1 said:

Another issue is the lack of 'Left Turning Tendencies' on departure. The 172 should require lots of rudder input when slow at high angles of attack, but the ball shows no deflection in the turn coordinator. Have you contacted Austin with your theory?   

No, I have not contacted Austin for the simple reason that I think he knows that something is wrong, but I think he doesn't know what that actually is (otherwise it would've been fixed by now).  All I can do is submit a bug once I have some substantial clue on what might be wrong.  All I found out so far is no news to anybody, I just have tried to quantify "lack of rudder authority" a bit but I think this is just yet another symptom anyway.

Re the "right rudder demand", this is probably a bit more complicated.  To begin with, there are many slightly different models of "172" and for example, our club's 'H' model has remarkably little right rudder requirements for a couple of reasons and doesn't differ at all from other planes we have. It depends a lot on rigging, power, prop. load balance and of course wind vector anyway and so I think that can't be generalized easily.  If you try the default 172 with absolutely no wind there is some noticeable right rudder requirement.  Of course I also think that something is wrong with the default 172 in this department, it flies almost auto-coordinated even in steep turns and "stability augmentation" has the exact opposite effect of what I'd expect - it needs more rudder to coordinate turns when SA is at 100% and very little when there is no augmentation. Just odd.

On 4/28/2019 at 5:47 PM, strider1 said:

Edit: Ok just been playing around with a 20kt crosswind using the default172 with the EFM(experimental flight model), and had no problem keeping the plane line up on 3mile approach. Plenty of rudder authority! I also tried it with the REP version and it was twitchy, with EFM on and off. EFM with the default 172 seems like the way to go with a crosswind!   

I just tried the default 172 again with the experimental flight model in my 20KTS crosswind scenario and I couldn't see any difference.  What do you mean by "line up on a 3 mile approach"?  Are you performing a long sideslip approach or are you flying crab angle down the glideslope?  I can't see any difference during the sideslip/wing down phase on final, at 20KTS I can't align the nose with the runway even at 70KTS over the threshold.  Mind you, if you have a wide runway it doesn't feel or look that wrong if you touch down with the nose not perfectly aligned with the centerline and just use a somewhat diagonal path to counter the missing few degrees of "kicking the nose straight".  This doesn't work on a typical, rather narrow European runway or back country strip tho.  Only a replay with the right camera angles tells the whole story.  IOW "video or it didn't happen!". 🙂

 

 

Edited April 29, 2019 by Captain Nuts

[jcomm 5,939]

Thank you Captain for your post,

and yes, losing hope is a somehow an "overdone" way of putting it, but those who know me, apply the necessary "filters" to my posts 🙂

Now, back to the effects that I notice even in that REP, and I had the REP and the Airfoilabs 172s, there's that thing with a huge right turning tendency ( instead of left ) that began somewhere along XP11. 

The first thing I did to overcome it in the default models ( and some add-ons too ) was to reduce the trim settings and eventual vertical stab and engine(s) cant that were there before to overcome the "old time" overdone torque effects. But even in those fine tuned prop models, and not only these 172s, there's that tendency to see a right turning tendency even climbing at high power settings. 

There are other theories about it's origin(*), but it feels so unnatural that I tried everything to forget about it, including using trim where it shouldn't be used... ( "Use trim" - as Austin once suggested when I was complaining about that overdone torque effect that plagued X-Plane from mid v9 until Murmur "fixed" it... )

Then, there's that "sucking" ground effect that came to live sometime ago. This is yet another newcomer to the XP FDM saga. Our airplanes have some sort of "inverted" ground effect where they're "sucked" into the runway in the final steps of a landing, when entering ground effect. It's another weird and anti-natural effect that appears to look a lot like the effect of a generalized implementation of an effect Austin "discovered" made sense regarding some pitching moments caused by tail interaction with ground in some airliners.

(*) like fuselage interference with propwash, or the way propwash effects over the inner sections of the wings affect the associated asymmetry of lift and create a differential that overcomes and actually outperforms the torque an slipstream effects that should instead contribute to the left turning tendencies , and / or a lack of proper propwash modelling in terms of it's interference with the tail surfaces and fuselage ... 

Edited April 30, 2019 by jcomm

[Captain Nuts 93]

 

11 hours ago, jcomm said:

Now, back to the effects that I notice even in that REP, and I had the REP and the Airfoilabs 172s, there's that thing with a huge right turning tendency ( instead of left ) that began somewhere along XP11. 

The current default 172 or B58 (which was really bad back then) or their REP combinations do not exhibit that crazy right-rolling tendency here, I think since 11.2x something.  The REP B58 is distinctly rolling left (not sure it should but at least it doesn't roll to the wrong side). Did you try any REP update lately?

But yeah, I almost forgot about the right rolling,  or pushed it to the back of my mind.  It was the first thing that really rubbed me the wrong way when I bought XP, the more torque, the more rolling to the wrong side, always taking off with crossed controls, I was not amused. 🙂 Back then I still had some strange and pretty far-fetched ideas like checking out the possibilities to build a pretty ambitious training device at our airstrip and bought XP11 (more or less on the way to purchase P3D) to see how it evolved and if it could deliver a solid software base, and that right rolling issue alone made it look absolutely hopeless to me.  Like 2 decades of bragging about the superiority of the "blade theory" still in my ears, it looked like it couldn't do any basic phase of flight right, maybe except straight+level cruise.  So believe me, I hear ya!

Luckily something else kept me from just throwing the DVD box into the trash bin - it was Ortho4XP and OSM overlays, an unheard before opportunity to meticulously recreate the surroundings of our airstrip and the usual action radius of our club and school in a way that looks absolutely familiar, with all the visual clues marking traffic patterns etc. then I added POIs, later super-realistic night lighting (we do a bit of night flying here), really intriguing prospects (not only) for my aforementioned silly ideas and a whole new way to spend insane amounts of time. 🙂

Eventually the right-rolling issue got fixed and meanwhile I had found a few GA planes with solid ground contact and things looked much, much brighter to me.  Like I said somewhere above, it wasn't much different in the other sim, very few planes were actually worth even thinking of building a home cockpit, not to mention silly ideas like getting some kind of training device certified.

11 hours ago, jcomm said:

Then, there's that "sucking" ground effect that came to live sometime ago. This is yet another newcomer to the XP FDM saga. Our airplanes have some sort of "inverted" ground effect where they're "sucked" into the runway in the final steps of a landing, when entering ground effect. It's another weird and anti-natural effect that appears to look a lot like the effect of a generalized implementation of an effect Austin "discovered" made sense regarding some pitching moments caused by tail interaction with ground in some airliners.

For some reason I never experienced this, maybe because I rarely use different planes than my few favorites.  I just noticed the lack of a ground effect, but this is something the experimental FDM seems to change.  I hope the increased interest in XP will also lead to more improvements in that regard and of course the basic, or should i say "neglected", weather simulation, as simulation of the stuff a plane flies in is just as important as the FDM IMO.  Yes, many more features are utterly half-assed, and depending on what preferences people have there might be better choices but for low-and-slow GA fans and tinkerers the alternatives look way more bland.  It doesn't even have to be LR improving stuff, like many of the really cool things for FSX came out long after development of the sim ended.

 

Edited April 30, 2019 by Captain Nuts

[Janov 2,383]

On 4/28/2019 at 2:51 PM, Murmur said:

Hi Jan, I don't have a pc right now so no access to XP, what is the max xwind speed you can consistently safely take off with the current C172?

Sorry, I was away from my main PC at home (with flight controls) for a few days.

Default C-172 at default weights and CG on a dry, paved runway I can:

• taxi just fine doing turns into wind, taxi straight etc. in winds of 30kts
• take off and land fine with straight crosswinds of 20kts

I run out of rudder authority at about 22kts crosswind, but this *can* be helped with some nose-down elevator (to increase nose-gear friction). Bear in mind that the demonstrated max crosswind for the C-172 is 17kts, if I am not mistaken. Sure there is "hearsay" evidence of it being able to handle MUCH MORE - but not sure if that evidence qualifies as being "scientific" enough for flight-testing.

Also bear in mind that "high altitude" winds will translate unabated to ground-level winds. Often people load up some "real weather" winds of 100kts at 35000 feet and then wonder why they can´t control their planes during taxiing.

Also remember that controlling a plane in a desktop simulator is much harder than the real one - you will pick up any unintended deviations from the desired attitude/heading much later, because you have to rely on visual cues instead of rotational sensation.

Cheers, Jan

 

[jcomm 5,939]

Thx Janov for the additional info.

I'll give it a try again one of these days. Also will test the latest REP version and see...

Will try to report back.

 

[FDEdev 955]

15 hours ago, Janov said:

Bear in mind that the demonstrated max crosswind for the C-172 is 17kts, if I am not mistaken. Sure there is "hearsay" evidence of it being able to handle MUCH MORE - but not sure if that evidence qualifies as being "scientific" enough for flight-testing.

The demonstrated xwnd is 17mph (15kts) and this is well below any really limiting wind speed. Even the 172 manual states that the mentioned 15kts are not considered limiting.

Demonstrated only means that this was the maximum crosswind component they encountered during the certification process, nothing else.

The even smaller and lighter 152 can be easily controlled in a 25kts crosswind IRL without getting the impression that you are at the actual limit.

Xwnd and turbulence are one of x-planes weakpoints since decades, but ar least everything is more shiny now with PBR etc.

[joelpacheco 9]

This ground handling "issue" is what sent me back to FSX after buying x-plane 10, and barely used it since.  Last week I bough XP 11 just to try zibo.  Had many enjoyable flights until I realized I had been flying with clear weather. Rolling out zibo with 17kts crosswinds from the left was nearly imposible. The plane was pushed to the right like a paper plane, and once off the runway it was violently pushed to the left that almost hit the terminals. I'm not pilot, just many years of 737 passenger experience, but that doesn't seem right.

[Janov 2,383]

Oh well, you believe what you want 🙂

Jan

[strider1 575]

On 4/29/2019 at 3:28 PM, Captain Nuts said:

Re the "right rudder demand", this is probably a bit more complicated.  To begin with, there are many slightly different models of "172" and for example, our club's 'H' model has remarkably little right rudder requirements for a couple of reasons and doesn't differ at all from other planes we have. It depends a lot on rigging, power, prop. load balance and of course wind vector anyway and so I think that can't be generalized easily.  If you try the default 172 with absolutely no wind there is some noticeable right rudder requirement.  Of course I also think that something is wrong with the default 172 in this department, it flies almost auto-coordinated even in steep turns and "stability augmentation" has the exact opposite effect of what I'd expect - it needs more rudder to coordinate turns when SA is at 100% and very little when there is no augmentation. Just odd.

I just tried the default 172 again with the experimental flight model in my 20KTS crosswind scenario and I couldn't see any difference.  What do you mean by "line up on a 3 mile approach"?  Are you performing a long sideslip approach or are you flying crab angle down the glideslope?  I can't see any difference during the sideslip/wing down phase on final, at 20KTS I can't align the nose with the runway even at 70KTS over the threshold.  Mind you, if you have a wide runway it doesn't feel or look that wrong if you touch down with the nose not perfectly aligned with the centerline and just use a somewhat diagonal path to counter the missing few degrees of "kicking the nose straight".  This doesn't work on a typical, rather narrow European runway or back country strip tho.  Only a replay with the right camera angles tells the whole story.  IOW "video or it didn't happen!". 🙂

 

 

2

Its strange because 172N that I rent requires no rudder on departure and no rudder at cruise(magic rigging!) But once upon a time it did require lots of rudder on departure. But the S-models that I rent require lots of rudder on departure. Regardless I would still like to get my money's worth out of my rudder pedals 🙂  

In the flight configuration menu, you can set up a 3m approach instead of starting on a runway. I was using the sideslip for wind correction to keep the plane line up with the runway centerline. Make sure you are using a 100% default 172 with no 3rd party tweaks. I have my 'Flight Models Per Frame' set to 4 and I have a custom 'Edit Response Curve', not sure if that makes a difference. In real-life landing on the centerline with a 20knt crosswind is very challenging! And if you have 20kts of wind then it's probably gusting above that! Good luck! Of course, practice makes perfect, but that's an expensive proposition if you are renting! 

Remember when everybody used to complain about rotating planes when parked/stopped in a crosswind? About a year ago after a windy/gusty day of flying, I parked the rental plane on the ramp and forgot to set the parking brake, next thing you know to my amazement the plane starts to rotate with a gust of wind! First thought was shi-t The second thought was Austin crosswind-ramp-flight dynamics are spot on 😉 lol........  

 

Edited May 2, 2019 by strider1

[soaring_penguin 428]

I did a flight in south of France (edit: with the default Cessna), 20 knots on the map (real weather), it is spring so expect there to be some thermals.
Without the Experimental Flight Model (EFM): the plane keeps turning right most of the time while airborne.
With EFM enabled: I get tossed left and right (as expected) but with left hand turn tendency hands free.
I don't really have a problem with taxiing in either mode, but with EFM it is a bit easier.

Edited May 2, 2019 by soaring_penguin