Thinking about making the switch to FAR.

[WhiteKnuckle]

Rapiers are the backbone of my spaceplane fleet. Very few of my designs don't use them. To me, turbojets are supplementary engines, I'll use them to give me a little extra power between sea level and 25k.

One way I do like to use turbojets is if I'm out of easy engine mounting locations and want to add engines above or below CoM. If you try this with Rapiers they'll flip you out when you switch to rocket mode. But with a TJ you're only using them at altitudes where you should have enough control authority to counteract the offset thrust.

[Crzyrndm]

It's either a massive urban myth, or they're very slightly more effective.

10% extra maximum thrust, 6.25% higher "0 thrust" speed. If the air/fuel ratio is different it's a stock thing.

Rapiers have the a definite advantage on large planes IMO. Once you start needing a rocket mode with a 200+kN of thrust, positioning the rockets during building and dragging enough up through the atmosphere is a right PITA. On smaller planes where you can just add a few small radials and still get a decent TWR, turbo's are obviously superior, but having 10 rapiers on your cargo hauler is far easier than 8 turbo's and a half dozen (+) rockets (and probably more efficient).

Once in space I'll still leave it to w/e rocket engine fits the bill, but rapiers make orbital ejection a whole lot easier.

[UnusualAttitude]

RAPIERS certainly have many advantages. Punching through the upper atmosphere with a good TWR is one of the keys to getting a spaceplane to orbit efficiently. RAPIERS, by being both jet and rocket, will give you a superior TWR over a jet-rocket combo.

This scales up to larger vehicles in theory, but I found that one of the challenges of heavy multi-RAPIER designs is their overheating when fitted as clusters. There are ways round this (with struts, or putting them on smaller fuel tanks), but this may destroy the looks of your aircraft and make the part-count soar.

I eventually gave up on them for my Mk3 heavy cargo, and settled for 10 turbojets and a KR-2L to get an orange tank to orbit. But if your computer can handle it, then you should definitely consider the RAPIER for many spaceplane uses.

[capi3101]

Hey all. Probably something dumb to report today. Last night I decided to do a "static firing" test of an individual Turbojet and an individual RAPIER engine on the runway just to get some data on those. With FAR in effect, a single Turbojet puts out 88 kN on the runway while a RAPIER does 80 kN, if anyone's curious. Now, what I did with that data was calculate a mass-per-engine guideline based on eddiew's earlier suggestion of a 0.75 TWR on the Runway. Given those thrust values and TWR target, a guideline of 11.75 tonnes per Turobjet or 10.75 tonnes per RAPIER is suggested. The actual values are somewhat larger than that but they aren't cut and dry; these values give a TWR ever so slightly greater than 0.75.

Anyways, last night I designed a new plane to test the guideline - turned out to be a 17 tonne craft with dual Turbojets. 100 m/s takeoff. Hit Mach 1 at 7k and Mach 2 around 17k, and topped out at Mach 4.2 at 30k - so I take it that was actually too much thrust? Perhaps a better thing to do is to use the guideline to figure out how many engines you need but then limit the thrust based on the plane's final mass. That'd take some math but math rarely daunts me.

Still not sure about a rule of thumb for an Intake to Engine ratio; does 0.03 intake area per engine still apply or not? Last nights engines were built on Nacelles, each one with a Shock Cone on the front end and six Structural Intakes on the sides.

Also had a problem with Procedural Wings last night - damn SPH wouldn't let me attach ailerons to the wing! Well, it would've allowed me a pair of flaps along the inner wings, and I suppose some roll control is better than none. Still, that made me more than a little nervous about the flight. I went without - a stupid, stupid move, but I wasn't really just fiddlin' around anyways. Plane was okay without them, until the inevitable stall and spin-of-death...

[UnusualAttitude]

While TWR remains important in FAR, I wouldn't worry too much about trying to find an absolute reference value which should be valid for all craft. Unlike stock, where drag depends on the mass of the craft, and performance is therefore closely linked to TWR, FAR also looks at the shape of your airframe. Therefore a craft with draggy elements (such as the external tanks on your original design) will have poor performance even if it has a high TWR.

To compare with what others have said, I have spaceplanes that commonly have a take off TWR of 0.5 to 0.6 that can get to orbit with nearly half of their fuel left (no nukes). Less engine means a higher proportion of fuel and/or payload. Correct engine choice is also important.

Also, I would see no reason to limit the thrust of your engines. If you're carrying them, they should be pushing as hard as they can.

UA

[eddiew]

Anyways, last night I designed a new plane to test the guideline - turned out to be a 17 tonne craft with dual Turbojets. 100 m/s takeoff. Hit Mach 1 at 7k and Mach 2 around 17k, and topped out at Mach 4.2 at 30k - so I take it that was actually too much thrust?

...

No such thing as too much thrust More seriously, the only reason not to have a second engine is the weight of it will affect your available delta-v after escape, which is why I usually aim around the 0.75-0.85 TWR mark. SSTOs can absolutely be made to fly with less, but the flight to orbit can be painful. My preference is to take a two engine plane with 1.0 TWR over a single with 0.55, since the latter will spend a lot more time flying in atmosphere and need much more optimisation. A high TWR can blast you through trouble zones, and multiple gimballing engines can counter a lot of instability.

My general thoughts on intakes is 'one shock cone plus something else' per engine. The something else shouldn't be a 2nd shock cone, but could be a radial engine body or structural intake. Remember that each one you add increases the drag and slows you down. It's not always useful to trade a cap of mach 4.5 @ 27km for mach 4.2 @ 30km.

Have you tried the B9 Procedural Parts wings? I find them very reliable and easy to customise.

. All wings are completely tailored to my requirements, down to 1/8th of a metre granularity.

And something I finally tried last night was re-ordering my engine and intake placement. Turned off mirror-mode and put engine-shock-engine-shock on in that order... and gained about 5km to my peak airbreathing altitude. In fact it now flies so high that it suffers from reducing thrust long before the engines auto-switch to rocket mode. This is so worth doing on multi-engine planes! Can't believe I never bothered before...

Edited March 4, 2015 by eddiew

[capi3101]

You should try out Intake Build Aid - it does the re-ordering for you. All you've got to do is push F7.

Guys, I'm still having problems making that final push into space with planes. I don't have access to my latest craft at the moment (the dual-engine one that makes Mach 4.2 by 30k), but I think my problem at this point is oxidizer load; the stock figure I go with for oxidizer is "25 units per tonne of theoretical maximum take off mass", but that assumes that you switch over to rockets going no slower than 1750 m/s. I'm lucky to max out on the jets going 1400. I did try upping the oxidizer load but the test flight with the revised load went out of control during the attempt to switch over to rockets, and I haven't had time to try again (I bought a car yesterday and spent the rest of the day working on a VASSAL mod). As a rule of thumb, how much oxidizer do y'all typically put on your planes as compared to liquid fuel? Do you usually just go with the standard 9:11 ratio?

I should probably redesign that plane to reduce the intakes - right now she's running Shock Cone, Nacelle and six structural intakes per engine. Yes, the drag's fierce - the Q gets up around 20 kilopascals down low and the L/D is pretty atrocious if you want to know the truth. So I'll try taking off ten structural intakes and see if that improves matters.

I also need some help figuring out how to run simulations. I get that when a number is red (because red is bad) that you should look at that parameter in the simulation model and it will tell you whether the plane's behavior will stabilize or de-stabilize with time (I assume an oscillation that tones down with time is good, one that hits a asymptote is a sign that you really should redesign your plane). But honestly, I don't know how to use it. Can somebody give me some guidelines, or at least point me to a tutorial that shows me how to use the simulation models?

Edited March 5, 2015 by capi3101

[Van Disaster]

Don't forget that at mach levels over ~3.5 drag starts increasing quite drastically, so you do want to be as slippery as possible. I generally run out of jet thrust from hitting speed limits before I run out of air and if there's any reasonable payload fraction I won't have sufficiently low wing loading to fly high enough at top jet speeds to make use of more intakes, so it's generally 1:1.

No rule of thumb for l/o ratio, that can also change depending on payload fraction anyway. I never use the sim tab so I can't help there.

[Guest]

Check out this craft, It's fun to fly in FAR and a good introduction, it will break if you fly badly, not if you respect frame max tolerance. Precise and easy to control

Stock aero model quality at ksp release won't be close to FAR, and FAR has already been available for ages. Not even sure if aerodynamic stress is a planned feature

http://forum.kerbalspaceprogram.com/threads/108633-FAR-Optimized-F-51-Yorzin-%C2%A6-NEW-VERSION

Javascript is disabled. View full album

[eddiew]

Do you usually just go with the standard 9:11 ratio?

...

I should probably redesign that plane to reduce the intakes - right now she's running Shock Cone, Nacelle and six structural intakes per engine.

...

I also need some help figuring out how to run simulations. I get that when a number is red (because red is bad) that you should look at that parameter in the simulation model and it will tell you whether the plane's behavior will stabilize or de-stabilize with time...

My habit is to have mostly LFO tanks which I don't tweak, and then a small can of just fuel - either as engine nacelles, radial engine bodies, or just a little FL-400 tank in a cargo bay without any oxidiser in. You really won't burn a lot of fuel in airbreathing mode. One option is to max the oxidiser and just see the ratios you end up with in orbit Also, you really won't get past about 1400-1500m/s on turbojets with FAR. They fizzle out at this point whatever you do. This isn't so bad though, since the upper atmosphere past 30km, and especially past 50, is very thin. By and large if you hit 50 at a good speed, the only thing that can cause you to fail is a lack of fuel, or an inefficient rocket engine.

The structural intakes are probably overkill... shock cone and nacelle per engine should do fine. Radial engine body rather than nacelle, if you have them - slightly better intake capacity.

I don't really look at the graphs windows. Mostly I just check the data window and hunt red numbers. If I can fly at my usual altitude/speed checkpoints, then it'll do. The only occasional problem is that you can create a plane with insufficient control authority and it won't actually respond - but it is stable, so FAR greenlights it. You are however correct that if your oscillation damps itself out, then you can 'live with it'. If it doesn't, then you'll require active control input, and probably a lot of it. This doesn't mean unflyable however! It just means you can't go afk. You could use something like Pilot Assitant to point east and ascend at 100m/s and it'd do all the fiddly stuff for you

It might be worth pulling in a known good flyer from elsewhere, maybe RevanCorana's above, or Wanderfound's Kerbodyne Dolphin, and flying that to orbit. It took me a long time that my piloting was terrible and I was being much to aggressive with the ascents and needing too much from the rockets. Since then, I've been making planes that will hit LKO with 3-4km/s left in the tank and fly much more smoothly. It'll also help reassure you about intake and fuel ratios ^^ Most SSTO builders will have used other people's designs for reference in their early days, until they got a feel for what would and would not work.

Edited March 5, 2015 by eddiew

[capi3101]

So last night I went ahead and removed the structural intakes from my plane. That's all I did. Stupid thing started careening off the runway at 40 m/s - mind you this was with a plane that had gotten into the air successfully the night before. I wound up spending 40 of the 45 minutes with which I'd allocated to myself for playing KSP last night trying to fix the damn gear. Ultimately wound up having to put a second nose gear on. She did finally make it up into the air.

And then I almost made it. Plane got to Mach 4 at 27k, a nice 10-15 degree ride the whole way up (when it got up to fifteen I'd put the nose back down at ten). Made the mistake of switching the FAR piloting assist from "Pitch" to "AoA" and the nose dropped. Wound up getting my 1450 m/s - but by then the plane had fallen from 32k back down to 27k and I lit the rockets too low. 92x-31. I'm annoyed at myself. Maybe tonight.

[capi3101]

Quick update: I made it - finally. 85x83. Really it became a matter of not tweaking the pitch angle once the craft got much about 1400 m/s and 32k in the atmosphere. I took screenies, but forgot to put them on my flash drive before coming to work this morning, so I'll have to wait until I get home to put them up here.

Now begins the next challenge - either figuring out how to get down and land again, or doing something a little more practical with the plane...I'm also thinking about testing out RAPIERs with my next design. Why not.

[eddiew]

Yay! 32km and 1400m/s is very good; many will light rockets at 30, and something that's just bound for LKO can often get away with 25.

And yeah, there's a lesson there that once you're flying stable, do not press any more of the FAR buttons Extreme AoA is normal at altitude, and FAR's AoA limiter is dangerous because you bounce off it. If it was a gentle creep to the limit then it locked you there that'd be fine, but - as you've found - it tends to completely cut your pitch control whenever you go over what it deems a safe limit and you lose the climb. Much better to manage the pitch yourself and try to 'feel' where the tipping/flipping point is. Unless you're very aggressive with the S key, then it's quite slow and you realise you're suddenly pressing W instead.

Getting down can be easy; I build all my birds with enough chutes at the CoM so that they'll fall at 12m/s or less when dry. It's good to try for a proper runway landing, but sometimes you just want to get the darn thing out of the air

[capi3101]

So...what's a good place to make the burn for re-entry in FAR if you want to make a Runway landing? I know DocMoriarty used to suggest putting your periapsis at about 44k over KSC (assuming a 100k orbit) to give yourself a long, slow de-orbit, but that was with the stock atmo. Would that still work or not?

If not, I've seen a YouTube video where Scott Manley did his de-orbit over the crater and put himself into the ground roughly midway on that peninsula to the west of KSC; that was with FAR so I know that should work.

Not an issue for this design necessary - but what do y'all typically do in regards to solar panels? Do y'all just slap OX-STATS all over the outside, do you stick 'em inside equipment bays and stow them before re-entry, do you use the shielded ones on the exterior of the craft, what all do y'all do? Mainly just curious.

[eddiew]

I tend to be somewhere in the middle of the desert on the continent to the west of KSC, and burn until my trajectory would splashdown just a little to the east. I turn off structural failures so it's possible I'm coming in far too steeply

The perfect re-entry varies a lot, depending on the plane, its drag, its structural strength, and the height of your orbit. But my theory is that you know that with a dead stick, you'd always lose speed and impact further west than your trajectory says, and that your wings are there to turn the fall into a glide; hence I take a guess at where the plane would 'impact' after aerobraking, and try to put it just west of the mountains. Most times I end up skimming the peaks with about 2-3000 metres clear and it's a relatively easy glide slope down to the runway from there.

Aiming at 44km over KSC would have you round to 'indonesia' before you landed. Most planes will still be at mach 7 at this altitude in FAR.

[Agost]

Not using spaceplanes right now... but I actually started directly with FAR. Current stock aerodynamics just don't make sense

[Van Disaster]

So...what's a good place to make the burn for re-entry in FAR if you want to make a Runway landing? I know DocMoriarty used to suggest putting your periapsis at about 44k over KSC (assuming a 100k orbit) to give yourself a long, slow de-orbit, but that was with the stock atmo. Would that still work or not?

In the current version I deorbit about 1/3d to a half an orbit before whatever runway I'm aiming for - assuming we're starting from lowish orbit - and aim at about 20km above the target, and re-enter pitched up a couple of degrees. I am flying relatively draggy craft at the moment though.

If I'm going for KSC rather than some alternate runway I'll usually fly over at about 13km and do a big diving circle to line up on runway 27, you get plenty of time to align & slow down without dodging mountains or terrain then.

The versions of FAR for 0.25 usually made me spend a couple of orbits in atmosphere slowing down...

[capi3101]

So...I haven't landed my first flight yet; I'm in the process of building a plane that'll do something practical. Tonight I managed to crash FAR in the SPH; I don't think that it and Procedural Wings are playing together nicely. I think that largely because what I did right before going to check the new stability parameters was to go replace the center rudder with a procedural control surface - I was trying to build a large rudder to combat the sideslip on my latest design. Needless to say, the new design did not make space.

Brings up a question - when you need a large rudder, how do y'all usually go about building one with stock parts? My largest one so far has consisted of a Standard Canard. I haven't got the memory for B9, so don't bother.

[NathanKell]

Use B9 Procedural Wings (it's in the Addon Dev forum).

[Wanderfound]

Brings up a question - when you need a large rudder, how do y'all usually go about building one with stock parts? My largest one so far has consisted of a Standard Canard. I haven't got the memory for B9, so don't bother.

Wing pieces.

[Crzyrndm]

Wing pieces.

^^

Tails don't need a huge amount of control (rudder can be quite minimal), but larger area can be useful hence wing pieces make up most of my tails.

I tend to build along rather than up like Wanderfound's section though. A little less effective for the same area but it looks better to me (in my head it's less draggy, but I can't say I've tested that)

[eddiew]

You can always double up your rudders if you're really having problems This guy isn't magnificently stable in the upper atmosphere, but falls under 'flyable with concentration'. For light craft on low-g worlds, it can even bring the option of landing on the tail.

[capi3101]

My latest design has a triple rudder - three Standard Canards - and is still experiencing issues with side-slip (either that or I've got a bigger issue - plane has a tendency to want to flip around and fly backwards once I get going good). I'll try swapping the middle one out with a wing piece tonight and let y'all know how that turns out - it was that same one I swapped out last night that crashed FAR.

I should get a screenie up of that one - y'all might be able to tell me what's wrong with the design. I'm actually thinking at this point my problem is piloting - especially once I'm above 25k before switchover.

What can y'all tell me about DCA? I was kinda half-watching a YouTube tutorial on FAR yesterday (hoping to get an idea of how to go about using simulations) and it was mentioned as something the guy "really liked" but I didn't quite get what it did...

[eddiew]

If the plane wants to fly backwards, you're probably trying to have your CoL too close to the CoM. Some build guides suggest putting them on top of each other, but that's actually really bad for stability since it'll fly either way round just as happily, but you'll have a monster of a task actually reverting it if it does turn around.

Also watch out for the order that tanks drain; a common thing is that the front tank empties first so all the fuel weight is at the back - an easy scenario not to have looked at while in the SPH, since generally everyone focusses on the wet and dry CoMs only, not the intermediates. The only way I know to have a totally stable CoM at all fuel levels is to have fuel only in the nacelles, and have them dead centred on the CoM of the plane.*

Above 25-30km, pretty much everything gets slippery until 40-45, and there's not a lot you can do about it. If the FAR calcs window says 25km @ mach4 is stable, the plane is respectable. 30@5 is superb, but not required. Best advice I have is make sure you've got a reaction wheel, and if you have vernors, put some on the nose and activate RCS, because that will really help.

IMHO it's also possible for a tailfin to be too responsive if it's a canard or control surface. You can find that you tap yaw, and you end up overshooting and pointing further to the other side than you were originally off course in the other direction - which escalates in a series of over-corrections until you get a flat spin. I agree with Wanderfound that wing pieces make the best tailfins, since most of the area will be rigid with only a small control surface behind it. If that's clearly too large for the plane, you can try reducing the control deflection or yaw % that's assigned to the fins. You really don't need a lot of yaw authority, it's only for minor adjustments; for bigger turns, roll into them and let the nose come round on its own.

As for DCA - dynamic control adjustment. As I understand it, FAR will dial back the sensitivity of your controls as your speed increases. Seems to be to stop you trying to change course so fast that your plane falls apart. Seems to be some level of compensation for having lead fingers or a badly designed plane that tends to oscillate, but mostly it just gets in the way and prevents making a course/pitch change with any decent speed. If you know the sorts of AoA your bird can handle, you're better off without it and just using sensible piloting to not overdo things.

/walloftext

* For this reason, a good place to start a new build is the cargo bay, then the nacelles, lining them up to put the CoM in the middle. Everything else then goes on afterwards, maintaining that centre of mass from the start, and your cargo doesn't upset the balance of the plane.

Edited March 11, 2015 by eddiew

[Van Disaster]

@capi3101: B9 pWings are standalone, you don't need any of the B9 mod to make them work. I build tails out of pWings unless it's a really small plane and I can't be bothered, a couple of the tail pieces are all I have left of stock ( or anyone else's ) fixed wings.