[1.3.1] Ferram Aerospace Research: v0.15.9.1 "Liepmann" 4/2/18

[D06P0]

On 10.07.2018 at 23:26, lordcirth said:

Какие еще функции? Все дело в изменении аэродинамики. Калькуляторы и тому подобное? Они имеют смысл только в том случае, если вы используете FAR aero.

Aerodynamics is a good thing!

 But when you want to play, and Kraft goes to the runway at a speed of 0.8 M and airborne brakes do not reduce the speed ....

maybe my settings / ship are not optimal

. question"?"

How to return to sink aerodynamics and use the possibilities of setting control surfaces?
 

 PS. Displaying EAS EXTREMELY HAPPY!

[The_Joe]

On 7/11/2018 at 9:48 AM, Gordon Dry said:

For convenience I added this small patch to my install:

@PART[*]:HAS[!MODULE[FARWingAerodynamicModel],@MODULE[ModuleLiftingSurface]]:NEEDS[FerramAerospaceResearch]:FINAL
{
	+description ^= :^:<color=orange>Not patched for FAR yet!</color> :
}

@PART[*]:HAS[!MODULE[FARControllableSurface],@MODULE[ModuleControlSurface]]:NEEDS[FerramAerospaceResearch]:FINAL
{
	+description ^= :^:<color=orange>Not patched for FAR yet!</color> :
}

 

How would I add this?  Im a bit of a noob.

[Gordon Dry]

38 minutes ago, The_Joe said:

How would I add this?  Im a bit of a noob.

Just create a text file, give it a proper name, change the file extension to .cfg, place the file into GameData folder.
I suggest to also create a folder inside GameData, like zFinal, where such custom patches are placed in.

On my system the file has this path:
GameData\zFinal\zzz_FARnoFAR.cfg

[Daniel Prates]

On 6/9/2018 at 2:41 PM, kcs123 said:

It is nice plane you have created here, but you are using moded parts that are not compatible with FAR. Whenever you use incopatible parts, you have arrows in SPH/VAB editor for COL mark. FAR disable those, but for some reason arrows showing up with such parts. You may not notice that something is wrong in flight at first, but you may expect strange issues, like NaN errors, FAR screen in flight not showing up properly values for AoA, Cl, Cd values, etc. It may be possible that on re-entry you don't get enough drag to slow down craft and similar issues.

Hi @kcs123 , only now I saw this post of yours. 

It never occurred to me that  FAR would only work with compatible parts, I always thought it  changed the most general aspects of aerodynamics and would hence work with every craft. Well, as far as ksp 1.3.x goes, I detected no difference or problem whatsoever (lucky me), even though my planes rarely have a stock part on them! I am not saying that there were no issues, but there were none I could promptly see anyway - planes would fly perfectly, with that extra difficulty that  FAR is supposed to provide, compared to the stock aero anyway. I actually never even thought it could not be working fine. 

But indeed, I have recently tried out a heavily modded KSP 1.4.4 and FAR will make a mess of things, sadly. The CoL will not appear anywhere in the craft (it just lies there dead on the floor of the SPH), and no wing at all will provide lift. 

Thanks for the heads up.

[kcs123]

I noticed that issue only with Airplane Plus mod, with cockpits, fuselages and such, beside wing surfaces and control surfaces. I didn't investigate config files in more details to find out why. Usualy cockpits and fuselages should not influence FAR at all, due to voxelization shape based on colliders that FAR use for more accurate calculations. Usualy all other mods that are not wings and control surfaces work just fine with FAR. Must be that some mods use something extra in config files for drag cubes that confuse FAR somehow.

Could also be some other mods around that cause same or similar things, but first indication that something is wrong is strange look and behaviour of CoL in SPH/VAB, even when you don't notice errors in log. I'm glad that it was useful to you and others to detect issues.

Edited July 17, 2018 by kcs123

[Cooper42]

On 7/17/2018 at 5:46 PM, Daniel Prates said:

Hi @kcs123 , only now I saw this post of yours. 

It never occurred to me that  FAR would only work with compatible parts, I always thought it  changed the most general aspects of aerodynamics and would hence work with every craft. Well, as far as ksp 1.3.x goes, I detected no difference or problem whatsoever (lucky me), even though my planes rarely have a stock part on them! I am not saying that there were no issues, but there were none I could promptly see anyway - planes would fly perfectly, with that extra difficulty that  FAR is supposed to provide, compared to the stock aero anyway. I actually never even thought it could not be working fine. 

But indeed, I have recently tried out a heavily modded KSP 1.4.4 and FAR will make a mess of things, sadly. The CoL will not appear anywhere in the craft (it just lies there dead on the floor of the SPH), and no wing at all will provide lift. 

Thanks for the heads up.

I’ve started reinstalling KSP.

 I used FAR last time and wanted to use it again. 

I had assumed, like above, that it automatically worked with parts mods due to the voxel based approach, and had thought about adding Airplane Plus to my mod list, so I’m glad I saw these posts.

 

Indeed, the first post here seems to suggest it should work fine with parts mods.

 

Is there any way of telling, other than testing every single part, whether a parts mod will or will not work with FAR?

Edited July 24, 2018 by Cooper42

[damerell]

14 hours ago, Cooper42 said:

Is there any way of telling, other than testing every single part, whether a parts mod will or will not work with FAR?

Just upthread there is a shim to tell you if a given control surface or wing is supported.

[kcs123]

4 hours ago, damerell said:

Just upthread there is a shim to tell you if a given control surface or wing is supported.

That only helps with wings and control surfaces, but not in case of cockpits, fuselages and similar parts. But at least it is something. For other parts you probably need to test it.

Edited July 25, 2018 by kcs123

[damerell]

36 minutes ago, kcs123 said:

That only helps with wings and control surfaces, but not in case of cockpits, fuselages and similar parts. But at least it is something. For other parts you probably need to test it.

I don't think there is any problem with other parts, whose aerodynamic behaviour simply follows from their shape. I think upthread you have probably inferred something incorrect from a botched mod install of your own. It's unfortunate that other users have been misled by that. As you'll see from the Airplane Plus thread, a user who has been using them together for a long time reports no problems other than the lack of support for control surfaces.

[dlrk]

Development still going on?

[ss8913]

So.. I compiled a version of this back on april 4, for 1.4.x; that worked up until 1.4.5.  Recompiled again (master branch from github) against 1.4.5, that seems to work.  JIC anyone was interested in having this working now.

[Pds314]

Regarding the accuracy of FAR's model...

1. I have noticed that large, long cones with nothing on the back have EXTREMELY low drag in the transonic, up until the angle of the mach cone is reached. Is this realistic?
2. I have heard from some people complaints that it is "too easy to make a plane supercruise." I have always dismissed this as stuff internal to the intakes/engines and mach reduction of compression ratios of real jet engines that simply don't apply to nearly-stock KSP, rather than drag as a barrier to supercruising, as things like the XF-104 can go faster IRL on 45 kN than an F-15 can on 160, with the primary difference being that the XF-104 is using an afterburner.

Edited July 27, 2018 by Pds314

[Pds314]

3 hours ago, ss8913 said:

So.. I compiled a version of this back on april 4, for 1.4.x; that worked up until 1.4.5.  Recompiled again (master branch from github) against 1.4.5, that seems to work.  JIC anyone was interested in having this working now.

Definitely interested. Not sure why 1.4.5 broke FAR but I'm happy to hear that there's a way to make it work.

[AtleSt]

4 hours ago, ss8913 said:

So.. I compiled a version of this back on april 4, for 1.4.x; that worked up until 1.4.5.  Recompiled again (master branch from github) against 1.4.5, that seems to work.  JIC anyone was interested in having this working now.

im very interested but i dont know how to recompile a mod, any chance someone could send me a link to an already compiled file?

[Maeyanie]

2 hours ago, Pds314 said:

2. I have heard from some people complaints that it is "too easy to make a plane supercruise." I have always dismissed this as stuff internal to the intakes/engines and mach reduction of compression ratios of real jet engines that simply don't apply to nearly-stock KSP, rather than drag as a barrier to supercruising, as things like the XF-104 can go faster IRL on 45 kN than an F-15 can on 160, with the primary difference being that the XF-104 is using an afterburner.

I used to think it was the unrealistic engines too, but turns out even using RJE engines, and with FAR on hard mode, it's still pretty easy to make a plane supercruise. Even a really bad one.

I've posted those screenshots both here and in a Github issue before and Ferram says it's legit... I'm not entirely convinced, but, I'm not an expert on aerodynamics, so I'm willing to accept that I'm wrong. It's definitely much better than stock KSP, anyhow.

Edited July 27, 2018 by Maeyanie

[Pds314]

14 minutes ago, Maeyanie said:

I used to think it was the unrealistic engines too, but turns out even using RJE engines, and with FAR on hard mode, it's still pretty easy to make a plane supercruise. Even a really bad one.

I've posted those screenshots both here and in a Github issue before and Ferram says it's legit... I'm not entirely convinced, but, I'm not an expert on aerodynamics, so I'm willing to accept that I'm wrong. It's definitely much better than stock KSP, anyhow. 

Well I mean, I'm not entirely sure RJE makes intakes realistic. It may be that RJE is doing some fudging with intake to allow actually-supercruising engines and turbojets to supercruise as expected, especially seeing as it is based on NASA's program and not a ground-up mod.

Edited July 27, 2018 by Pds314

[Maeyanie]

Just now, Pds314 said:

Well I mean, I'm not entirely sure RJE makes intakes realistic. It may be that RJE is doing some fudging with intake to allow actually-supercruising engines and turbojets to supercruise as expected, especially seeing as it is based on NASA's program and not a ground-up mod.

Those engines are copies of the ones the Su-27 uses, and it can't supercruise. The plane is a similar weight, too, so TWR should be similar.

But yes, it's entirely possible it's due to RJE not being realistic enough, particularly with the intakes.

[Pds314]

2 minutes ago, Maeyanie said:

Those engines are copies of the ones the Su-27 uses, and it can't supercruise. The plane is a similar weight, too, so TWR should be similar.

But yes, it's entirely possible it's due to RJE not being realistic enough, particularly with the intakes.

Well the Su-27 can't supercruise but I'm fairly sure some of the derived engines and later variants can. (Thinking of the late model AL-31F and AL-41FS and the Su-35 and variants)

Edited July 27, 2018 by Pds314

[Maeyanie]

Just now, Pds314 said:

Well the Su-27 can't supercruise but I'm fairly sure some of the derived engines from it and later variants can.

True, the Su-57 can, and it does use an improved version of the same engine. So that theory is certainly reasonable.

[Pds314]

5 minutes ago, Maeyanie said:

True, the Su-57 can, and it does use an improved version of the same engine. So that theory is certainly reasonable.

Well 57 is very different. At this point it's basically a Russian Wunderwaffe xD. I think the 35 can though as well though, which is important because it's a plane that actually exists in significant numbers and is cheap to make.
Literally the Su-57 has cost $10 billion for 10 flyable prototypes! It makes the F-22 and F-35 look affordable, and Russia's killed it too, meaning that even though most of that cost is R&D, it really is a billion dollar fighter.

Edited July 27, 2018 by Pds314

[Maeyanie]

4 minutes ago, Pds314 said:

Well 57 is very different. At this point it's basically a Russian Wunderwaffe xD. I think the 35 can though as well though, which is important because it's a plane that actually exists in significant numbers and is cheap to make.

Yes, sounds like it can too, at least with lighter payloads, and it does also use an improved (though less so) version of the same engine. So again, could be.

[fitemesquirrel]

14 hours ago, ss8913 said:

So.. I compiled a version of this back on april 4, for 1.4.x; that worked up until 1.4.5.  Recompiled again (master branch from github) against 1.4.5, that seems to work.  JIC anyone was interested in having this working now.

 

I'd love to have the binary for this.

Alternately, if you could quickly brief us on how to compile the source like via MonoDevelop, that would be awesome. 

[ss8913]

2 hours ago, fitemesquirrel said:

 

I'd love to have the binary for this.

Alternately, if you could quickly brief us on how to compile the source like via MonoDevelop, that would be awesome. 

I'm taking the easy route on this and just using the [free] community visual studio, since it seems to be set up for that, as most mods are.  you basically just have to load the .sln file into VS, add references where you point it at your KSP install's dlls . the references thing, you right click on the references and add references, point it at your ksp dir's ksp_data_x64/managed/ and the DLLs you need are in there.  add the references that it asks for until there are no warning symbols on any of the listed references.  then just F6 to build, and you'll have a binary now somewhere under your source tree; copy that to the appropriate place in your KSP install.

to get the source, look at like CKAN or the first page of this thread, download the .zip of the 'master' branch from github.. extract it somewhere; it uses long pathnames so somewhere like C:\temp\ is good to avoid windows's limit on path length.  Then do as above with VS.. VS community edition is a free download from microsoft.

I'm sure you can do this with monodevelop as well but I wouldn't know the specifics.

[fitemesquirrel]

Just now, ss8913 said:

I'm taking the easy route on this and just using the [free] community visual studio, since it seems to be set up for that, as most mods are.  you basically just have to load the .sln file into VS, add references where you point it at your KSP install's dlls . the references thing, you right click on the references and add references, point it at your ksp dir's ksp_data_x64/managed/ and the DLLs you need are in there.  add the references that it asks for until there are no warning symbols on any of the listed references.  then just F6 to build, and you'll have a binary now somewhere under your source tree; copy that to the appropriate place in your KSP install.

to get the source, look at like CKAN or the first page of this thread, download the .zip of the 'master' branch from github.. extract it somewhere; it uses long pathnames so somewhere like C:\temp\ is good to avoid windows's limit on path length.  Then do as above with VS.. VS community edition is a free download from microsoft.

I'm sure you can do this with monodevelop as well but I wouldn't know the specifics.

 

Thanks for the procedure, I'm sure it will help others. I've only got Linux so I don't have that option, but I'll keep kicking monodevelop and hope I can get it to load the references without crashing. Cheers.

[kcs123]

On 7/25/2018 at 7:37 PM, damerell said:

I don't think there is any problem with other parts, whose aerodynamic behaviour simply follows from their shape. I think upthread you have probably inferred something incorrect from a botched mod install of your own. It's unfortunate that other users have been misled by that. As you'll see from the Airplane Plus thread, a user who has been using them together for a long time reports no problems other than the lack of support for control surfaces.

There is problem with other parts. That is whole catch. It is issue that is hard to notice if you are not aware of. Sure, you can make a craft with those parts, sure it can even fly in some cases, depending on other parts used, but does craft behave like intended, to allow FAR calculating lift and drag properly over such parts ? That is why we love so much this mod, to calculate drag and lift on craft in more precise way to real life examples as much as possible. For me it is hard to belive that you will get good output data when you have wrong data on input, to start calculation. And for some reason FAR get wrong data on input for some reason when such parts are involved.

If you read few posts above, I'm not only one that noticed this:

On 7/17/2018 at 6:46 PM, Daniel Prates said:

But indeed, I have recently tried out a heavily modded KSP 1.4.4 and FAR will make a mess of things, sadly. The CoL will not appear anywhere in the craft (it just lies there dead on the floor of the SPH), and no wing at all will provide lift. 

Everyone can use parts from any mod in any way that fits in their game, but it is just something to be aware of and do not complain to either, ferram or to authors of other mods. FAR is not "officialy" updated to latest KSP, to tell that all issues are fixed from FAR side, so use it as it currently is, or don't use it at all. It is not nice to nag author of other mods to fix things on their side either, at least until it is discovered what exactly cause issues and be able to create MM patch on those parts to fix conflict. So, have to be patient and wait for proper FAR update first, before asking others to provide support. My best "educated" guess here is that is either, something configured in not ordinary way in config files, or colliders on part mesh were placed in some strange way to cause issue in voxelization.

Only way ordinary users can help in meantime is try to identify parts that cause conflicts, so it will be easier to find exact issue and what parts would need some kind of MM patch to solve issue.