I would delete the effectiveness scalars from the config, myself. Like you said, if the areas, angles and lift slopes are correct, there is only max_angle_scalar to play with
2 Likes
Interesting take! One airplane which stands out of the list with an similar approach is the
Douglas DC3: updrag 0.011 - downdrag 0.005
Itās also using a reasonable fuselage drag coef with 0.8 and it has CFD including vtail CFD activated, adjusted elevator and rudder lift coefā¦
so this is one airplane which ticks all the boxes. So itās almost like a developer is using all the tools and has some really good knowledgeā¦
and guess what, the flight model is created by our hero right here: Flight Model Review
definitely gonna try that one out now 
3 Likes
Someone begs to differ. Iāve never been this confident. 
The thing with these āāimprovement modsāā is what are they improving regarding the flightmodel and even more important: what are these āāimprovementsāā based on?
Did they correct numbers based on reality/ research or based on āāit should fly like thatāā ?
Not saying the durckworks mod is doing this, but most of these mods are killing the flightmodel, making stuff more stable/ easy based on nothing and people blindly believe this must be right.
Comparing the Duckworks mod with the MS version, he didnāt changed much in the flight model, besides adding the effectiveness scalars back in. And changing the vtail geometry to a higher value which is probably wrong, because this number should exclude the rudder area.
3 Likes
All package up with a pretty bow that reads ārealismā.
2 Likes
I guess Iām on similar lines regarding that mod. In my opinion, it is a must have for the improvements it makes with systems depth, but the aim of the flight model adjustments have always been questionable to me. From the get-go, we had a video from the developer explaining what he did and the real-world accounts to back them up, and then we had Duckworks with nothing more than āimprovement.ā Reading threads related to the subject, the main things I could glean from the changes were that they were motivated by first impressions - that it didnāt fly how people thought it should fly, and that it was difficult to land - and that the Duckworks mod fixed this.
Rob Randazzo from PMDG trashed the stock flight model at one point but didnāt offer any details beyond it seeming to not have enough rudder authority, too much aileron authority, and lack of adverse yaw āin the aileronsā (NOT a PMDG addon question, but RR flies a DC-3 so... - PMDG Simulations). He said the real DC-3 has neat qualities that were missing from the stock addon, but didnāt bother to name any of them. I guess it wasnāt worth his time.
Meanwhile, at least two posters with experience in the DC-3 were offering advice in a thread for people having trouble(What is wrong with DC3 takeoff ? How to fix and tailwheel lock unlock); they didnāt mention anything about a lack of realism with the flight model. One related to difficulties landing, especially in crosswind, by recalling that he had nearly ground looped it on his final mission after 900 hours flying it in Vietnam.
The Duckworks mod owes its initial success to solving negative first impressions of the DC-3 that ended up growing into generalized statements about its supposed lack of quality, etc. It ended up evolving into a very fine mod, and thereās no doubt Imenes has done some quality work, but I still prefer using it in combination with the stock flight model.
2 Likes
Here he explaines adverse yaw and the real world data at 28:09
and at 54:17 he showcases the adverse yaw in the sim, by simply using the compass:
I wished more developers would make these real world comparisons with their flight model,
but there are probably āāa few reasonsāā why they donātā¦
4 Likes
He got almost 10Ā° out of it and he was descending with an airspeed of 120mph indicated (20 mph faster than the tests). Iād say it passed the test. I love that bird.
1 Like
āthe cfd stuff messed up a few things and I have to dial it back inā
Did he? is the current default DC-3 flight model updated, or do we need a mod for that?
Itās diffuse for the users.
I donāt know how I could trust this gentlemanās undoubtedly theoretical knowledge but lack of real DC-3 handling experience, when every real DC-3 pilot commenting on the model has trashed it more or less.
exemplarily only one voice, R. Randazzo, PMDG owner, many hours on his own DC-3 as pilot:
ā It is so bad. So so so so so bad. It flies nothing like a DC-3. It sounds nothing like a DC-3. It operates nothing like a DC-3.
ā¦
It hurts, honestly. The DC-3 is such a marvelous machine to fly and this is aā¦ i donāt know what it is. but it isnāt a DC-3.ā
so I donāt know who we can trust as an expert with the Asobo MSFS SDK. Is there anyone?
āitās a sad sad situationā¦ā
We lack experts who know how to create and document(!) a comprehensive flight model SDK.
We lack experts who know how the SDK works.
We lack experts who know how the airplanes feel and fly IRL.
We know nothing. 
1 Like
My thoughts exactly.
And while I may not have any DC-3 time, my many thousands of hours in various light and medium-weight aircraft in all situations tells me one thing: the one video of his I watched part of (the DC-3 flight model one) should be subtitled āI make it painfully obvious Iāve never flown a plane beforeā. Theoretical based on SDK-numbers is alright but thatās not anything close to the biggest chunk.
1 Like
Few things irk me more than when people trash something or make some kind of dramatic comment and then donāt elaborate, as if they expect everybody to walk away thinking they have just been graced with the presence of pure wisdom. In that entire PMDG thread, I could pick out two people who had claimed time in a DC-3. Neither one of them bothered to mention anything specific about the flight model beyond control surface authority and adverse yaw, or that it ādoesnāt fly like a DC-3.ā In which ways? People want to know these things. There were a few tidbits about the sounds, the cockpit, and systems, which are all very important but have nothing to do with the flight model. Ironically, the most cogent comment came from somebody who had tried the two āimprovementā mods available at the time, but didnāt even get off the ground because there was no p-factor and he concluded they were both garbage. Again, please elaborate.
It has been updated as of SU12. I have both copies if you wish to compare, but it should be easy enough to see it was updated from its original version just by looking at it. I suppose it is easier to ask a sardonic question here than to check yourself. The Duckworks one was updated recently after having not changed much since it was originally released. It now includes a lift_coef_aoa_table that doesnāt go below zero lift and has no zero lift AoA, so figure that one out. Again, Iāll stick with stock for now and use the Duckworks mod for the improvements it does well, and for which its credit is well deserved.
We have at least one person in this thread who has provided explanations that are more than adequate and far more informative than whatās provided in the SDK documentation. Not all hope is lost.
Why would anybody take your comment seriously if you just admitted this?
Itās not that I donāt believe you, but what is the biggest chunk, and how does it so vastly overshadow ātheoretical based on SDKā?
1 Like
I updated the list integrating vtail CFD: VT
There are maybe a few aircraft which are already in the list with vtail CFD activated which I canāt check,
if you can find any feel free to comment, any info is very much appreciated.
search for CFD_ReinjectVTailX
in the flightmodel.cfg
Nice. Could you put a hyphen or red x mark next to the ones you know have the vtail CFD disabled? Iād be willing to check any that I have. Iāll also check on some others that I have to add to the list.
1 Like
I donāt think it would be explicitly disabled, it will just be absent from the config.
Iāll run a sweep of what i have tonight.
1 Like
on the side, donāt want to derail anything here:
1.) RS Randazzo owned a DC-3 for several years, his type rated first hand expertise is well documented, he doesnāt need to āclaimā anything. He just knows.
2.) he is an expert and business man with no time to waste, who knows where his input matters and where not. Why should he specify detailed feedback in a forum, where there are no direct beneficiaries for such information who would make a differemce? He offered his free(!) consulting role in the development of the DC-3 from early on and was ignored. As he is ignored by Asobo in his unique expertise in other relevant issues (environment model/turbulence)
What REALLY worries me, is how, evidently, Asobo seems not to care at all how much they ignore and frustrate their high end 3rd party developers.
That raises a flag. (that also directly affects the topic of this thread, flight model realism)
Arguably for many it is the high quality 3rd party makers, who make the difference for many for us, to invest into this sim and related products.
But Asobo obviously has no problem to ignore PMDGās offered free consulting. And make lemon aircraft (new ATR) that kill the known efforts and investment of 3rd party collaborators like Milviz who developed similar planes.
Which can only mean, Asobo/MS do not see this sim/game in their business strategy for the realism minded simmers. Itās meant to be a McDonalds kind of franchise. Mass market, not quality.
Because that was literally the subject of the thread - the question of the OP - and Randazzo, in the midst of being a business man with no time to waste, replied not once, not twice, but five times, never elaborating on why he thought it was so (so so so so) bad. That a few gauges donāt operate and control surface authority is off doesnāt really add up to the degree to which he very apparently abhorred it. It wouldnāt need to be that detailed - just something to distinguish honest criticism from just trashing the work of a competitor out of bitterness that Asobo didnāt welcome his free (with no time to waste) advice. He doesnāt exactly present without bluster. Where you are right is that, in that particular thread, it doesnāt look like too much was being asked in terms of detail. It seemed more tongue-in-cheek than anything.
No hard feelings; it just struck a chord with me and has been a constant frustration with the DC-3. A lot of people in the forums just seemed to jump on the bandwagon of declaring the FM was garbage because of initial impressions of it taking off weird and being hard to land, but never articulating anything beyond that other than the rivets look bad, gauges suck, systems depth - things not related to the FM. Then there were other people saying they donāt have an issue with the FM and trying to offer advice, but a lot of it fell on deaf ears in preference of the vacuous impression that it was all trash.
1 Like
This thread is about numbers and the effect they produce, to find out why adverse yaw is missing in most aircraft and weird yaw/ slip/ sideforce behavior.
That way we can maybe find out some common mistakes/ oversights developers are making and helping to improve the flightmodel.
Emotional statements like āyour complete flightmodel is trashā are useless and doesnāt help in any form if they are not specified and backed up by example. Doesnāt matter if this statement comes from a pilot, astronaut or the head of the US Air force.
So far we know aileron up and down drag doesnāt produce the big adverse yaw effect as advertised, the number size and their meaning are unclear.
Fuselage CX is much too low in most aircraft, because of the low default and suggestions and lots of developers doesnāt look into it + one drag value for many different areas is very limited/ inaccurate.
Then we have the vtail CFD which has a great effect on yaw behavior, which is not used very often and the elevator + rudder lift coefs, which have to be corrected in relation to the aspect ratio (default values are used very often, which is wrong).
Using oversized airplane geometry to achieve a certain flightmodel looks like a mistake somewhere else from developers.
I updated the list where itās sure they donāt use any vtail CFD.
Thanks to everybody contributing.
4 Likes
I have a couple to add:
SWS RV-14 (and RV-14A):
UD=0; DD=0; FX=0.2; VT=false
SWS RV-10:
UD and DD absent but the DD scalar is set to 1.5 in flight tuning, so DD would be 1.5; FX=0.6; VT=true
1 Like
Thank you very much, both added.
Maybe for UD and DD they are using a custom code somewhere,
Fuselage CX of 0.2 is funny 
SimWork Studios is known to be against the usage of CFD simulation, Iām quite suprised they are using it for the RV10, also the FX of 0.6 is more reasonable.
I wouldnāt say they were against it. More like needing convincing of the merit of retroactively adding it to an existing FM.
CFD has obviously come on some from when it was first launched with the reinjection. Also with Asobo deprecating so many of the legacy tables, CFD is clearly the way to go with new builds.
Perhaps they were using the RV-10 as a way of gaining experience to apply to bigger projects like the PC-12 (this is pure personal speculation).