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

[foamyesque]

@m4ti410: For wing parts with built-in control surfaces I haven't added code to properly account for this yet; you'd probably want to use only FARControllableSurface though.

The way FAR here's the way FAR handles unused attach nodes:

First, get the orientation of the attach node with respect to the airflow

If it's oriented into the flow, multiply the value of drag added by the maximum pressure coefficient at that Mach number (M = 0, Cp = 1; M = 1, Cp = 1.28; M = Infinity, Cp = 1.86) This models the drag from compression on the front of the object.

If it's oriented backwards, if M > 1, divide by M²; this gives a limit on the minimum pressure coefficient and models the drag due to viscosity over the back of the object.

If it's towards the front, move the CoD to account for the little bit of positive stability given by the unused attach node.

Does this imply that aerospikes have less drag than other engines? Also, what about things like Clampotrons, which have both nodes and are radially attachable such that the node can be buried inside another part?

[ferram4]

@Esulin: Don't even bother trying to use the calculator; with stock aerodynamics the drag force on each part varies with that part's mass, which means that the ship's mass can be ignored in the calculation. However, with FAR installed, this isn't the case. The calculator's assumptions do not match the physics being simulated with FAR, and no matter what you do you will not get a proper solution.

@Bloodbunny:

How to report a bug so I can fix it in a timely manner:

1. Narrow down the plugins and mods that you are using until you can re-create the issue; if the issue is with FAR alone and you're running other plugins narrowing down the source becomes more difficult.
   
2. Recreate the issue using the smallest craft possible; this leaves fewer options for the source of the error.
   
3. Post detailed and specific instructions on how to recreate the bug, along with the entire output_log and the problem craft; this will allow me to easily recreate the bug on my end.
   

I'll need more info before I can go to work on it, especially since I haven't used that plugin. Without a log, I don't know where the source of the bug is.

@foamyesque: Aerospikes might have slightly less drag, but they taper so much more than other engines that I doubt the difference is that much. FAR also detects if an attach node is buried in another part just to fix the problem you came up with. Basically, that's handled in the "is this unused?" section of the code.

[Delta Force]

Any advice on how to attain lateral stability and control? My aircraft has a rolling motion to the right despite being completely symmetrical. It's manageable until I hit around Mach 3.5, then the aircraft begins rolling out of control. It's not dangerous by itself, but an uncontrollable roll rate at Mach 4.25 and 21,000 meters is concerning.

I also have a very slow turn rate from the rudder, with the aircraft occasionally stopping its yaw motion and flying straight despite full rudder application. The aircraft also has a minor roll motion towards the inside of my turns. When I stop making turns the aircraft yaws left and right several times like a pendulum and while the motion dissipates it continues for some time.

[grom]

How do you use flaps / air brake part of FAR? I would like to setup air brakes when doing reentry to smooth out the deceleration when doing reentry (slowing down more in thin atmosphere). And flaps for takeoff and landings.

@ferram regarding my issue that other week I believe that was a PEBKAC. Still trying to get my head around all of what this mod does.

[SkyHook]

Has anyone solved the issue with FAR that makes gravity turns impossible? I can't seem to do a gravity turn without making any craft spin end over end, falling out of control. I am pretty sure that FAR is the cause of this behavior..

[ferram4]

@Delta Force: First, make sure that you've struted your wings up properly; there is a current KSP bug where parts placed with symmetry under the same load do not deflect the same amount. The second thing I'd do is make sure that your vertical stabilizer is perfectly straight; at very high dynamic pressures it can cause noticeable yaw and roll rates. Those should take care of the causes of a constant rolling moment.

As for yaw stability, what it sounds you're really after is yaw damping; for this purpose what you're looking to do is maximize S_v * l_v, where S_v is the area of the vertical tail and l_v is the distance from the aerodynamic center of the vertical tail to the CoM of the plane.

I hope you're not trying to turn the plane using the rudder. If you are, instead roll the plane and pull back to turn. Use the rudder for nothing more than countering adverse yaw (yawing in the opposite direction you intend to turn due to drag differences caused by deflected ailerons) and damping yaw motion.

Keep in mind that as Mach number increases the lift slope near 0 AoA decreases (for very small angles, C_l = 4 * AoA * (M² - 1)^-0.5). So planes tend to get a little unstable in yaw as they speed up; you may be interested looking up the wedge tail of the X-15 since it also had yaw stability problems at high Mach numbers.

@grom: Well, first you place control surfaces where you would like spoilers or flaps. Then you use the Editor GUI's control surface tab to set their functions and how much they deflect. Then you set what action groups you want the flaps to respond to; there is one command for increasing deflection and another for decreasing. Then, in flight, you use those action groups to control the flaps while the spoilers respond to brake inputs (which I should change to be an action group itself).

@SkyHook: It sounds like you're building aerodynamically unstable rockets or taking a highly aggressive ascent profile that is not actually a gravity turn. You can try posting a picture and all of us here can try diagnosing the problem, but a good first step is to read the new suggestions section of the readme.txt, which has some basic suggestions.

[foamyesque]

So I built a thing.

I needed some big fins.

[SkyHook]

@SkyHook: It sounds like you're building aerodynamically unstable rockets or taking a highly aggressive ascent profile that is not actually a gravity turn. You can try posting a picture and all of us here can try diagnosing the problem, but a good first step is to read the new suggestions section of the readme.txt, which has some basic suggestions.

Well back to the drawing board I guess, I think my ascent profile is too aggressive. Thanks!

[grom]

I know terminal velocity with FAR depends on your design (that is how much drag is being produced), but does getting mach (white) or heat effects mean past optimal speed when ascending? Or in other words what do these effects mean when using FAR?

One thing I've been doing is limiting acceleration to 25m/s which seems to avoid these effects (cause sometimes my payload is so light that even with most efficient ISP engine still have too much TWR).

[ferram4]

@SkyHook: If you still end up having problems, just post a pic of the rocket and the ascent profile you're using and we can figure it out.

@grom: The mach effects and aerodynamic heating effects don't actually mean anything specific with FAR... they're just kinda there. Honestly, I'd love to hijack the mach effect and have it vary properly (only shows up near Mach 1, and very prominently) but they're just visual effects right now.

[grom]

@ferram4 so they just activate based on speed/acceleration and have no relation to drag? if so then I can probably be going harder with some of this rockets then. Get into orbit quicker

[ialdabaoth]

@SkyHook: If you still end up having problems, just post a pic of the rocket and the ascent profile you're using and we can figure it out.

@grom: The mach effects and aerodynamic heating effects don't actually mean anything specific with FAR... they're just kinda there. Honestly, I'd love to hijack the mach effect and have it vary properly (only shows up near Mach 1, and very prominently) but they're just visual effects right now.

Have you looked at how I'm doing it in DRE2?

[ferram4]

@grom: Based on my knowledge they are based on velocity and pressure.

@ialdabaoth: I looked at that when the first versions of DRE came out, but didn't consider it too much of a priority. Also, I didn't understand the code initially and I like to have at least some basic understanding of exactly what's going on before I use the code. So maybe next update I can adjust this stuff.

[Van Disaster]

Skyhook: usually means the nose is too draggy which makes the rocket unstable if it isn't pointing straight up it's direction vector; drag will be making it turn more. If putting fins at the base of the rocket doesn't work then you'll have to start turning higher where there's less atmosphere ( and consequently less drag ), or reduce the launch TWR a bit so you're not going so fast initially, or pilot very very carefully. Ferram mentioned an issue with the current KW Rocketry set where the rocket parts aren't draggy enough, so the fairings move the CoD way up.

[Carzum]

So I just installed the most recent version of FAR from spaceport, and I've run into a problem. I'm experiencing a complete absence of atmospheric drag...

I began with a clean installation of KSP along with a few mods: B9, KW, MechJeb, Tav's Aerospace, Quantumstruts and also the procedural wings mod. All versions of these mods being 0.20 compatible and installed in the gamedata folder. So after launching my first rocket, and getting to ridiculous speeds I figured something was wrong.

According to MechJeb I had 0.0 m/s^2 drag. 0.000 Drag coefficent, and no Dv lost to drag. So I figure one of the mods is responsible, and remove them all from the GameData folder. So now I have all stock parts, after letting Steam redownload the squad folder, and FAR installed. First thing I build is a 1-man capsule attached to the large SRB and I fire it straight up. It broke the sound barrier at 2 kilometers, and got a maximum velocity of about 2.3km/s. I don't think it should be possible to get to an altitude of ~600 kilometers only using 1 SRB, so I think it's save to say that even using pure vanilla, FAR is still not working properly. Now the strange thing is: I do experience decelaration when falling down: up to 25G... When coming down I will accelerate from 0m/s to 2.5km/s and when hitting the thick atmosphere to about 200m/s.

I have installed FAR by extracting the FerramAerospaceResearch folder and the modulemanager.dll into my GameData folder. Please note that extracting the contents of the 'Ships' and the 'Source' folder into my KSP route folder have not helped to resolve the problem.

EDIT: So after downloading KSP again, and also redownloading FAR from the mediafire mirror I am still experiencing these issues. What I did discover is that attaching parachutes to the rocket seems to activate some sort of drag. It's very little during ascent: about 150m/s loss in total. (Which means my test craft consisting of a 1-man module and the large SRB gets to about 150km. In vanillia it will go to 35-40km max) Descent however seems to happen in a nominal fashion.

Edited June 19, 2013 by Carzum

[Klajan]

One quick question about the memory usage of FAR. Does FAR increase the RAM usage significantly? Especially for large Rockets? Because since I have FAR installed my game crashes every second flight due to lack of memory (damn, I can´t wait for a 64 bit version of KSP) I have a lot (around 30) of mods in use, and noticed this issues only after I installed FAR.

[ferram4]

@Carzum: MechJeb works off of the stock drag model, which is zeroed out for almost all parts (excepting parachutes) with FAR installed. So, long story short, MechJeb is wrong and everything is working as it should.

If you don't believe that a large SRB pushing a tiny payload should be able to accomplish what it did with FAR installed then you're still too used to the stock drag model. I'd suggest you try looking at what the dV of that rocket is and figure out how high it would go if you only lost 300 - 500 m/s of dV to drag.

@Klajan: It does increase RAM usage for each part on rockets, as well as using up some memory just by being installed. This is because more memory is needed to calculate the aerodynamic forces and to reduce computational overhead (by quite a bit; I'm storing the solutions to the equations here and here). I could look into some memory optimizations, but I don't know how much I'd be able to do.

[Carzum]

@Carzum: MechJeb works off of the stock drag model, which is zeroed out for almost all parts (excepting parachutes) with FAR installed. So, long story short, MechJeb is wrong and everything is working as it should.

If you don't believe that a large SRB pushing a tiny payload should be able to accomplish what it did with FAR installed then you're still too used to the stock drag model. I'd suggest you try looking at what the dV of that rocket is and figure out how high it would go if you only lost 300 - 500 m/s of dV to drag.

Thanks for the response! It's always awesome seeing mod developers giving so much active support.

After trying a lot of things I kinda came to this conclusion for myself. Even though it feels strange considering the differences with the vanilla drag model, it does add up. My testing craft, the module/SRB one, has about 3000 m/s dV according to Mechjeb, but only reaches around 2.2km/s. And gravity losses don't fill that 800m/s hole. I always knew FAR changed the drag model in a pretty radical fashion, but the difference is pretty stark. In vanilla I could only reach a maximum altitude of about 40km, against 700km using FAR.

Anyway, thanks for responding and I'm glad nothing is wrong.

[MR4Y]

Well, after a few days messing with things, here's the SpaceShipOne + White Knight flying. Still needs two high thrust engines, so Touhou Torpedo's Mk IV's were used:

SpaceShipOne released:

And here is reentry position:

[grom]

@MR4Y impressive! What mod are using for the wings there? It looks to me like you using Damned Robotics but then the wings fit around that part so nicely if that is the case.

[praise the suuun]

Ive run into a bug after installing b9 aerospace. the cotrol surface configuration tab recognises b9 control surfaces but not stock surfaces. perhps this is an easy fix?

[ferram4]

You don't have a proper version of ModuleManager.dll, so the parameters for stock wings haven't been added. It is included in the GameData folder.

[Subcidal]

Is it just me or do capsules come in sideways during the last bit of their descent no matter what you do? Went back a few pages and haven't seen anybody talking about it.

[m4ti140]

Is it just me or do capsules come in sideways during the last bit of their descent no matter what you do? Went back a few pages and haven't seen anybody talking about it.

Well, that's interesting, bc there is a whole page in this thread about this issue...

[Chingus]

You don't have a proper version of ModuleManager.dll, so the parameters for stock wings haven't been added. It is included in the GameData folder.

I have the same issue, and I'm using the modulemanager.dll that came with your latest release. It works with stock surfaces before I install b9, but not after. Edited June 20, 2013 by Chingus