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

[Northstar1989]

@Northstar: In practice, in real life there is little difference between the thickness of the wing and how much lift you get from it (assuming the same angle of attack and camber) until you get near stall. At that point, thinner wings will have much more sudden stalls than thicker wings (the difference being leading edge versus trailing edge stall). In general, anything with a thickness greater than ~12% chord will have its peak lift dropped severely by early onset of trialing edge stall while anything thinner than ~12% chord will have its lift dropped severely by leading edge stall at much lower AoAs.

FAR doesn't bother with this, since it's kinf od complicated to figure out and would make procedural wings a de facto required mod. All wings are set up as 5% thick, with whatever shape is ideal for either subsonic or supersonic flight.

Maybe there's something I'm missing, but aren't you constantly on the edge of stall when you are flying near altitude ceiling? That is, the thinner the air, the lower speed the aircraft, and the heavier the payload, the closer an aircraft operates to stall. Thus, thicker wings would allow flight in more marginal conditions, to higher altitudes in subsonic flight (which is what I'm interested in- especially with thick wings and relatively weak propulsion in the form of propellers, Flying Fortress style), etc.

Also, thicker wings allow you to build in greater camber, and according to NASA:

"In general, low to medium speed airplanes have airfoils with more thickness and camber. "

http://quest.nasa.gov/aero/planetary/atmospheric/aerodynamiclift.html

There is a lot of data out there on "The Effects of Airfoil Thickness on Lift Coefficient". See the charts in this one, for instance:

http://itlims.meil.pw.edu.pl/zsis/pomoce/BIPOL/BIPOL_1_handout_8A.pdf

Lift Coefficient hits a peak at a thickness of around 12-16% chord length (t/c = 12-16%)

But, see the next page- it also has an effect on drag coefficient- though surprisingly some of the best Gliding Ratios were obtained for the thickest airfoils in this study...

So, I wouldn't call the effect of thickness of lift trivial. Less important than wing area or camber, perhaps, but definitely important.

It may be revealing to reflect on the fact that one of the reasons that biplane/triplane designs were abandoned is because advances in materials enabled the construction of longer, thicker, wider wings. Thickness was a consideration as well as size- and of course wings that are thicker at the base can support a greater maximum span if they gradually taper.

I find the perspective of history and the evolution of aircraft particularly interesting. Take, for instance, this quote from a discussion )on Reddit) of the replacement of biplane/triplane designs with thicker-winged planes after WWI:

"during WWI wings were thought to be most efficient when thin. This was logical since thin wings would simply slice through the air. It was not until later that it was discovered that thicker wings could actually create more lift with less drag by way of the Venturi affect (thicker at the leading edge, thinner at trailing edge). This negated the advantages of the second wing, and inadvertently created new uses for the wings such as fuel storage. "

http://www.reddit.com/r/AskHistorians/comments/1k5fjn/why_were_biplanes_the_preferred_method_of_early/

I've heard a lot of similar lines in historical documentaries on WWI-era aircraft, as well as on the History Channel (back before it went to Hell), etc.

IMHO, Procedural Dynamics (aka Procedural Wings) SHOULD be a huge advantage and practically a requirement for any serious plane-builder in KSP who wants to build craft with low wing-load and high Glide Ratio. Of course, the utility of such craft is questionable- since KSP is a game about space exploration, and the best spaceplanes don't have the largest or thickest wings- as that's a LOT of mass to push to orbit... So I wouldn't call it unfair to non-Pwings players to include thickness in calculations of lift coefficient.

As it currently stands, anyone using B9 Aerospace with FAR is unfairly penalized, as they have extra mass to haul around, and greater drag, but none of the increased lift that thicker wings generate (wihin certain limits- as I said the best lift coefficients were measured at t/c of 12-16%, with declining lift at greater thickness)

Regards,

Northstar

Edited July 21, 2014 by Northstar1989

[ferram4]

Yes, I am aware of how efficient thick wing sections can be for subsonic flight well below the critical Mach number. However, the data you're showing is for airfoils alone, which means that it does not include the effects of finite wing sections, which will dominate drag effects and hide virtually all of the benefits from thickness effects. Further, FAR does not simulate camber within a single wing part at all due to the fact that not only is camber useless for supersonic lift, but it also raises the question of determining which way the wing should be cambered, which is not necessarily trivial.

Overall, FAR already allows you to reach the lift and drag of thick subsonic wings at low Mach number flight, because those wings are closest to the optimum possible and FAR's analysis that ignores thickness effects for subsonic flight already brings things close to the optimum.

[Cerebrate]

@ferram4 , or anyone else with a good feel for FAR's internals:

Is there any relatively simple/supported/possible way to pull the numbers as seen in FAR's flight data window out for use in another mod for the current ship in flight?

(Reason being as here: http://forum.kerbalspaceprogram.com/threads/84409-0-24-0-NanoGauges-tiny-ana1og-gauges-for-kerbalnauts-0-5-9?p=1287094&viewfull=1#post1287094 - I want to be able to keep an eye on Q (in particular) during launch and reentry without having to have the flight data window open all the time on my increasingly crowded screen...)

Thanks muchly,

-c

[ferram4]

No, as that field is currently private in FARControlSys. I can make it public for the next update though.

[TheFireRodan]

Procedural Fairings work properly, right? In launch tests no Proc fairings flew faster than with... FAR was active, as default nosecones beat no fairings.

So is [Procedural Fairings< No aerodynamics < Default aerodynamics] Correct?

Edited July 21, 2014 by TheFireRodan

[Woodstar]

Mines not showing in toolbar and which dll's do i delete for mod statistics, This one ModStatistics-1.0.3.dll and the graph? or do i leave the graph?

[ferram4]

@TheFireRodan: They will, but only if the shape they are shielding is draggier than the equivalent fairing shape. Basically, going and place fairings over a nice, aerodynamic shape just makes things worse because you're increasing the surface area drastically. Only add fairings for payloads that require them.

@Woodstar: FAR appears on the new stock AppLauncher by default; you will need to switch an option in the debug menu in the Space Center to switch back to the Toolbar. As for ModStatistics, if you delete the ModStats dll that will remove it from FAR, though you would be much better served by just setting disabled=true in the config ModStats directory, which would be permanent instead of the temporary fix you're going for.

If you delete ferramGraph.dll FAR will not run. Do not delete that dll, it has nothing to do with ModStats and has been in FAR since v0.4, many, many KSP versions ago.

[Woodstar]

Thank you will do

[NathanKell]

Proc fairings are not so much about decreasing drag as about making your payloads not screw too much with the aerodynamic center and make your rocket unstable.

[guitarxe]

Hey guys, I'm new to KSP and wanted to give this mod a try. I'm not much into planes right now, focusing more on my rocket designs. I was wondering, what kind of impact am I going to see to rockets? For example, I use the KER mod to tell me how much dV my rocket has and how much TWR each stage has to figure out if I have enough to make it into orbit. So I judge that if my rocket's stages are all reasonably above TWR or 1, and in total there is more than 4,500 m/s of dV, then I should be able to make it into orbit.

Now if I were to install this mod and launch the same rocket design, would I find that it's now impossible because my rocket is not very aerodynamic? Meaning that my usual way of building rockets where I only need to take care of enough TWR and dV would now be not enough?

Also about procedural fairings - that should make my rockets more aerodynamic, right? So if I get my clunky and ugly payloads into a nice smooth cone shape with the fairings it should improve the performance of my rocket?

[ferram4]

You're correct, though the aerodynamic considerations are less about losing dV to drag (that's not a big deal really, total drag losses are about ~100 m/s, only 300 m/s for the worst vehicles, compared to the stock 1000 m/s) and more about keeping the rocket stable (big draggy stuff at front with lots of mass at bottom == launching a rocket-powered arrow backwards). Procedural fairings tend to help with that, since most of the payloads you launch aren't that aerodynamic really.

You're looking at a dV of ~3200 - 3500 to orbit and an initial TWR of somewhere between 1.2 and 1.6. Any higher than that results in stability problems normally. Also, you need to start the gravity turn slow and early (not 10 km straight up and then yank over, that's not a gravity turn) in order to keep the rocket from becoming unstable. Keeping at terminal velocity isn't a concern anymore, since you'll lose control long before you get there.

That's about it. Make your rockets look like rockets and they should fly like rockets.

[guitarxe]

That's about it. Make your rockets look like rockets and they should fly like rockets.

Perfect! That's exactly why I wanted to try this mod

[Northstar1989]

Yes, I am aware of how efficient thick wing sections can be for subsonic flight well below the critical Mach number. However, the data you're showing is for airfoils alone, which means that it does not include the effects of finite wing sections, which will dominate drag effects and hide virtually all of the benefits from thickness effects. Further, FAR does not simulate camber within a single wing part at all due to the fact that not only is camber useless for supersonic lift, but it also raises the question of determining which way the wing should be cambered, which is not necessarily trivial.

Overall, FAR already allows you to reach the lift and drag of thick subsonic wings at low Mach number flight, because those wings are closest to the optimum possible and FAR's analysis that ignores thickness effects for subsonic flight already brings things close to the optimum.

Like much of the more advanced stuff to do with aerodynamics, your response just left me scratching my head. I may be a genius (that's a tested fact, according to IQ tests as a child, not my opinion), but I'm a biologist- not an aerospace engineer, and it's 2 AM here...

What exactly are you saying? That thick wings actually have some benefit for low subsonic flight in FAR? Because as far as I was aware, they don't- and are strictly inferior to thin wings at all speeds in FAR, which was precisely my issue (since I want to build aircraft with VERY LOW cruising speeds- due to their reliance on electric propellers, which have terrible velocity curves- and have very high altitude ceilings, so they can also fly on Duna- and large, thick wings is the main way I could see to realistically accomplish that, especially since gravity is lower on Duna, thus penalizing the additional mass of thick wings less...)

Regards,

Northstar

P.S. I'm not complaining. I love this mod already. How else could I do something like THIS, all-stock (except MechJeb2 and FAR), by exploiting body-lift?

(That's a rocket of three 1.25 meter fuselages in a row, by the way. Note that the ASAS is only set to a pitch angle of 84 to cause it to pull up at maximum-torque: the rocket had very little SAS torque, and the aerodynamics kept trying to shove the nose into the ground...)

And yeah, that easily made it to orbit, with all remaining stages attached, and fuel to spare...

(The increase in Delta-V is due to shutting off the most inefficient engines- which were attached mainly to give those boosters fly-back recovery capabilities, as this is 0.24, *IF* they can survive the aerodynamic loads of re-entry...)

Edited July 21, 2014 by Northstar1989

[Cerebrate]

No, as that field is currently private in FARControlSys. I can make it public for the next update though.

That'd be very handy, thanks muchly!

-c

[ferram4]

@Northstar1989: Thick wings have no benefit in FAR compared to thin wings at subsonic speeds because both wing types are simulated as the same thing, with the same stall angle, both at the very limits of what can be achieved from moderately thick wings. They already stall with the lift coefficient of moderately thick wings, (~1.6), and so making thickness affect performance won't make thick wings any better, it will simply make thin wings worse.

And besides that, wing thickness doesn't matter in FAR anyway. They're all the same thickness as far as the code is concerned.

For your Duna plane, you probably want to look more into high AR wings instead of focusing on thickness, since that will provide much better L/D performance than what you're currently seeking.

[Grunf911]

You're correct, though the aerodynamic considerations are less about losing dV to drag (that's not a big deal really, total drag losses are about ~100 m/s, only 300 m/s for the worst vehicles, compared to the stock 1000 m/s) and more about keeping the rocket stable (big draggy stuff at front with lots of mass at bottom == launching a rocket-powered arrow backwards). Procedural fairings tend to help with that, since most of the payloads you launch aren't that aerodynamic really.

You're looking at a dV of ~3200 - 3500 to orbit and an initial TWR of somewhere between 1.2 and 1.6. Any higher than that results in stability problems normally. Also, you need to start the gravity turn slow and early (not 10 km straight up and then yank over, that's not a gravity turn) in order to keep the rocket from becoming unstable. Keeping at terminal velocity isn't a concern anymore, since you'll lose control long before you get there.

That's about it. Make your rockets look like rockets and they should fly like rockets.

Hi Ferram,

I have experienced an interesting behavior. Not sure if I can call it a bug or not. Also I use several mods, so I am not 100% sure it is the FAR issue, although seems like it. Here goes.

I have built simple rocket (2 stage) with each stage consisting of FL-T400 FT and LV - 45's. I am in the middle of the ascent, passing some 350m/s when all of the sudden, my ship flies off screen and the camera stays (as if something exploded, or that camera stayed with an invisible control part).

The funny thing is i can still rotate camera, KER still shows data, and even more strange I CAN STILL CONTROL the ship. Throttle, Staging, pitch and everything works, just that camera does not follow. I fooled a bit with it, and I managed to cut power, open a parachute (complete a contract for testing parachute), and then SAW the craft again on the way down, with opened parachutes, and it splashed OK, however I could not recover it.

Mods i use (am at work so i do not know them by heart):

- Kerbal Alarm Clock

- FAR

- KER

- TAC Life Support (beta pre3)

- RT2 1.4.0

- Better Atmospheres (everything except custom asteroids)

- Kethane

- RealChutes

- DMagic Orbital science

- Universal Storage

.. probably few more, but all of them claiming to be 0.24 compatible (except RT2)

I am posting here since I suspect FAR due to approx 350 m/s trigger in LOWER atmo (that part was consistent in bug)

I will though when i get home try to disable FAR and see if it occurs again

[DaMichel]

When i said that there is a strange interaction with Improved Chase Cam. This is not true!

Granted i have a few other mods but none of which mess with the camera though. And still it happens that the air speed display reverts to surface mode and the FARControlSystem.activeControlSystem apparently becomes null exactly during the time where i rotate the camera in flight. I mean really moving the camera with right mouse button down.

@Cerebrate & Ferram: It is already possible to obtain data from FAR. It does not matter what is private or public. See https://github.com/DaMichel/KerbalFlightData/blob/master/KerbalFlightData.cs line 163

[Volt]

Quick question - what are people's Mun lander solutions with FAR installed? I've been using it for aircraft saves but I haven't yet confronted the challenge of launching things to space with the stability changes, and I'm wondering how to do it. I assume some fairings will be involved, but even then, I'm used to making wide-based, single stage return craft which are the widest part of the stack I launch them on. My landers wouldn't even fit in fairings!

[BudgetHedgehog]

Quick question - what are people's Mun lander solutions with FAR installed?

Personally, I go for a lander can with either a fuel tank or ScJr under it and radially attach 4 x 0.625m fuel tanks with the landing legs on them. Nicely toes the line between wide stance and slim width. If there's a ScJr, I put a flat tank underneath. The 909 gives me enough thrust and dV to do a descent burn and return to Kerbin. That might change with costs now though.

As I didn't have the ScJr at that time, there was a fuel tank so I didn't need the radial ones. I replaced them with IR pistons which worked nicely. Decent thrust, decent dV and fits inside a 2m KW fairing.

Edited July 21, 2014 by ObsessedWithKSP

[Tsuki]

I know it was discussed before but it sill one of my biggest problems with FAR is spot landing. I did get fairly good results with practicing and repeating specific reentry procedures and I find it very satisfactory to land un-powered craft on the landing strip by only doing one single reentry burn in orbit. But... it would be great to have some computational assist

I did some thinking (however I am very inexperienced in the actual math involved) and wouldn't it be possible to calculate a landing spot/elipse using current AoA and the crafts drag/lift coefficients, reentry angle and speed. The calculation should update often, to display the landing location as it would be if all those variables stayed the same as they were at the moment of the calculation. It would help a great deal in knowing how are your inputs changing the spot of landing. Maybe such calculations are to expensive to do every frame, but maybe a lower update cycle would provide just enough time to calculate it and still being useful.

EDIT: for more clearance, what i am suggesting is some kind of: where will an unrpowered glider land if it keeps the current AoA if it is less than stall AoA. for angles above stall aoa, we could consider it an invalid solution, or just simplify the prediction to use some kind of average drag the craft is producing at the moment. So some kind of glide slope indication.

Edited July 21, 2014 by Tsuki

[Grunf911]

Quick question - what are people's Mun lander solutions with FAR installed?

I use Infernal robotics pistons to extend landing legs. Works like a charm. Very wide base, yet not too wide on launch. Below is the pic of my lander.

So the idea is 4 pistons in the small tank at the bottom, extending a small girder with 2 pairs of legs on each piston.

Edited July 21, 2014 by Grunf911
clarification

[Tsuki]

Or hide the whole lander inside the fairing like they did it RL:

[RoboRay]

Quick question - what are people's Mun lander solutions with FAR installed? I've been using it for aircraft saves but I haven't yet confronted the challenge of launching things to space with the stability changes, and I'm wondering how to do it. I assume some fairings will be involved, but even then, I'm used to making wide-based, single stage return craft which are the widest part of the stack I launch them on. My landers wouldn't even fit in fairings!

Here's my latest mission, my first Mun landing in 0.24, showing the lander launched below the command pod, covered in fairings.

Javascript is disabled. View full album

If you use Procedural Fairings, you can make them as wide as they need to be. Note that making them huge is going to impact the aerodynamics of the rocket, however.

[elfindreams]

Hi Ferram,

I have experienced an interesting behavior. Not sure if I can call it a bug or not. Also I use several mods, so I am not 100% sure it is the FAR issue, although seems like it. Here goes.

I have built simple rocket (2 stage) with each stage consisting of FL-T400 FT and LV - 45's. I am in the middle of the ascent, passing some 350m/s when all of the sudden, my ship flies off screen and the camera stays (as if something exploded, or that camera stayed with an invisible control part).

The funny thing is i can still rotate camera, KER still shows data, and even more strange I CAN STILL CONTROL the ship. Throttle, Staging, pitch and everything works, just that camera does not follow. I fooled a bit with it, and I managed to cut power, open a parachute (complete a contract for testing parachute), and then SAW the craft again on the way down, with opened parachutes, and it splashed OK, however I could not recover it.

Mods i use (am at work so i do not know them by heart):

- Kerbal Alarm Clock

- FAR

- KER

- TAC Life Support (beta pre3)

- RT2 1.4.0

- Better Atmospheres (everything except custom asteroids)

- Kethane

- RealChutes

- DMagic Orbital science

- Universal Storage

.. probably few more, but all of them claiming to be 0.24 compatible (except RT2)

I am posting here since I suspect FAR due to approx 350 m/s trigger in LOWER atmo (that part was consistent in bug)

I will though when i get home try to disable FAR and see if it occurs again

Actually that has occurred for me ever since I installed RT2. You can actually zoom back in (it somehow resets the zoom to really far away) using the mouse wheel. With all the other issues I have been tracking down and posting about, I just haven't gotten around to looking at that one too deeply since all it does is make me re-zoom back in.

You even sometimes get it on the launch pad

[elfindreams]

How do I build Rockets using this mod? I can always get my rockets to pitch but sometimes they don't turn with gravity, sometimes they turn the wrong way, sometimes they do turn with gravity but when I jettison the first stage the rocket will just start spining out of controll. What is happening? I've looked for guids on how to build rockets with far but I can't find anything, all the guides I've found are just for planes. I found something to do with adding wings to my rockets, but I found they make no difference and I don't want to add wings on my rockets. I use KW for rocket parts andd engines, I don't know if that makes any difference.

Another tip that has helped me immensely: Forget what you think you know about where center of mass should be.

When dealing with FAR you want your CoM as much higher from your center of "lift" (or is it better to refer to it as center of aerodynamic pressure?) as you can make it. Usually with light payloads this means tacking on control surfaces as far down the rocket as you can go. The problem is once you lose that stage, the CoL rises back to where it was before and if that is close (or above) the CoM your rockets will become a lot less stable.