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

[Taverius]

The quadratic unitScalar is labeled as an intake bug in that MechJeb code too, and it does look like a bug - probably something like a "speed = (speed + intakeSpeed) *unitScalar" line repeated twice or similar.

Re intakeSpeed, fortunately the required change if you only want to reduce area while keeping the final value the same is actually increasing it from 100: in the above case, to as much as 455.8.

Oh I'm fine with it, I just added a new block to my spreadsheets for this. But its funny just how many such issues there are.

Can you please make an issue on the KSP bugtracker about this?

[DaRocketCat]

What are the installation instructions for FAR? There aren't any here, and your spaceport description simply tells the user to refer to the forum thread for installation instructions.

[bulletrhli]

I don't know if anyone knows about this, but I took the flat structural pieces in 8 times configuration and placed them around my rocket to act as drag. After launching everything went black and I lost all planets. Interesting... If a screenshot is needed, I will repeat this process and get it as I have just restarted KSP.

[ferram4]

@rhoark: That's all behaving as it should; the only time that an airbrake in the location you described would be unaffected would be if the wing in front of it were stalled.

@a.g.: Welp, time to revert all my changes then.

@DaRocketCat: See the modified first post of this thread. You install it just like you would any other mod, copy it into the KSP root directory and merge it with the existing GameData and Ships folders.

@bulletrhli: If you manage to cause that again, could you try to remove all other mods and unnecessary parts from the craft until you get the smallest vehicle that can cause that bug? Having access to a craft file and detailed instructions on how to reproduce the bug would go a long way towards finding and fixing the issue.

As a heads up to everyone, here's what I've gotten done for the next release (no ETA though):

Cargo bays ~90% functional using Firespitter's animation module. A few bugs to work out though.

The huge amounts of drag from some of the new B9 intake parts has been fixed.

Transonic and supersonic wing drag getting tweaked.

Some optimizations in the wing code; should reduce lag a bit.

And a known issue that I've found: don't use a cargo bay as the root part of a ship. The entire ship will be unaffected by aerodynamics. This will be fixed in the next version.

[asmi]

Hi Ferram! I wonder if you've ever considered "FAR'azing" BobCat's Buran? I've just tried flying it, and I even managed to get it to orbit with no issues (just had to be VERY careful with steering), but during reentry it turns read-end first and no RCS nor anything else I could think of helped to turn it around the way it's supposed to fly. If you would be so kind to provide configs for that, I'd be the happiest person on the planet! I really love Buran, the way it flies, its' cargohold is roomy enough to assemble ISS, but having FAR renders it useless as I can't land it

[lyndonguitar]

I am using mechjebs auto-land feature for my Vertical Landing Capsules and it works pretty accurate, however if I use this will Mechjeb still calculate the predicted landing site properly?

[Doomydoom]

@rhoark: That's all behaving as it should; the only time that an airbrake in the location you described would be unaffected would be if the wing in front of it were stalled.

@a.g.: Welp, time to revert all my changes then.

@DaRocketCat: See the modified first post of this thread. You install it just like you would any other mod, copy it into the KSP root directory and merge it with the existing GameData and Ships folders.

@bulletrhli: If you manage to cause that again, could you try to remove all other mods and unnecessary parts from the craft until you get the smallest vehicle that can cause that bug? Having access to a craft file and detailed instructions on how to reproduce the bug would go a long way towards finding and fixing the issue.

As a heads up to everyone, here's what I've gotten done for the next release (no ETA though):

Cargo bays ~90% functional using Firespitter's animation module. A few bugs to work out though.

The huge amounts of drag from some of the new B9 intake parts has been fixed.

Transonic and supersonic wing drag getting tweaked.

Some optimizations in the wing code; should reduce lag a bit.

And a known issue that I've found: don't use a cargo bay as the root part of a ship. The entire ship will be unaffected by aerodynamics. This will be fixed in the next version.

I got that black screen a lot in 0.19, I'll try to find something that breaks it.

edit 1: album of screenshots of it happening with console open http://imgur.com/a/xI20Y

It seems to happen when loading one craft, launching, ending the flight, and loading a craft and launching it. But not every time. It took six tries reloading and launching the same craft to get it to happen. I got it again just now by launching a craft, restarting it ten times, then going back in to the hangar, loading the same craft again, and launching it. I was using a complex craft with a few mod parts on it, but no other plugin mods that I can see are active on it. This was all done in .19. Pic of the craft: http://i.imgur.com/PdpmfOW.jpg

Edited June 8, 2013 by Doomydoom

[foamyesque]

So uh

Am I crazy, or should this not have worked?

[ferram4]

@asmi: I can look at Bobcat's Buran, but I don't know if I'll actually create the configs for it... the fact is that since I don't use it there isn't much incentive on my end to do the work for it. Honestly, if you want to make them up you can find the relevant distances and angles yourself; the documentation for each value should be in the FAR readme.

@lyndonguitar: If your asking "will MechJeb properly account for how aerodynamics will affect my landing position?" the answer is no. MechJeb isn't able to interface with FAR for that and even if it was it would be quite possible to land in the wrong place as a result of the drag and lift changing due to the angle of attack of the lander during the descent.

@foamyesque: Those are all mainsails, correct? I think you've proven that any problem, even poor aerodynamics, can be overcome with enough thrust. I'll bet you could cut the dV needed down by ~500 m/s or so if you used a more aerodynamic vehicle. Now, realistically, that should probably fly apart, but I can't control the stiffness of the joints between parts independently of their attach strength; if they could this probably wouldn't happen. I suppose I can look into increasing the drag of the command modules when they aren't in a reentry configuration.

[foamyesque]

@foamyesque: Those are all mainsails, correct? I think you've proven that any problem, even poor aerodynamics, can be overcome with enough thrust. I'll bet you could cut the dV needed down by ~500 m/s or so if you used a more aerodynamic vehicle. Now, realistically, that should probably fly apart, but I can't control the stiffness of the joints between parts independently of their attach strength; if they could this probably wouldn't happen. I suppose I can look into increasing the drag of the command modules when they aren't in a reentry configuration.

Skippers. Liftoff TWR is 1.5, flies just fine; I rejigged the ascent path and got 3.4km/s dV to 100x100. I don't think the problem is the command pods; the issue is that each stack is being treated independently, as if air could flow freely around them. Since each individual stack is aerodynamic, so is the overall rocket.

[localSol]

[...about the closed wing not flying apart..] ..Considering the strength of KSP jet engines and the amazing failure strengths of parts, none of this is really all that wrong.

Thanks very much for getting back to me on this! I heard people mention the strength of certain parts..

..Maybe different aircraft materials with different mass, cost, heat resistance and strength...

To anyone new to this like me, I found out there is a large increase in drag followed by a decrease, called wave drag, between trans-sonic and supersonic speeds depending on something called critical mach for a particular aircraft. FAR actually models this and accounts for some of the instability I notice when going through those speeds. You can get it graphed for you in the static analysis menu before flying also.

other stuff :

FAR includes a cruise-missile-like jet (craft file). Fun to try to fly it. I needed to use trim/joystick.

I switched to Taverio's stock replacer engines, B9 mod has great looking ones for various speeds and purposes too.

Edited June 9, 2013 by localSol

[ferram4]

@foamyesque: While the aerodynamics of each stack are being determined independently (actually, the aerodynamics of each part are being determined independently) is part of the problem, at supersonic speeds (where most of the poor aerodynamics would be noticed) a half-decent approximation can be figured by handling the pressure difference acting over the front and back cross-sectional area of the vehicle and adding skin friction drag afterwards. Currently FAR doesn't actually handle how the cross-sectional area of vehicles change (it's difficult to determine for any given part and then there's the problem of the change between parts) which is where the majority of this problem comes from. Then I suppose a 3-D interference effect could be approximated on top of that if needed.

@localSol: I wasn't specifically talking about the strength of p-wings, I was talking about the strength of most parts in KSP in general; they all are a little too strong but this is necessary based on how the joints between parts behave.

[Weatherman159]

New (potential) FAR user here, I hope you can help me somewhat. I've always stayed away from FAR, mainly because of forward compatibility (me thinking: what if a different aerodynamics system is implemented by Squad and I have to remake all my planes again), but since 0.20.2 seems to be a long-term project for me, I thought I'd give it a try. To my astonishment, my most used planes required only small modifications to make them FAR capable, some of them even behave better using it. I then tried to launch a rocket. Oh god. Help.

This here is a fairly standard design I use for transporting rovers to the Mun. Rover size may vary a bit, but I always use this configuration (a rover on both sides) to balance the CoM.

Difference from stock is I added the nosecones. It does go up quite nicely in a straight line, but as soon as I try to do a gravity turn, it just flips end over end and heads toward the ground. I do my gravity turns at around 10 km high, going for about a 45° angle. And this is a small one. How I would go about delivering my Kethane mining base, I have no idea.

[daveboy2000]

New (potential) FAR user here, I hope you can help me somewhat. I've always stayed away from FAR, mainly because of forward compatibility (me thinking: what if a different aerodynamics system is implemented by Squad and I have to remake all my planes again), but since 0.20.2 seems to be a long-term project for me, I thought I'd give it a try. To my astonishment, my most used planes required only small modifications to make them FAR capable, some of them even behave better using it. I then tried to launch a rocket. Oh god. Help.

This here is a fairly standard design I use for transporting rovers to the Mun. Rover size may vary a bit, but I always use this configuration (a rover on both sides) to balance the CoM.

Difference from stock is I added the nosecones. It does go up quite nicely in a straight line, but as soon as I try to do a gravity turn, it just flips end over end and heads toward the ground. I do my gravity turns at around 10 km high, going for about a 45° angle. And this is a small one. How I would go about delivering my Kethane mining base, I have no idea.

Use fairings. Did you ever see a rocket go up with rovers sticking out? It makes the entire thing unstable.

I recommend either making a fairing in The Fairing Factory web-app, or to get KW rocketry and use their standarized fairings.

[ferram4]

@Weatherman159: I'd go with what daveboy2000 says on the fairings, but here are some other things you might want to think about:

• You might want to look into what a gravity turn actually entails; it doesn't mean "go straight up and then pitch over 45 degrees all at once" like most people here seem to think. Taking the gentler, smoother pitch schedule of a gravity turn will make controlling your rocket easier.
  
• Try to keep your rocket pointed within 5 degrees of the surface prograde marker (which is slightly more than the size of the circle on the marker). Large angles of attack can cause a loss of control in many designs.
  
• Reduce your TWR. You could probably replace every single one of those Mainsails with Skippers and it would still fly. Large TWRs tend to lead to overspeeding in the lower atmosphere, which can lead to aerodynamic forces overpowering control authority.
  
• Use more serial staging and less parallel staging; you don't need nearly as much dV with FAR installed, and longer rockets tend to be more aerodynamically stable. Further, as the mass drains out of the first stage's tanks, the CoM moves forward, making the rocket even more stable.
  
• Try to make your first stage last until you start getting into the upper atmosphere; early staging events can cause the launch vehicle's dynamics to change suddenly at an inopportune time.
  
• Add fins to the bottom if you need an extra little bit of control; their effectiveness will drop near Mach 1 due to transonic effects, but they can help on some troublesome designs.
  
• Don't use ASAS in the atmosphere; it can cause flexing oscillations that can cause a loss of control due to uneven aerodynamic effects.
  

I'd say to start with rockets that aren't designed to deliver strange payloads and instead focus on smaller, more aerodynamic payloads. Think back to when you first started playing KSP and go through those motions again.

This is a rocket that I use to send crew to a space station I have in Munar orbit; I'm sure it could easily be modified to deliver a payload to the Mun. You'll want to start pitching over when it's going about 25 m/s; just start to tilt it over the slightest bit and just keep it pointed forward. Gravity will turn it most of the way.

[rhoark]

ASAS in the atmosphere is fine as long as you don't have excessive pitch authority (ie, multiple mainsails with gimbals on.)

Tail fins are a big help with rockets flipping, but below 40k you should still take the turn 10 degrees at a time.

[Weatherman159]

Thanks for all the tips. Looks like I need to learn to fly all over again....and also build different types of ships .

Although I do see a problem in the future. Even if I get proper flying skills, I'd say there's no way to get something like this into orbit.

(I don't have a picture of the launch vehicle, but it was basically four asparagus stacks strapped under it)

Gah this is twisting my mind. Tall rockets with low dV, that's like...the opposite of what one wants in the stock game. This is gonna take some time to adjust to...

Gah n.2 the fairings, the bloody fairings...I don't want to install KW rocketry, but the biggest one I can make in fairingfactory is just too small.

Edited June 9, 2013 by Weatherman159

[DaRocketCat]

Where do I put the "Source" folder? I apologize for my ignorance, I have no idea how the new system works .

[camlost]

Gah this is twisting my mind. Tall rockets with low dV, that's like...the opposite of what one wants in the stock game. This is gonna take some time to adjust to...

I think it's the notorious aerodynamics of stock game that forced people to create asparagus abominations that contradicts the reality simulating which is exactly the purpose of all sandbox games

[Weatherman159]

Where do I put the "Source" folder? I apologize for my ignorance, I have no idea how the new system works .

You don't put the Source folder anywhere. It's just the source code for the mod.

[foamyesque]

Gah n.2 the fairings, the bloody fairings...I don't want to install KW rocketry, but the biggest one I can make in fairingfactory is just too small.

There's a zoom out button. You can make absolutely enormous fairings in FF with it.

(I missed it at first too)

As for the rovers, I've been doing some research in that direction. I've found that one solution is to build a spar out of the cubic octagonal trusses with a not-rockomax node in the middle. Strap the rovers to the node, and your pod to the top of the spar. Strut. Enclose in fairing.

EDIT:

For example:

[Weatherman159]

There's a zoom out button.

Well I noticed that, but when I clicked it nothing happened so I assumed it was broken. Turns out one has to hold it down. Cheers.

[localSol]

I'd go with what daveboy2000 says on the fairings, but here are some other things you might want to think about:

• You might want to look into what a gravity turn actually entails; it doesn't mean "go straight up and then pitch over 45 degrees all at once" like most people here seem to think. Taking the gentler, smoother pitch schedule of a gravity turn will make controlling your rocket easier.
  
• Try to keep your rocket pointed within 5 degrees of the surface prograde marker (which is slightly more than the size of the circle on the marker). Large angles of attack can cause a loss of control in many designs.
  
• Reduce your TWR. You could probably replace every single one of those Mainsails with Skippers and it would still fly. Large TWRs tend to lead to overspeeding in the lower atmosphere, which can lead to aerodynamic forces overpowering control authority.
  
• Use more serial staging and less parallel staging; you don't need nearly as much dV with FAR installed, and longer rockets tend to be more aerodynamically stable. Further, as the mass drains out of the first stage's tanks, the CoM moves forward, making the rocket even more stable.
  
• Try to make your first stage last until you start getting into the upper atmosphere; early staging events can cause the launch vehicle's dynamics to change suddenly at an inopportune time.
  
• Add fins to the bottom if you need an extra little bit of control; their effectiveness will drop near Mach 1 due to transonic effects, but they can help on some troublesome designs.
  
• Don't use ASAS in the atmosphere; it can cause flexing oscillations that can cause a loss of control due to uneven aerodynamic effects.
  

These are really good, especially for new people. Maybe add these to the readme and first post?

But if people out there are still having trouble, I wouldn't sweat it too much, you can take your time, keep trying and get creative with FAR. You might be surprised what you can launch even with lots of forces acting against you, especially wobble-physics with massive loads. I launched a pretty big DEMV rover packed with all kinds of surface attached gear with no fairing (and plenty of people have, because there weren't as many options at the time) on a maybe needlessly big rocket , and was able to get it into out of the atmosphere by being very gentle. I try to generally keep the G-force meter below two G's and for extra questionable loads, do several test launches first to see where the stress will cause problems. For that rover payload I had to keep it near 1.3-1.5 G's for a lot of the pre-orbit.

In FAR I learned to be careful of near mach 1 (and trans/supersonic speeds). The idea is that there are tremendously powerful shockwaves/vibrations going around and through your vehicle through those speeds.

Avionics packages can be gentler than ASAS if you need help with unstable launches. They don't guarantee stability but If you notice what ASAS does with the controls, it is rarely gentle, so that's why you might notice quick direction change followed immediately by tearing apart or loss of control. There are lots avionics part types to choose from between stock, B9 and Firespitter. Depending on the vehicle sometimes it's best to use none of them or FAR flight assistance, and I think those options are fun if confusing at first. It's a good game balance in my opinion, until there's a better balance anyway..

Something new I have been tinkering with are the CMG gyros by a modder (esentially they give you the same forces of a command pod and use a little energy to operate), and somewhere I think there was a mod for finer RCS control.. Gyros can help sometimes with gravity turns, though sometimes just the engine gimbal is plenty.

Edit : Oh but to answer weatherman :

I can't really tell as much as I would like from the picture, but especially with large fuel tanks, as they empty the shifting center of mass can cause the (center of) thrust (as it changes also due to jettisoned engines) to react weirdly to the changed CoM. I guess you have different length and capacity of tanks and if they all burn at once, some will run out sooner than others, leaving some heavy fuel in some tanks and some still heavy but lighter empty fuel tanks and engines (until they are jettisoned - but that may not necessarily be better, because the center of mass shifts again, to where who knows!). The picture makes it look like there are only tanks on about half of the side of the whole rocket, kudos to you if you can launch it but i would think it would be very unstable even though the gimballing engines can help, they are also very heavy , and heavier stuff is also less effected by the torque in the command pod that you'd use to help manuever compared to lighter stuff. The thrust and gimbal from mainsails is very strong though so you'd have to turn them off to see only what the command pod is able to do, which at ground level can be hard to determine since you lose so much fuel so fast with mainsails.

Edited June 10, 2013 by localSol

[Queue]

hey guys i love the mod but i have a few questions. Im sorry if i sound like a complete idiot and i have googled things questions but alas i can find no answer.

first question. when i am programming the control surfaces to be pitch yaw roll etc how do the numbers effect this (i think the default is 15) i have read the read me in the game and really do not understand it if you could please explain how a 15 may be different from a 5 or 50 would be awesome (pretend i am a complete moron because i might be

second i know ASAS does not work well it seems with far (which is fine i have used it as a crutch way to long but since i do not have a joystick i find keeping a plane stable difficult i know FAr has a built in system but like the above question i do not fully understand how the numbers effect the flight like wing stabilization .05 vs .10 etc if it could be explained i would really appreciate that.

thank you so much for the mod and the help i cant get enough of my SSTO space planes

[ferram4]

@Queue: The numbers are the maximum deflection of the surface in degrees; basically, putting 5 in there will cause it to deflect up to 5 degrees while 50 will cause it to deflect up to 50 degrees. Larger numbers will result in more lift and drag from the control surface, up until it stalls, at which point lift will drop off but drag will continue to increase. Depending on how your plane is designed, a control surface stalling can result in a slight reduction in control authority at best and a complete loss of control at worst.

The numbers on the control systems controls the gain of the control system; the gain determines how quickly the system responds to disturbances. Basically, a gain of 0.1 for the Wing Leveler will result in the system responding faster than a gain of 0.05. However very large gains can cause the vehicle to oscillate; for an abnormally large Wing Leveler gain, the plane might continually roll left, right, and back again, never actually leveling out completely; if the gain is large enough if might even roll all the way over. Very large gains are the reason that ASAS can cause serious problems with FAR.

@foamyesque: That's a very cool way of packing that payload. Everyone who can't figure out how to launch their rovers, space station sections, etc. learn from this guy.