Control Surface, etc. Tweak Question

[Luckfish]

I've recently begun running several test flights in order to determine how to improve the controllability of my rockets. I'm running with FAR, DR, KIDS, and a few other mods. I'm using the same craft over and over, aiming for the same circular orbit, and and seeing how stable the ascent is and how much dV I have left over when I hit said orbit. With each flight I'm adjusting a different parameter. I'm using the ascent guidance autopilot in MechJeb only to try and improve the repeatability of subsequent tests by taking out pilot error (since I bought a joystick I enjoy flying my rockets a lot more than letting MJ do it, but I like the info that many MJ modules give me).

Some of the parameters I've been adjusting have been launching with and without FAR's Dynamic Control Adjust, having MJ limit ascent acceleration to various numbers, adjusting the ascent profile, having precise controls turned on, etc. One of my variables has been tweaking is the maximum deflection range of my tailfin control surfaces. In the VAB, right-clicking on one of the AV-8R's brings up the window to adjust a couple of parameters on the winglet, but I'm a little unsure about what the difference is on some of the tweakables. It appears I can toggle on/off the ability of the winglet to respond to pitch/yaw/roll commands, and adjust the amount of maximum deflection each gives.

1) The fact that there's two sections for each surface means that I can assign it to have a certain range of motion for pitch as well as yaw (for example)?

2) Also, I would imagine that pitch and yaw are more important to rocket ascent, so should I disable roll control altogether for all four winglets? Would engine gimballing and SAS be the only roll control I (or MJ) would have, and would that be enough?

3) One variable I haven't tried messing with yet is the variables (k and d, if I remember correctly) FAR has for the control limiters (accessable after you leave the VAB). I'm thinking that these are more akin to a type of 'gain' that gets factored into control inputs (whether made by me or MJ). Are these only active when the their corresponding FAR limiter is toggled? Should I be messing with these at all?

4) I'm sure I'm not the first person who has been down this path, so any advice on which variables I would be better adjusting, and which I should leave alone would be appreciated. Again, I'm only letting MJ get me to orbit for the repeatability of these tests. I've read that MJ doesn't do well with FAR installed, but I've gotten some pretty stable (albeit sluggishly controlled) ascents during some of my trials.

Edited February 18, 2014 by Luckfish
Moved question to FAR thread

[Claw]

Luckfish,

I tend to temper my comments when someone asks about FAR because I have no experience with it. But I feel bad nobody is trying to help, so here are my thoughts. (Maybe I need to get FAR so I can answer some of these posts...)

1) I would guess that if there are separate sections, then it probably controls the range of motion for each axis.

2) Pitch and yaw are more important for rockets (generally). You can disable roll if you want, but to go back to your original point of testing, I doubt that disabling roll will get you much improvement in performance. If you are going to selectively disable roll, make sure you do it symmetrically or you will end up with unwanted yaw/pitch when you try to roll. Engine gimbal will only help roll if they are off axis (i.e. you have more than one engine). SAS will help roll no matter where you place it. Either of these can be sufficient to control roll, but that really depends on the size of the craft.

3) I can talk all day long about flight control gain parameters, but I don't know if k & d are gains or limits in FAR. Sorry. If they are gain and limit parameters, I still doubt you will see much performance improvement. MJ already tailors inputs to not overcontrol (in most circumstances). Gain and limiters are usually implemented to aid in "feel" of a craft while you're flying it, so that it reacts similarly to inputs at different points of flight. MJ is already imposing this on itself.

[Rhomphaia]

Engine gimballing doesn't provide any roll authority, off axis or not (at least in stock, I too have no experience with FAR).

[Claw]

Engine gimballing doesn't provide any roll authority, off axis or not (at least in stock, I too have no experience with FAR).

Oops, my bad. You're right. I was confusing my thoughts with another quad-coupler thread. They can cause roll off axis.

[Luckfish]

Thanks for replying, Claw and Rhomphaia. After thinking a bit and reading your posts, I think I'll be setting the lateral fins to only control pitch, and the ventral/dorsal ones to only control yaw. If my rocket is fairly large, I might give each one a minor amount of deflection control in the roll category also, to assist SAS and any off-axis engine gimballing. It makes sense that the roll axis would be touchy when flying a rocket through an atmosphere, because there isn't much in the way of atmospheric negative feedback to dampen the inputs like what pitch and yaw would feel. The only thing really counteracting roll would be inertia.

With FAR I would tend to get some pilot-induced oscillations (especially in the roll axis) during ascent, whether I flew manual or used MJ. Since FAR thins out the atmosphere to something reasonable (as well as fixes the silly simulation of drag) I can understand how the control surface response would need to be adjusted. My first few flights with FAR didn't go so well until I figured this out. I actually uninstalled FAR for a short bit, but it made everything seem way too easy, and the drag equation and fake gravity turn kept bothering me. I've decided to put my space exploration on hold until I get pretty sharp at designing and flying in the FAR atmospheric model. I had been pretty intimidated by both FAR and Deadly Re-entry at first, mainly due to hearing about other people's frustrations and problems and seeing a few YouTube videos, but going without them now would feel a bit like cheating. I think one of the biggest draws KSP had for me was both the realism (compared to other space games) and the steep learning curve. I know it's not for everyone, but I love it.

[Claw]

Try not to think of it as "cheating" or "not cheating." The beautiful thing of an open ended game like KSP is that you can play how you like. If you get more enjoyment out of using FAR and DRE, then do it!

[Luckfish]

Well, "cheating" as in nerfing down the difficulty so much as to remove the challenge I enjoy. I don't fault anyone for playing however they wish, but I wouldn't enjoy it near as much without FAR & DRE. The modability of KSP is truly awesome, it allows people to tweak it to just the game they want.

[Luckfish]

So I'm trying to assign & tweak control surfaces in the VAB, and either something isn't working right or I don't know what I'm doing. I'm trying to assign the two lateral AV-R8 tail fins to both deflect 10 degrees max for pitch control (upper half of the tweak window) as well as 3 degrees max for roll control (lower half of the tweak window). After I close the tweak window and go back into it it has the pitch max deflection set to 3 degrees. Sometimes it just resets back to 15 degrees. I've tried closing the tweak window by right-clicking somewhere else in the VAB, as well as right-clicking on the opposite tailfin. I've tried tweaking with symmetry both on and off.

Am I not understanding the tweak window or how to manipulate it? I understand that KSP is still in alpha, but if there's some tutorial on tweaking control surfaces out there I would appreciate a link.

Edited February 18, 2014 by Luckfish

[Rhomphaia]

[snip]

Those tweaks are from FAR, stock KSP only lets you tweak which imputs the surfaces react to. Your best bet would be to ask in the FAR thread.

[Luckfish]

Thanks, I'll re-post over there...