[SSTO] Help my pride to orbit!

[Evanitis]

Good day fellow collegaues!

This is Bob Kerman of the Recreational Deaeronautics. Help a fellow out with some design and fly hints, will you? The thing is: our space program prospered really well under the aegis of the policy 'Math is for sissies, manly kerbals add moar boosters!' But lately the manager's wife moved onto our prospect, since she wanted to see the glaring lights of vast science. That she loved very much. But for some reason she couldn't extend her admiration to the sonic waves that rapid combustion of the inevitable 'moar boosters' happen.

So she insited on asphalting the potatoe field where we test our rovers and presonally led the undusting of the SPH. She wants a rocket that doesn't have exploding parts. Blasphemy! But what could we do, she's the bosses wife afterall. So we started testing of such craft. Geting to space on these is haaard. Well, harder than it was in Jeb's early days he says. Anyways, let me show you the most succesful variant. Or at least the best looking one.

Javascript is disabled. View full album

Yeah, I know, we could have uploaded the plans, but here at the Recreational Deaeronautics, we believe in eyeballing stuff. Anyways, having trouble gettin' to orbit... or, um... even out to space for a quick sip of vacuum. Regardless how -right- it looks (and handles). Not sure if it's a design flaw or flight error.

We tried adding more boosters or bigger engines with less success. Mind you, we don't have the really advanced kinda' stuff - you can see our options on the above screens. Am I just dreaming too high with our capabilities? That payload is only like 3 tonnes... well, that sounds much now that I better think of it. *note: do an empty testing too*

Anyways, Jeb says that all is well until he goes fast or goes high. He heard somewhere that one sould be able to get to 25k on such with like 1200 m/s. Well, he tried fast ascension, but stuff started to explode, and the point was to avert that. It seemed better to do the acceleration above 18k but Jeb couldn't gain speed and altitude at the same time. Air is too thin I feel, but adding moar... umm, of those small intakes didn't help. Boss's wife is furious.

So, fellow Kermans... what are we doing so seriously wrong?

PS.: How fast is Mach-3?

Edited June 8, 2015 by Evanitis

[cybersol]

Mach 3 is 340 * 3 = 1020 m/s

For your plane, whats the TWR of the 3 909s when you first switch to rocket mode? I'm guess its a little low. Have you thought about using a Poodle engine there?

For spaceplane ascents to LKO:

- Lift off

- 10-20* pitch up until speed above 270 m/s

- Raise pitch to maintain speed below 290 m/s (high drag above this speed)

- At 10-14km, pitch down to 0-5* and accelerate until 500 m/s

closer to 10km for only turbojets and to 14km for only rapiers

- Slowly raise pitch to 15* while building speed

5* pitch up per 50 m/s speed increase above 500 m/s seems to work well

- Grab the last bit of airbreathing speed by pitching back down to 5-10* upslope

Do this around 17km for turbos and 20km for rapiers

- As soon as your speed indicator stops going up, punch the rockets and/or switch modes on the rapiers

- Get out of dodge by aiming at 20* upslope until 32km

- After 32km follow the top of prograde marker until your at 5-6* pitch

- Cruise until your desired apoapsis

- Cruise out of atmospshere to apoapsis and circularize

You don't have to do all that exactly, but its what I do when efficiency is more important than lackadaisical piloting.

[Cairol]

How is your CoM / CoL balance?

What's your athmospheric TWR?

You seem to have a rather small wing area and small control surface area.

Mach 3 is about 3 x 330 m/s = 1000 m/s.

I'm not so expierienced in building SSTOs yet. So other users can surely help you further.

[MarvinKitFox]

Umm. You have two of the MKII *rocket* fuel fuselages there, and two FL-T100 fuel tanks.

Just what is your actual LiquidFuel and Oxidizer loadouts, on the runway?

hmm...

I have copied your design as exactly as I can.

Fuel loadout: remove OX from side tanks, and 1/2 from rear fuselage.

Flying technique.. Very complex!

full throttle

take off

point 30 degree up.

When jets die, press space. Also now, tell your pilot to hold prograde.

achieve orbit without pressing another button.

Admittedly, this leaves me with only about 25m/s left in the tank, at 75kmx75km orbit.

If you replace that tri-coupler and 3 engines with one T45 engine, the deltav remaining is about 160m/s, but you need to fly more than just "point prograde"

Replace the airscoop with a shock cones and do a bit better yet.

Moral of the story:

You don't have enough engine to achieve level flight over 22000m altitude, or 1050m/s sustained.

But you *can* scoot directly out of the atmo at a 30* angle, at about 1280m/s(at 24km)

Difference? Airscoop storage capacity, i think Lasts about 2-3 seconds, while you are transiting from 20 to 22km

Edited June 7, 2015 by MarvinKitFox

[rofltehcat]

Try using two swivel engines instead of the Terriers (or maybe a mix). My attempts at similar designs on that tech level were prettyunsuccessful until I used swivel engines. Before that the TWR was just too low. Poddle might work as well but didn't for me.

It also looks a little heavy. What is its current weight?

[Evanitis]

Thanks for all the replies.

- Raise pitch to maintain speed below 290 m/s (high drag above this speed)

- At 10-14km, pitch down to 0-5* and accelerate until 500 m/s

*tries to hide stupid expression* That may be the root of my problem. I was under the assumption that all is well until the flames don't erode the engines.

Whats the TWR of the 3 909s when you first switch to rocket mode? I'm guess its a little low. Have you thought about using a Poodle engine there?

Can't check the number right now, but it feels low indeed. Poodle you say? I didn't think about such bulky looking ugly engine. Swivels didn't feel much better, but I'll give both another go. In case your advanced flight advice won't be enough on it's own.

How is your CoM / CoL balance? What's your athmospheric TWR? You seem to have a rather small wing area and small control surface area.

CoL is bit behind CoM, stays the same when fuel and payload is out. No lift or stability problems at any point, so it feels just right on the wing department.

Fuel loadout: remove OX from side tanks, and 1/2 from rear fuselage.

I got as far as removing side OXes (and having as low amount of jet-fuel as the slider allows). What's the point of removing from the fuselage OX? I had the impression that the rocket stage needs all of it, while I still have excess for the jets.

Replace the airscoop with a shock cones and do a bit better yet.

Will try it, thanks. Well, once technology advances a bit more.

Try using two swivel engines instead of the Terriers (or maybe a mix). (...) It also looks a little heavy. What is its current weight?

Yup, playing around with engines is on the list. It's 21 tonnes with payload, 19 without it.

Edited June 8, 2015 by Evanitis

[Jesrad]

You can probably settle down to 2 Terrier only, that'll help streamline the craft some more, and aerodynamic drag is of utmost importance to spaceplanes. That will mean a flatter path into the 12-20 km altitude zone, at a higher speed and higher temperature, though. I advise you to add precoolers between the ram intakes and FTL-100s on your ship - it'll make them endure the heat logner and it'll also add the extra liquid fuel you'll need.

And here's a little-told dirty secret of KSP spaceplanes: if your plane's atmospheric TWR is > 1.0, then you should ditch the wings entirely and take off vertically, because:

- during the jet-powered ascent you'll already be climbing at a steep angle and from then on pitching down continuously, so the lift in this phase is wasted

- the wings will drag your craft down during the very critical rocket phase (20-40 km altitude) and you'll be pointing your ship prograde anyway to minimize drag, so they won't be generating lift here either (zero angle of attack)

All told, the wingy bits are there only for stability.

[rofltehcat]

21 t is actually a lot lighter than I had expected. Then 2 turbos are more than enough. Still not sure about the terriers and possible poodle. Your design is lighter than my monstrosities so you'd probably be fine with a poodle. Do you use Kerbal engineer reloaded? It allows you to move the slider to e.g. 20 km and it'll tell you the TWR/dV for activation on that altitude.

[capi3101]

Me, I'd swap out that three LV-909 concoction you've got on the back with a single Swivel engine. It'd save you the 0.15 tonnes for the mass of the Tri-Adapter, you'd get 125 kN more thrust, and (the big ones) you'd see a huge boost to your delta-V - those Terriers have an atmo Isp of 85 while the Swivel's atmo Isp is 270 and you would lose a very drag-inducing part in the process (namely the Tri-Coupler again).

Yes, the Terriers and the Poodle have better Isps than the Swivel once you're in space...but remember, you're going to be lighting the damn things while you're still in atmo. In the meantime, they're gas guzzlers.

Your wing setup looks a little anemic to me. How does the plane handle?

[eddiew]

*tries to hide stupid expression* That may be the root of my problem. I was under the assumption that all is well until the flames don't erode the engines.

Actually I agree with your assumption. My usual ascent is:

- get to 10km as fast as the plane can handle

- pitch between 8 and 10 degrees

- don't fiddle the controls until you're upwards of 30km and in rocket mode

Sure, the intakes may get pretty toasty, but they start cooling after 19km whatever speed you're at, and the shallow climb allows you to really get some mileage from the air breathers. If you've got the bottle to redline the thermal indicators, you can see over 1400m/s (rapiers) before switching to rockets. Even with turbojets, a km/s or more is not impractical and highly desirable

[A_name]

And here's a little-told dirty secret of KSP spaceplanes: if your plane's atmospheric TWR is > 1.0, then you should ditch the wings entirely and take off vertically, because:

- during the jet-powered ascent you'll already be climbing at a steep angle and from then on pitching down continuously, so the lift in this phase is wasted

- the wings will drag your craft down during the very critical rocket phase (20-40 km altitude) and you'll be pointing your ship prograde anyway to minimize drag, so they won't be generating lift here either (zero angle of attack)

All told, the wingy bits are there only for stability.

Can you elaborate on this a bit? How would you achieve the needed horizontal speed with this profile? Wouldn't you have to constantly point above prograde to not keep from falling back when you start leveling out in the middle atmo?

[Warzouz]

... and (the big ones) you'd see a huge boost to your delta-V - those Terriers have an atmo Isp of 85 while the Swivel's atmo Isp is 270 ...

That's mostly irrelevant. When you have to fire your rockets, the ISP is already near VAC value. ISP rises very fast with altitude. This is very easy to check as ISP is shown in the menu of your engine. Just open the menu while climbing (it must be activated though)

But the drag issue is very relevant, though.

When speeding up, I usually use "prograde" to reduce drag. I do it twice : on first acceleration from 250m/s to 500m/s at 12000m and after 35/40km on final rocket burn. The rest of the time I'm on a 5/15° AoA (500 to 1200m/s acceleration) and start rocket phase (25 to 35km).

Hitting F12 helps a lot to see drag increase.

[capi3101]

That's mostly irrelevant. When you have to fire your rockets, the ISP is already near VAC value. ISP rises very fast with altitude. This is very easy to check as ISP is shown in the menu of your engine. Just open the menu while climbing (it must be activated though)

But the drag issue is very relevant, though.

Yeah, I did the math on that one right after I posted it...came to the conclusion that the Isp of the Terriers/Poodles surpassed the Swivels when the atmospheric pressure got down to the neighborhood of 1/10th of an atmosphere. That happens in the stock atmo around 10k, so the whole Isp argument is moot. And then when it comes to thrust - the Poodle starts outdoing the Swivel at 0.3 atmospheres...right before 7500 m.

I think I'd still recommend the Swivel, though, judging the OP's current tech level from their plane. At least until they get Advanced Aerodynamics (and the Mk2 to 2.5 meter adapter that comes with it); for now they could replace the tri-adapter with an emptied Mk2 to 1.25 adapter and mount the Swivel to that. Doing that would cut down the drag profile.

[ajburges]

And here's a little-told dirty secret of KSP spaceplanes: if your plane's atmospheric TWR is > 1.0, then you should ditch the wings entirely and take off vertically, because:

- during the jet-powered ascent you'll already be climbing at a steep angle and from then on pitching down continuously, so the lift in this phase is wasted

- the wings will drag your craft down during the very critical rocket phase (20-40 km altitude) and you'll be pointing your ship prograde anyway to minimize drag, so they won't be generating lift here either (zero angle of attack)

All told, the wingy bits are there only for stability.

What the hell flight profile are you using? Wings still are useful for the 10-30 km range where you build speed since they allow more efficient flight profile there. (where you do 40% of your orbital acceleration) Granted a high TWR jet needs less wing, but it still wants some. The only time jets don't need wings is when you enter the wasteful 1.8 TWR range where they behave as aspirated rockets. Wings also improve margins for landing.

For orbital ops space planes, TWR of the anaerobic engine is the prime consideration. With the extra mass they take, they will never be an economic choice for operation beyond LKO. So you only need a few hundred dV. However a TWR above .7 at the start of anaerobic mode is very important to a successful accent to 40 km where drag falls off and an efficient gravity turn doing so. The smaller the engine to do so, the more efficient because of mass and drag. Too much engine surface at the back is surprisingly draggy.

These factors are why the aerospike is such a good space plane engine. It has good TWR, minimal drag, and top tier efficiency.

Low TWR (.3-.6) designs either need to fly at 30-36 km for extra lift while building horizontal speed for an accent like .90 (wasting dV to the reduced but still present drag), or waste a lot of dV in radial thrust.

[Evanitis]

Thanks for all the answers again! Got the plane to orbit without changes (so I could test and feel how better are the various engine suggestions). Halfway there towards the K-Prize challenge - now I just have to figure out landing. But it looks easy, gravity is on my side.

You can probably settle down to 2 Terrier only, that'll help streamline the craft some more, and aerodynamic drag is of utmost importance to spaceplanes.

That was my initial tweaking idea too, but I didn't have the two-way separator unlocked. One more setup to test besides the single swivel or poodle.

Your wing setup looks a little anemic to me. How does the plane handle?

Feels perfect. I had a struggle with FAR in the earlier versions. I won, but it was a long, hard fight. I'm not sure if I learned my lessons than, or the new stock aerodynamics are that forgiving. I was told to use less wings in 1.0, so I slapped on the first sleek-looking ones. I was pretty surprized how smoothly it took off on the first try.

Do you use Kerbal engineer reloaded?

I'm familiar with the math-basics, but I'm having more fun if I never look at any numbers. Eyeballing, ballpark estimations, lot of testing. Or doing stuff untested. Didn't look at a single TWR or dV readout since 1.0 is out. Some days I envy the expertise of the above forumers, but I love my silly designs, hilarious mistakes and the occasional gargantuan miscalculations. Normal Career still feels too forgiving, even with that mindset.

[capi3101]

Feels perfect. I had a struggle with FAR in the earlier versions. I won, but it was a long, hard fight. I'm not sure if I learned my lessons than, or the new stock aerodynamics are that forgiving. I was told to use less wings in 1.0, so I slapped on the first sleek-looking ones. I was pretty surprized how smoothly it took off on the first try.

That's good to know. Myself, I use newFAR, and I find with each passing day that I'm forgetting the tricks of the stock aero model. Not a good thing; I like to be able to help out spaceplane jockeys regardless of what air they try to fly in if I can.

I would be careful when you come back down to do your best and try and control your rate of descent - too fast, and you'll burn something up. A.I.R.B.R.A.K.E.S. are your friends - I'd suggest getting them unlocked at the earliest opportunity - but so is having plenty of pitch authority. If you can control your pitch, you can control your temperature - just slow down your rate of descent if it looks like something's trying to overheat.

Your CoM position will also have changed on account of your plane having used up its rocket fuel to get into orbit. Now, it looks to me like you've got more mass ahead of your tanks than behind them, so the CoM should shift towards the nose; your plane will probably be slightly less maneuverable than it was on the way up but still controllable, and that's good as long as it doesn't shift forward too far (and then you get lawn dart behavior - which is bad on entry since you can't effectively slow sown your descent if you need to). The CoM going forward is a damn sight better than the alternative - the CoM shifts aft and winds up aft of the CoL, which is bad since at that point you're dynamically unstable and if you overwork the controls in the slightest, you're going to head into a spin and likely a crash. If in the future you want to be sure which direction your CoM will go (or if you want to set things up so it doesn't move much at all), try out a mod called RCS Build Aid - among its useful features is a "dry center of mass" indicator, which will show you where the CoM will be when your plane is out of fuel.

One other mod I'll throw out there is NavUtilities. It's an integrated ILS, which will make a Runway landing much easier to accomplish. Barring that, there's the old "flags on either end of the Runway" trick, which I'd be happy to elaborate on further if it's something you're interested in.

[cybersol]

Congrats on getting to space. You posted your spaceplane weight later in the thread after my poodle suggestion, and for that weight the 3 terriers or 1 swivel is likely the sweet spot.

For landing tips:

1) Airbrakes are great additions. Turn of all control like pitch/etc on them, and use their braking feature to slow down or change your landing position. Bias them towards the top surface to help with 3).

2) Move your remaining fuel to the front of the plane before re-entry (usually a good fix for capi's CoM thoughts above)

3) You can also pitch up (cobra style) to slow down faster, but you may want a quicksave your first few times with a particular plane, until you find the limits.

[ajburges]

Congrats on getting to space. You posted your spaceplane weight later in the thread after my poodle suggestion, and for that weight the 3 terriers or 1 swivel is likely the sweet spot.

For landing tips:

1) Airbrakes are great additions. Turn of all control like pitch/etc on them, and use their braking feature to slow down or change your landing position. Bias them towards the top surface to help with 3).

2) Move your remaining fuel to the front of the plane before re-entry (usually a good fix for capi's CoM thoughts above)

3) You can also pitch up (cobra style) to slow down faster, but you may want a quicksave your first few times with a particular plane, until you find the limits.

I find that with sufficiently stable craft, reentry with too much AoA trends to be self correcting. If your craft is engineered to be able to stall (in the manner of early shuttle concepts) just remember to exit stall before 500 m/s when you start entering the trans-sonic drag.

Airbrakes are definitely OP. Deploying one takes you out of heating effects in a few seconds. Like lift surfaces, slipstreaming has a greater effect the further from the CoM they are. I tend to place mine far back. That means they grant little control authority, but they are also less likely to put me in a flat spin!

Alternative to fuel balancing in orbit is to align (near) dry mass appropriately for landing, and make sure you don't bring rocket fuel back down. Drop off unused fuel at an orbital station keeping 80 m/s for deorbit.

[Evanitis]

I like to be able to help out spaceplane jockeys regardless of what air they try to fly in if I can.

Thanks for that habit. xD

CoM shifts backwards, though not as seriously as it did with my early version SSTO designs. Added the wer-dry CoM, CoL images to the album in the OP- with included MJ readouts for the math enthusiasts. The rearranging of the remaining fuel sounds like a neat idea, and I have enough space in the cargobay to fit some small tank in for that purpose (and for an extra sip of dV for the rocket phase.)

Tried to land the plane, got a flat-spin. Back than I used 'chutes to negate this effect, now it didn't feel that effective as in the earlier versions. Not that it was too reliable even than. Crew survived though, the payload is in orbit and it could finish two Sat-missions, so the experiment still turned some profit. I also got a perfect alignment of 5 Minimus missions in one go, so tech-level is maxed and I can unlock all the neat gadgets that were mentioned in this thread. Starting with aerobrakes. Think I'm good for now.

My comp is from the last decade, so I tend to skip mods that I can workaround in stock. Think I got a pretty accurate dryCoM readout by removing the payload and the fuel. I didn't know about the flags trick - anything more to it than putting them there so my runaway is marked?

Edited June 9, 2015 by Evanitis

[Warzouz]

When reentering, I usually turn the plane upside down and force it to go downward. I create a lot of drag and the lift pull me to the ground faster which alose create more drag faster.

Doing that maneuver, I quickly cut my speed and I can target the landing point more accurately. If I'm too far from the landing target, I do the same maneuver but no upside down so I can glide further to the target.

[capi3101]

CoM shifts backwards, though not as seriously as it did with my early version SSTO designs. Added the wer-dry CoM, CoL images to the album in the OP- with included MJ readouts for the math enthusiasts. The rearranging of the remaining fuel sounds like a neat idea, and I have enough space in the cargobay to fit some small tank in for that purpose (and for an extra sip of dV for the rocket phase.)

Tried to land the plane, got a flat-spin.

And looking at the dry screenie, it's easy to see why...the CoM shifts aft of the CoL - your plane becomes flip happy in that configuration. Definitely want to get a little more mass up towards the nose if you can. Barring that, you might try moving your wings back ever so slightly. A little dihedral (angling your wings upward 5 degrees or so) wouldn't hurt either.

My comp is from the last decade, so I tend to skip mods that I can workaround in stock. Think I got a pretty accurate dryCoM readout by removing the payload and the fuel. I didn't know about the flags trick - anything more to it than putting them there so my runaway is marked?

Same boat on the CPU - on the chip it says "Made in Byzantium".

There's not much more to the flags trick than that. I usually use rovers instead of flags because my flags have a tendency to vanish or spontaneously combust on me - and once they're in place, I set the parking brake and re-label them as bases; this gives you the bonus of being able to see where they are A) from orbit and a hundred klicks out, which is reaaaaly handy when you want to get aligned well in advance. I usually also add one or two more rovers on the inland side of the runway (the west side) at 1 kilometer and 5 kilometers out; you can continue adding them at five kilometer increments if you like - the most I had at one time went out to 30 kilometers (that was in 0.21 or so and with the next version of the game the Runway shifted over, rendering my ILS useless for alignment - otherwise I might still be using that one). The more markers you've got, the more points you have to align, the more you can be assured that your alignment is good from further away.

And then, if you label them judiciously, you can use the same ground markers to help you find a good glide slope (I did the math once and it came out to something like a 5.7 degree slope, a tad steep but it does help you avoid the mountains west of KSC pretty effectively). Here's how that works - label your ground markers (regardless of if it's a flag, rover, whatever) by the distance they are to the Runway (I label the ones on the Runway as KSC 09 and KSC 27, and the ones inland as "Meatball - 1 km", "Meatball - 5 km", etc.). When you're looking at your markers, note the distance you are from the marker, and add to that the distance the marker is from the Runway (for example, let's say I'm 4.7 kilometers from Meatball - 10 km; 4.7 + 10 = 14.7). You multiply that total by 100, and then add 100. The result is what you want your altimeter to read (so in the aforementioned example, (14.7 * 100) + 100 = 1570); if your altimeter reading is higher, you're too high and if it's lower, you're too low, simple as that. You have to be able to do the math in your head pretty quickly (or just hit pause frequently), but it's a pretty effective system. I used to use it before I installed NavUtilities, and even to this day I'll still set rover/bases on either end of the Runway pretty early on.

Note that if you're trying to land on KSC 27, the system still works (you can still align with the inland ground markers), but you'll only be able to use the KSC 27 marker to judge your glide slope.

No rover tech? No problem - you've already at least unlocked Aviation, otherwise you wouldn't be asking about planes in the first place. Unlock Basic Science (for 45 sci) if you haven't already so you've got the Stayputnik. Then build an assembly consisting of a Stayputnik, a battery pack or two, a Mk1 Fuselage, a Wheesley engine, a Radial Intake, and add a set of tricycle bush-plane landing gear. Add a radial chute for good measure. Limit the thrust of the Wheesley to about 10%. For the KSC 09 and inland markers, start off slowly - just enough thrust to get going, get turned around, and drive it just off the edge of the Runway. Kill the thrust, pop the chute, set the brake, relabel as needed. For the KSC 27 marker, you can open up the thrust a bit but pay attention to the piano keys - you don't want to be going too fast when you get to the end of the ski-jump ramp, otherwise your little pseudo-rover is probably going in the drink. Same as before - get off the raised green area, kill the thrust, pop the chute, set the brake, relabel.

And of course, you can always use flags if your Kerbals can plant them; just because they commit random acts of magic for me doesn't mean they'll do the same for you.

I hope that's helpful to you.

Edited June 9, 2015 by capi3101

[Evanitis]

And of course, you can always use flags if your Kerbals can plant them; just because they commit random acts of magic for me doesn't mean they'll do the same for you.

Don't think I ever had a disappearing flag - though I don't think I ever zoomed above them at supersonic speeds either. Anyways, placing various markers at set distances with any technology is an easy enough task - compared to landing an SSTO. Tech-tree is almost done by the way - and I never even left Kerbin's SoI. I also think computer performance will be a bigger limiting factor than finances. That 'normal' difficulty is too easy.

Ahh, going off topic here... Thanks again, see ya out there!