Why does my spaceplane run out of fuel attempting a West-bound/CW Orbit?

[sebjf]

Hi,

I am trying to make a spaceplane. I've managed a couple of craft that can get into orbit with 500-1000 m/s dv remaining, but when I try to launch westward (to pick up crews from Asteroid intercepts), I cannot get into orbit at all.

Below is the latest. The ascent profile I aim for is to get to 1300-1400 m/s around 1-2km, then pitch up to 20 deg, and around 9km engage the nukes. (I just tried it just now - 1200 m/s by 3km, engaged the nukes at 12km, burned all engines at 20 deg until the air breathing engines ran out, then burned prograde until AP hit 70km. Orbit established with 830 m/s dv remaining.)

When I aim west though it doesn't work.

I can never get the AP as high before reaching it as I do going east. I know I have to fly against the rotation of Kerbin, but thought that was only 175 m/s or so to cancel the initial rotation?

Can anyone see where my plane is going wrong?

Here is a picture of it, and the craft file: https://www.dropbox.com/s/oo3hpgzdmwh84b1/NX3.craft?dl=0

I haven't unlocked RAPIERS yet.

Sj

Edited October 23, 2019 by sebjf

[Starhawk]

When you launch eastward you already have the rotational velocity of Kerbin added from the point of view of the orbital frame of reference.  When you launch westward you lose that and, indeed, start with negative velocity from the point of view of the orbital reference frame and your desired orbital direction.

Since Kerbin's equatorial surface velocity is 175 m/s, you need twice that or 350 m/s extra velocity to achieve a westward (retrograde) orbit.

Happy landings!

[sebjf]

So that's why its double, thank you!

But still, if I can get into orbit with ~800 m/s to spare east, where does the other ~500 go?

Obviously I could be flying badly, but I have tried *many* times and I can get no where near orbit like I can east. I can't even get the AP above 40km at any point, let alone raise it high enough and keep it there to escape the atmosphere and achieve orbit.

Is there something else about going west that I've missed?

[Fierce Wolf]

Build a rocket instead of a plane and turn west. Why a rocket? It goes straight up as default and deals less with atmosphere and the not-so-good-at-flying factor of a plane. See how much does it cost to achieve orbit going west, then east with the same design. Post results!

[Starhawk]

Well, the longer it takes your spaceplane to get to orbit, the longer you spend fighting drag.  I imagine the losses come mostly from that.

I once actually toyed with the idea of creating a 'Go West Young Kerbal' challenge for launching spaceplanes westward into orbit.

I've done it myself and been surprised at how large the loss was, but that was quite a long time ago and I don't remember the details.

If my analysis is mistaken I am sure some helpful forum member will come along and correct me.

Happy landings!

[bewing]

Your description of your ascent profile sounds pretty inefficient. Blasting along at 1200 m/s in dense air is a lot of fun, but it's going to waste immense amounts of fuel. If you'll be a little more patient and save the majority of your speed run until you've climbed over 15km with a much lower pitch angle, I think you'll find you have plenty of fuel to make orbit. Also, those particular swept wings you are using are a trick to trap newbies. Every other type of wing gives twice as much lift for the same drag as those ones.

Edited October 23, 2019 by bewing

[Starhawk]

18 minutes ago, bewing said:

Your description of your ascent profile sounds pretty inefficient. Blasting along at 1200 m/s in dense air is a lot of fun, but it's going to waste immense amounts of fuel. If you'll be a little more patient and save the majority of your speed run until you've climbed over 15km with a much lower pitch angle, I think you'll find you have plenty of fuel to make orbit.

This is exactly correct.

I hadn't paid enough attention to your ascent profile when I first looked at your opening post.
I believe that @bewing's advice will net you much more dV on orbit and make it very easy to achieve even a retrograde orbit.

Happy landings!

[sebjf]

Thank you all for the advice!!

I have replaced the wings.

I have also tried the new ascent, but I don't think I am doing it correctly.

I aggressively get to 15km (by starting at 45 deg or so, keeping speed at about 250 m/s) before leveling out. The air engines to push the speed 1300-1400 m/s, before I engage the nukes. The speed continues to increase and I rise slightly to keep from overheating.

The problem comes when the plane reaches about 21 km and the air breathing engines cut out. At this point the surface speed (1500 m/s or so) continues to increase but too slowly, and the time to AP (a few seconds to a minute) decreases. I encounter AP before gaining enough speed and the plane falls back into the atmosphere. Then its a case of either burning up, or letting the nukes push against the atmosphere until fuel is exhausted.

I can push the AP away by pitching up, but then surface speed doesn't increase at all. If I aim the plane down, surface speed increases quickly, but the time to AP decreases even quicker!

I don't think my nukes have enough thrust to overcome the drag, which is why I need to ensure the trajectory had the AP above 50km before the air-engines cut out in my original ascent.

I have tried this many times, I managed to get into orbit by starting a speed run at 10k and pushing up at 20 deg with nukes, but with no more dv than originally

Is it a reasonable conclusion that my rockets don't have enough thrust? Apparently people have got planes up with just one LV-N.

[Vanamonde]

The nuclear engine is very fuel-efficient, which is important for flying to distant places after you're already in orbit, where the rate of acceleration is unimportant. But it also has a very low thrust/weight ratio, which makes it a terrible choice for spaceplanes because they need to build up speed within a time limit to reach orbit in the first place. It can be done, but you're making things harder on yourself by choosing that engine for your upper stage. Try a higher thrust engine for the rocket phase of the flight, or save even more weight by using Rapier engines. Those can function both as an air-breathing jet and as a rocket in vacuum, so you can replace two engines with one. 

[Signo]

Try switching your airbreathers with rapiers, give some angle of attack to your wings, 5 degrees should be good and easy, try moving your command surfaces to the back to reduce drag. Rapiers Will work till 27/28k meters. Try to push prograde as much as you can while only nuke propelled from 28k to 38k. I usually stay at 5°/6° until my Pe goes positive then I level to 0°/1°.

Rapiers will push you at their best from 6k to 15k, try to gain as much airbreathing Speed as you can in that range and then try to maximize your vertical Speed from 15k to 19k before powering nukes. Good luck.

 

Edit: I am a big fan of LF-only designs.

 

Edited October 23, 2019 by Signo
Pic

[Streetwind]

10 hours ago, sebjf said:

I aggressively get to 15km (by starting at 45 deg or so, keeping speed at about 250 m/s) before leveling out. The air engines to push the speed 1300-1400 m/s, before I engage the nukes. The speed continues to increase and I rise slightly to keep from overheating.

The problem comes when the plane reaches about 21 km and the air breathing engines cut out. At this point the surface speed (1500 m/s or so) continues to increase but too slowly, and the time to AP (a few seconds to a minute) decreases. I encounter AP before gaining enough speed and the plane falls back into the atmosphere. Then its a case of either burning up, or letting the nukes push against the atmosphere until fuel is exhausted.

A few suggestions for further optimizations:

You're flying with ramjets, which do poorly at subsonic speeds. While @bewing was right about not hammering through the atmosphere at mach 4 a scant kilometer above the surface, you actually do want to go supersonic reasonably quickly. How exactly you do that depends on the exact performance of your plane, but as a starting point, try a 30° ascent off of the runway, full throttle. Your speed should be climbing at a reasonable rate. If it climbs very quickly, then go steeper. If it climbs very slowly, or not at all, go flatter. Once you hit 300m/s, feel free to flatten out a little, because the sound barrier is incredibly draggy and you want to spend as little time in it as possible. Once you hit 400m/s, return to your 30° pitch. Your ramjets should be starting to seriously get going at this point, and effortlessly maintain that pitch.

Now your speed and altitude should both be climbing at a brisk and steady pace, up until a point around 1100m/s when you notice your ramjets slowly starting to run out of breath. At that point, check your altitude. Are you above 10km? If not, your ascent was way too shallow, and you should revert and try again with a different profile. Are you above 15km? You were going too steep, flatten out immediately, or revert and try again. What you should be doing right now is eke out the most speed possible from your wings and ramjets before they call it quits, so from here on out, stay reasonably flat at around 15-16km and just accelerate. You can slowly let yourself drift higher, but only slowly.

When the time comes that you notice your acceleration fall off sharply, it's time to stop being a plane and start being a rocket. Fire up the nukes, and pitch up, up, up. Not too suddenly - you don't want to slap your flat bottom into the airstream in the blink of an eye. But get that nose up, or you'll experience exactly what you described and not make orbit. The exact pitch you need depends again on your plane's performance, but 45° should be a decent start. The goal is to find a pitch setting that will push your apoapsis just high enough that you'll be able to insert into orbit on rocket power before reaching it. From here on out, it's like you're a rocket's upper stage, in a slow gravity turn.

If you have excess liquid fuel, you can let your ramjets run dry, but beware of asymmetric flameouts near the end. You'll know best whether or not that is a problem for you.

Edited October 24, 2019 by Streetwind

[Signo]

I built a version of mine of the craft and actually it is a tough craft to fly, worth a "Go West" challenge. There are issues starting when you reach 15000 because the ramjets are "spent". They actually flameout at 21k but you are left with a -400Km Pe to fill. It was however fun to fly: I tried to stay as much low tech as possible.

Craft file available on request.

 

Cheers.

 

[FinalFan]

I'm just spitballing here, but I want to share a thought about why it might be that there is so much trouble over "only" 350m/s.  

First, let's consider airbreathing mode.  This is the "easy" part, not in terms of designing it or flying it but in the sense that the very efficient jets give you a huge head start on the dV you need for orbit.  Let's say the jets do all the work of going up to 1200ms, and 2/3 the work of 1200-1500m/s, for 1400 out of 2400.  (More than 2300 because orbital speed is higher at lower altitude.  For simplicity I will ignore the altitude and atmosphere and hope it isn't too far off.)

So you're going east at 1500 (surface) and need to be going 2400 (orbital) and you get 175 for free (surface speed assists orbital speed).  That leaves only 725 dV that the rockets need to supply, plus 100 they already did for 825 total; and the rockets are always less efficient than jets and also less powerful on a NERV spaceplane like yours is.  

Now you want to go west instead, and the difference in dV of orbiting the other way is only 350 (2400+175 instead of 2400-175), but it's not 350 added to 2225, it's 350 added to 825, or a more than 40% increase in the work the rockets have to do!  No wonder you had trouble, when putting it that way.  

[sebjf]

Thank you for all the great replies! 

I'm glad to hear it is a tough problem, opposed to a really simple mistake! 

I've tried to improve my ascent efficiency but am still having trouble. Here is a video of my attempt to do a high-altitude speed run with the NERVs: https://www.dropbox.com/s/cnmpttc31wjx4hl/Nx3 High Altitude Speed Run-1.m4v?dl=0. I don't seem to have a good TWR for long enough to get the AP high enough. Is there an obvious place its going wrong?

Still so far the only way I've managed to orbit, is to accelerate to Mach 4 at 1-2 km, then throw the plane up at a high pitch, using the entirety of the air-breathing altitude (+ NERVS after 9 km) with the supersonic performance of the Whiplashes to build up enough vertical speed such that it leaves the atmosphere, even after the TWR drops. But that only works pro-grade.

I am in career mode and don't have RAPIERS yet, but I did redesign the plane to use Aerospikes, thinking that the higher TWR would be more forgiving when doing high-altitude speed runs. Unfortunately I don't seem to do much better! I achieve Mach 4 at ~18km before engaging the rockets, but it seems they have too much work to do and I barely make orbit. https://www.dropbox.com/s/lndebycxwz7fohl/Nx4 High Altitude Speed Run-1.m4v?dl=0]

Signo seems to do much better with engines that are less efficient than the Aerospikes if I am correct, so this must be a problem with the ascent? 

I have tried with a sharp and shallow pitch. With a sharp pitch I can indeed push the AP to 70, but the PE is so low I don't have enough fuel to circularise. On the other hand if I try not to fight gravity it seems I waste a lot pushing against the atmosphere and end up with less dV than with the NERVs.

I will try again with some combinations of pitches over time...

[Signo]

I watched your video and I will try to write what I do "different": mind that it is just the way I do it and we may surely find there is a better way.

First of all, try to maximize the speed you can reach with the ramjets in their "sweetspot" around 10k meters. Do not be in a hurry to reach your desired Ap and do not be scared of the heat gauges popping, that is the way it should be. You should not be too fast either: if you are over 1400m/s the ramjets can not push you so much further so you will not have enough kinetic energy to transform your horizontal speed in the vertical speed you need to overtake your "bitter spot" that comes from 15000 to at least 25000. Turn on the nervs as soon as you decide to go for your "jump" above that spot and of course do not turn off the ramjets to exploit whatever they may provide until they flame out.

Do not pitch too much or your wings will be airbrakes instead of helping you with their lift. Try to push prograde instead of "up" and to always gain speed: it will be slow at first but the more you go up the better it will be. The angle of attack of your wings will push you up even if your craft is flying a 0° pitch course.

Do not look for you Ap in the usual place you look for it: with the craft I pictured in this thread I reached my actual Pe when I stopped the engines - flying an "airplane" like this really makes you go "ballistic" so your real objective here is to fill up your Pe on the other side of Kerbin; you will then end up your ascent being more or less at the Pe of your orbit and you will ascend out of the atmosphere to your desired Ap. Usually it takes me almost a full orbit to circularize.

About the craft, it seems to me that nobody mentioned but I do lock the gimbal of the airbreathers to avoid the thrust "going around" instead of perfectly prograde and through the CoM. You have an airplane and not a rocket so you should have plenty of control surfaces to move wherever and whenever you want. You just do not really need an active yaw control because there is no side wind on Kerbin.

 

Keep up your flights and enjoy, this is maybe the best game ever because you can learn new ways of flying even after thousands of hour of play.

 

 

[paul23]

Honestly, if you have trouble getting the required orbit with 13k delta-v another engine might be better to use.

 

13k is a lot, normal rocket decent uses about 4k for a retrograde orbit. I think the biggest offender are the nervs: they're simply not the right tool for the job.

[Signo]

@paul23: it is 13k in airbreathing mode as far as I can see. I do not even look at the readings of KER/MJ when building stuff like this but just at the actual amount of fuel I am carrying. 

[Laie]

12 hours ago, sebjf said:

... throw the plane up at a high pitch, ... such that it leaves the atmosphere, even after the TWR drops.

Pulling up with a low-TWR plane (low TWR in rocket mode) is a risky affair.

I generally prefer the "just keep flying" approach: try to maintain a pitch where airspeed increases at all times, still relying on the wings to keep me up. The faster I go, the higher I get, quite naturally -- with increasing speed I get closer to orbital velocity, requiring less lift to keep me up, allowing me to fly higher. Where I also have less drag and hence can accelerate faster.

All the "more speed, less lift required, more climb" stuff happens automatically if you just maintain an attitude where airspeed increases at all times. The plane needs to be be able to accelerate in level flight after the airbreathers have run out, though. Yours looks as if it should work. If it doesn't, try adding more wing.

[mk1980]

if the plane is desgined to fly to LKO with a few hundred m/s left i don't think it makes any sense to use nukes.

i'd cut the nukes altogether and use 1 aerospike instead. saves about 5 tons of engine mass (which you could simply replace with 5 tons more fuel) and gives you 50% more thrust when the jets flame out (nuke has 60 kN per engine, a single aerospike has 180).

when doing spaceplanes with decent amount of thrust, the most simple approach is to just aim for something like 10-15° climb and let them run hands off. no need for fancy ascent profiles - those are only really relevant for very tightly optimized configuations that have to go supersonic in level flight because they lack the thrust to accelerate beyond transsonic while climbing. you have 2 ramjets on a tiny plane, so you can just find a suitable climbing angle you don't have to change at all.

[AeroGav]

I have built many liquid fuel only SSTOs,   the problem with yours is that it has far too much fuselage drag and far too little lift (wing area).

With Whiplash engines i prefer to get my speedrun about 17km in level flight.   After that raise the nose gently so it's pointing just  5 degrees above prograde after starting the nukes.

With nukes you always have low TWR because they are heavy and weak.    Your craft can still gain energy so long as the goodness of your lift to drag ratio exceeds the badness of your TWR,  to use simple language.

Optimum lift drag ratio in high speed flight is when the nose is 5 degrees above prograde, so that is what you should aim for once the  nukes are on.

 

In KSP about 80% of the drag is from fuselage parts.  The ideal for a liquid, nuclear spaceplane is to put all your fuel in Big S wing parts,  but as that tech is not available to you, I'd say go with mk1 liquid fuel tanks instead.

Mk2 parts look nice and have good heat tolerance,  but they make twice as much drag as mk1 parts with the same fuel capacity.  They only really make sense for kerbal containing bits, since they give twice as many seats per part which offsets the higher drag.

 And for the love of god,  give it more wing area.   Like four or five times as much.

I put this together really quickly,  trying to base it off your design as much as poss.

https://www.dropbox.com/s/qpnu93jw84yhhgz/snouty.craft?dl=0

This ran out of fuel just before circularising ! 

So i redesigned it with 2 more mk1 liquid tanks.  Whilst i also converted the crew bits to mk1 size,  it may not have been the best move on account of how hot they all got.   

 

https://www.dropbox.com/s/01t94h8kja37fp8/snouty2.craft?dl=0

We made it though

 

 

 

[paul23]

I still question: why the nuclear engine?

[AeroGav]

34 minutes ago, paul23 said:

I still question: why the nuclear engine?

I did win a payload mass fraction challenge with a nuclear engine,  liquid fuel only ship.    However the things you have to optimise in a nuclear ssto are completely different to a high twr rocket one .     The nuclear ssto has a much lower fuel fraction ,  needs high lift and low drag.     The rocket ssto needs high fuel fraction and low dry mass.   So long as it can take off and get supersonic the rocket one doesn't have to worry about lift or drag too much,  it quickly blasts out of the atmosphere.

The retrograde orbit and the amount of payload - crew cabin for 6 seats  and a cargo bay -  make this a fairly demanding job for an ssto at this tech level.    Tough to do it without building an enormous craft whose fuel tanks/engines dwarf the actual payload.

[mk1980]

depends on the goal of the design. getting kerbals to LKO is trivial at this stage in the tech tree. the only pragmatic reason why you'd even want to use an SSTO  plane rather than a simple disposable rocket is to save some money.

to realistically save money, the plane must be reasonably efficient in both fuel spent and playtime spent. if the mission with the plane takes 20 minutes longer than with a rocket, it's hard to justify using the plane even if the plane costs close to 0 funds - simply because a disposable rocket would get the job done for maybe 10 or 20k currency and you can make a whole lot more than that in the 20 minutes you saved by doing some other trivial contract.

a reasonably fast plane with a pair of aerospikes and a pair of ramjets can hit orbit witin 5 minutes or so. landing takes a lot more time and effort than the reentry of a dumb capsule at 4x time warp of course, but if the design is somewhat stable aerodynamically and you figured out where you have to retroburn and what trajectory to aim for, you can also get a spaceplane down to the spacecenter within a few minutes.

 

of course none of that matters if you have more fun using planes or if planes are one of the goals / constraints you set for yourself in that specific campaign. having fun trumps pragmatic effiiciency considerations. it's a game after all.

[AeroGav]

On 11/1/2019 at 10:16 AM, mk1980 said:

depends on the goal of the design. getting kerbals to LKO is trivial at this stage in the tech tree. the only pragmatic reason why you'd even want to use an SSTO  plane rather than a simple disposable rocket is to save some money.

to realistically save money, the plane must be reasonably efficient in both fuel spent and playtime spent. if the mission with the plane takes 20 minutes longer than with a rocket, it's hard to justify using the plane even if the plane costs close to 0 funds - simply because a disposable rocket would get the job done for maybe 10 or 20k currency and you can make a whole lot more than that in the 20 minutes you saved by doing some other trivial contract.

a reasonably fast plane with a pair of aerospikes and a pair of ramjets can hit orbit witin 5 minutes or so. landing takes a lot more time and effort than the reentry of a dumb capsule at 4x time warp of course, but if the design is somewhat stable aerodynamically and you figured out where you have to retroburn and what trajectory to aim for, you can also get a spaceplane down to the spacecenter within a few minutes.

 

of course none of that matters if you have more fun using planes or if planes are one of the goals / constraints you set for yourself in that specific campaign. having fun trumps pragmatic effiiciency considerations. it's a game after all.

I don't think you're the first person to say my designs take too long to reach orbit,  so i had a go at making a more powerful version.

I must admit, I really like cruising through the upper atmosphere,  because it's the only time you get to see it (too high for airplanes, too low for satellites) so that's why i build them like i do.  I have a savegame from the point where my spaceplane is ready to stage in the NERV engines so i can skip straight to this part of the flight.   Having reached orbit with as much delta V as possible,  i then usually revert or just leave the guys floating up there forever, and go design another spaceplane.   Even though i got the delta V to go some place, i don't particularly enjoy that part of the game - plotting manuver nodes and waiting around.

Anyway,  instead of trying to do it with as little as possible (one Whiplash,  two LVN),  this has twice as many jet engines and one extra LV-N.  On my first attempt I got to MECO in ten minutes,  second time i screwed up and climbed too steep,  had to dive down again for the speedrun , which is why the screenshot shows MECO at 11 minutes.   That's about half the time of the original design.    Although the launch mass is higher due to the extra engines, it doesn't use much, or any,  more fuel to get into orbit.

Once again,  this is a RETROGRADE orbit.    Going East West like a sane person would leave a lot more delta v,  save time and not get so hot.

https://kerbalx.com/AeroGav/Snoopy-SSTO

In terms of the NERV or not to NERV debate,  I agree that at this particular tech level it makes less sense especially if you're only going to low orbit in an east-west direction.

NERVs really shine when you're doing Panther / NERV because the top speed of the Panther is low and the rocket mode burn to get to orbit is  a long one.    This tech level is important i think because whilst NERV take a few science points,  they can be done on a level 2 R&D building,  whereas the Whiplash/Aerospike node is tier 8 and needs the Max level R&D.   Getting the cash for that upgrade is the longest grind in career mode !

Paradoxically,  NERVs also work well at higher tech level when you've got BIg S wings to hold all your fuel in.     This gives a liquid fuel only design a big advantage in drag reduction over something that needs large LFO tanks.         For example,  check out my flying wing -

https://kerbalx.com/AeroGav/korten-fledermaus

I think this is more your kind of ship MK,   about 7 minutes to MECO - still plenty of Delta V left too !

[sebjf]

Thanks again for the great responses! 

I added more wing area, to start with, on my Mk 2 craft.

I built an Aerospike version. For this I need to use its high TWR to get out of the atmosphere quickly but I can get it into a (regular) orbit with about 550 m/s of dV to spare.

I added even more lift to the Nuke version, but with two engines it just can't accelerate quickly enough once in the upper atmosphere, before the plane falls back down (after which the drag becomes overpowering and it loses speed).

Looking at AeroGav's designs I think I am just using too many Mk2 parts.

Based on this thread I also created a Mk 1 plane that successfully achieves retrograde orbit with enough dV to rendezvous!

It took a long time, partly because I spent days trying to figure out why my efficiency dropped precipitously between two game sessions, only to realise that somehow at some point the cooler had become Mounted BACKWARDS!  I then had to figure out how to get it back to Kerbin without exploding... 

It may not be pretty, but finally I have a craft that can de-orbit my Kerbals!

 

On 10/29/2019 at 9:06 AM, Laie said:

I generally prefer the "just keep flying" approach: try to maintain a pitch where airspeed increases at all times

Interestingly, I still find the best ascent for the above plane is:

• Take off at 10-20 deg.
• Increase speed to ~1200 m/s by 10 km.
• At 10 km pull up gently (30 deg)
• At 20 km engage Nukes
• Continue at 30 deg until Alt. reaches 40 km
• Level out and push forwards until orbit is achieved

I tried the just fly faster approach, but I get into orbit with much less dV. I am not sure why.

In fact, I find using the jets to throw up the AP while they still can at high speed/low altitude the best ascent for all my planes. I wonder if there's a quirk or mistake I am repeating in all of them, since I can't seem to figure out how to make good use of the high wing area. I've never been able to push the PE high enough while in the atmosphere at all regardless of engine. (Actually, I have trouble getting to any speed that can present temperature problems with the Mk 2 craft - is this a sign that the craft is too draggy to make it?)

 

On 11/1/2019 at 10:16 AM, mk1980 said:

of course none of that matters if you have more fun using planes or if planes are one of the goals / constraints you set for yourself in that specific campaign. having fun trumps pragmatic effiiciency considerations. it's a game after all.

Yes I originally started this because I had Kerbals on ships without re-entry pods and it felt super wasteful to send up rockets just to de-orbit them. These planes have taken so much time, and have been frustrating, but once they started to work as Signo says there's so much to explore in different ways to tweak and fly them.