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

[taniwha]

KSP's symmetry is not perfect. That (usually) tiny asymmetry will cause asymmetric aerodynamic forces. Those, in turn, will cause the vessel to turn in various directions, most readily roll. 69km (vertical) is a long way so the acceleration from those forces will a cause tremendous increase in spin rate.

Remember the paper + paper clip helicopter.

In other words: everything (except KSP's symmetry) is working properly. FAR is working as designed.

[SkyRender]

In other words: everything (except KSP's symmetry) is working properly. FAR is working as designed.

If the system is incompatible with the engine it's implemented in, at best you can say that the design of the system is insufficiently designed for the limitations of the engine. "Working as designed" is meaningless if "as designed" is a flawed specification.

[ferram4]

@UnanimousCoward: I cannot reproduce the issue, and I've been using Ven's stuff for a long time. If there's an issue, it's probably on that end; there's nothing I can do for parts that have been set up differently than all the other ones that exist.

@SkyRender: Of course the QBE dropped to only 70 m/s, it's a probe core. It's light and bulky, of course it's gonna slow down, regardless of whether it spins. As for the spinning, I'm sorry that FAR accurately represents the aerodynamics of the system that you have constructed in it. I know this results in things doing what you told them to do, not what you want them to do. The only way around this is for FAR to ignore flexing entirely, and to hand-wave away the aerodynamic effects of wings bending under load or rockets bending under high amounts of thrust. I will not do that. I will not compromise the simulation accuracy because you did not feel like designing for realistic effects; yes, real life vehicles flex unevenly too, and must deal with those effects.

It is working as designed, and models a real-life design problem. Now, adapt to it.

[NathanKell]

Ah, I see, so FAR should magically figure out exactly how KSP will cause this two year old unfixed bug in every situation, and then magically compensate?

[kcs123]

Flexing wings is actualy very usefull. It will show weak spot on your craft, so you can add struts in that area or increase strength/weight ratio on wings.

RCS Build aid mod is very usefull to find out tiny asymmetry, not only from RCS, but also main engines. You can see example in my post page or two before this one.

[DaMichel]

The uneven flexing can be quite annoying i guess. I have some planes which tend to fly stable in a slightly banked attitude. I presume this must be due to flexing. Quite oddly though, i also made hypersonic designs which fly wings level.

@Ferram4: Thanks for taking care of the intake drag

[Rakaydos]

Ignoring aerodynamics, adding a nosecone to the end of a tank will cover up the exposed node on the end of the tank which also matters.

Does that mean the Aerospike, nasa 4x, and LFB are naturally lower drag than their attach-point-allowed relatives?

[UnanimousCoward]

@UnanimousCoward: I cannot reproduce the issue, and I've been using Ven's stuff for a long time. If there's an issue, it's probably on that end; there's nothing I can do for parts that have been set up differently than all the other ones that exist.

I didn't mean to imply that it was a problem with FAR. Sorry if I gave that impression. I was just wondering if anybody might have any idea what's causing it.

From what I've read, it appears that not everybody has this problem, but I'm definitely not the only one. I posted about it in Ven's thread as well, and it turns out there's a couple of workarounds, so I can continue to enjoy both of these superb mods. Keep up the good work - it's mods like all of yours which make KSP such fun.

[Sparker]

I am making a Mars lander. It has a shape of a classic command pod. I want to create some lift during the atmospheric entry, but it behaves quite strange(in my opinion):

And compare it with Mk 2 command pod:

First, my lander has very low Cl. Second, it decreases very fast with increase of AoA, although my lander is more flat than Mk 2 command pod. Third, my lander has an absolutely reverse behavior of Cd curve.

I think that all it happens because of quite a complex design of the lander - sure, there are very many parts in it!

Do you have any ideas how I can make my lander create more lift?

[Rebelgamer]

Try building it in the SPH, not the VAB, and rotate the bottom towards the front. Will come out more accurate than using the VAB inverted. If your using DRE, don't bother sweeping past 25 on the AOA, you'll put too many parts in harms way.

[Rakaydos]

My eve lander is having problems with the parachute-spars shearing off the decouplers during reentry testing. (been testing at kerbin to save time.)

Which decouplers best resist aerodynamic failures from a lateral shear force?

[Sparker]

Try building it in the SPH, not the VAB, and rotate the bottom towards the front. Will come out more accurate than using the VAB inverted. If your using DRE, don't bother sweeping past 25 on the AOA, you'll put too many parts in harms way.

In SPH it is the same.

Currently i'm stuck on 10 AoA limit, because past that lift decreases, but i would like to have more lift, like on the command pod.

[NathanKell]

Are those proc fairings? The fairing detection code might be interpreting the heatshield as shielded by the fairings...

[Sparker]

Yes, these are procedural fairings, but i used them without the fairing base(i've found out it doesn't change much). No, the heat shield is not shielded. It says that it's not shielded and it also heats up.

Added later:

by the way, i've found one issue with procedural fairings. If you try to attach fairings in symmetry mode, they are not shielded(fairings themselves shouldn't be shielded, right?). If you attach fairings one after another without symmetry mode, they shield each other and they are all shown as not shielded.

Edited February 2, 2015 by Sparker

[Van Disaster]

Does that mean the Aerospike, nasa 4x, and LFB are naturally lower drag than their attach-point-allowed relatives?

That... is a good question which should be easy to test with the tools. I would say "yes" but sometimes these things get custom configs.

[Dermeister]

Dumb question probably but I'm trying to figure out How much lift my craft will have... Will it even fly? Does any 1 know? I've tried to look for tutorials but none really explain where to find lift only AOA and how it affects lift.. but I want to know if I put this bad boy on a Big plane then Unclamp it.. Will this glide? or fall like a brick.

[Van Disaster]

From the static derivatives:

* AoA at given speeds & altitudes - for instance to take off ( and land! ) at 100m/s your craft needs an AoA angle of 21.6 degrees.

Now from the main graph in the first tab you can check cD and cL at that speed and see if it will even stay in the air. If you want to to take off with somewhat less pitch then you need more wing, and the wing area is in the static derivative tab again ( although after a fairly short time you should be able to eyeball it ).

Repeat for other key speeds, like going supersonic at some altitude/mach 2 at 15km/mach 5 at 25km/mach 7 at high altitudes.

RE: engines with stack nodes, quick & dirty testing suggests a LV-909 is only very slightly more draggy than an aerospike, and I'm not sure how much that is due to AoA change for level flight from being slightly different mass.

Edited February 2, 2015 by Van Disaster

[Dermeister]

Thanks I don't want this to be able to take off like a plane... It's meant to glide like the space shuttle but I have to know if I use a bigger plane and mount it under then take off... then decouple( Like they did with Enterprise and the 747) to test how it glides I need to make sure it will fly lol. but an AOA of 22 degrees is defo a bit high...... I think I need more lift... I recon I think until I have a AOA of 10 Deg @ 100ms do you think that would make sense?

Edited February 2, 2015 by Dermeister

[tetryds]

Thanks I don't want this to be able to take off like a plane... It's meant to glide like the space shuttle but I have to know if I use a bigger plane and mount it under then take off... then decouple( Like they did with Enterprise and the 747) to test how it glides I need to make sure it will fly lol. but an AOA of 22 degrees is defo a bit high...... I think I need more lift... I recon I think until I have a AOA of 10 Deg @ 100ms do you think that would make sense?

That will be the last of your problems, look at the stability derivates, positive Mw is the cause of several kerbal casualities.

That means your COL is in front of your COM, probably because you have fuel + all engines on the back.

Try putting one of those tanks right behind the cockpit, using a ballast or shifting wing.

But remember that using ballasts adds even more weight and shifting wing drops your pitch authority.

I believe you should get more familiar with FAR before making a shuttle, they get exponentially more complicated as you build.

[Dermeister]

That will be the last of your problems, look at the stability derivates, positive Mw is the cause of several kerbal casualities.

That means your COL is in front of your COM, probably because you have fuel + all engines on the back.

Try putting one of those tanks right behind the cockpit, using a ballast or shifting wing.

But remember that using ballasts adds even more weight and shifting wing drops your pitch authority.

I believe you should get more familiar with FAR before making a shuttle, they get exponentially more complicated as you build.

<< <but I've clicked Update COL and it shows it's behind... but from those numbers they look like I can move the COL just a little bit further back...

[tetryds]

http://i.imgur.com/r2OehNz.jpg << <but I've clicked Update COL and it shows it's behind...

Check that table for different speeds, as the editor indicators do not compute at different speeds.

Also calculate them again after updating COL.

I guess what you should do then is what you said before, stick it on the top of a plane and check how it flies.

[Van Disaster]

Well if you're making a realistic shuttle then it did land very fast, around 215kts max ( although I keep finding that as mph in places - I'm going to assume they really mean nautical mph ), which is 110m/s; as a delta you can sustain pretty high AoA so you can just go with whatever you're comfortable with - check how much drag the craft is producing at various AoAs - but if you're going realistic then you should be able to get the pitch angle at landing the pilots aimed for from somewhere. It appears the final glideslope was 1.5 degrees, so with your craft as is that'd still be 20 degrees AoA to touch down which seems a bit excessive. So yeah, basically indicated AoA - 1.5 at 110m/s at whatever you feel comfortable landing at.

Edit: You are looking at your CoL ball at 0 AoA - the stability derivatives are giving numbers at 15 deg AoA, and I suspect CoL is now forward of CoM. If you go to the static graph make sure the yellow line isn't crossing the X origin before 15 degrees.

Edited February 2, 2015 by Van Disaster

[Dermeister]

Well if you're making a realistic shuttle then it did land very fast, around 215kts max ( although I keep finding that as mph in places - I'm going to assume they really mean nautical mph ), which is 110m/s; as a delta you can sustain pretty high AoA so you can just go with whatever you're comfortable with - check how much drag the craft is producing at various AoAs - but if you're going realistic then you should be able to get the pitch angle at landing the pilots aimed for from somewhere. It appears the final glideslope was 1.5 degrees, so with your craft as is that'd still be 20 degrees AoA to touch down which seems a bit excessive. So yeah, basically indicated AoA - 1.5 at 110m/s at whatever you feel comfortable landing at.

Edit: You are looking at your CoL ball at 0 AoA - the stability derivatives are giving numbers at 15 deg AoA, and I suspect CoL is now forward of CoM. If you go to the static graph make sure the yellow line isn't crossing the X origin before 15 degrees.

Qif you notice that's why I had my Graf set for .29 because 220mph = 0.2890147016968497 Mach wich is roughly 100 ms and I say roughly... and the shuttle lands at about 10 Deg AOA give or take but I'm not 100% sure about the landing AOA because my sources are dodgy so feel free to correct me.

But yea this is why I don't expect it to take off it only needs to land...BUT! thanks to your Input guys it's helping me correct the small issues that frankly I did not detect by looking at the grafs so your inputs are very helpful SO in short .. This -censored- needs more lift Twards the back.. !

Edited February 2, 2015 by tetryds
No need to say that.

[Duxwing]

Dear ferram,

Would you please make FAR write "Shielded" and "Unshielded" instead of "isShielded: True" and "isShielded: False"? I ask because changing this output would make your mod would become much-prettier.

-Duxwing

[undercoveryankee]

Dear ferram,

Would you please make FAR write "Shielded" and "Unshielded" instead of "isShielded: True" and "isShielded: False"? I ask because changing this output would make your mod would become much-prettier.

-Duxwing

Those context menu entries are provided for debugging and aren't displayed unless you've specifically enabled them in your FAR settings. For diagnostic output, I like to see the internal name in the debugging display as a reminder of what it's called if I'm working on something that needs to refer to it by name.