Interplanetary SSTO with nerv engines

[jsisidore]

Looks creepy, I'm trying to build something closer to real life craft, like nasa designs.

And I'm currently having a weird glitch with my new craft, for some reason it veers off to the left and then I have to juggle A D to keep it straight until the end of the runway. I checked everything I could, everything is straight, at least no less than the old design, the only major difference is than the torque now is twice as big, from ~7kNm it grew to ~14kNm, with 1 tonne of added weight. I don't understand.

[AeroGav]

39 minutes ago, jsisidore said:

Looks creepy, I'm trying to build something closer to real life craft, like nasa designs.

And I'm currently having a weird glitch with my new craft, for some reason it veers off to the left and then I have to juggle A D to keep it straight until the end of the runway. I checked everything I could, everything is straight, at least no less than the old design, the only major difference is than the torque now is twice as big, from ~7kNm it grew to ~14kNm, with 1 tonne of added weight. I don't understand.

Landing gear often attaches a degree or two out of alignment, especially when you're not putting it on the root part.

To ensure straightness 

1. attach landing gear to the component you want it to be on, and adjust it so it looks ok to the eye, then

2. ensure angle snap mode is on, rather than fine rotatation

3. press 3, to select "rotate part" mode

4. highlight the landing gear

5. press F to toggle between Absolute and Local rotation modes.  Ensure you are in Absolute mode, then..

6. rotate your gear by one notch in the left/right rotation axis, then snap it back to centre again.  do the same for the up / down axis and also for the roll axis.  This ensures everything is pointing dead ahead.

My designs , with their large wings, can take off at pretty low speed.  I don't like leaving the plane on the runway faster than 60 m/s because of the way the landing gear can get squirrely and also it makes it hard to get into a stable initial climb  - tends to zoom up into a very steep rise, then nearly stall and dive back down, then go back into another crazy climb.  If I lift off slightly below my subsonic climb/gliding speed that is less likely to happen.

[jsisidore]

Thanks. I did not know there is an absolute mode, it did help a little, now I can go all the way to 35m/s, but I need like 150m/s basically all the way to the end.

It might be a bug, or something inside cargo bay but nothing is sticking out and is symmetrical.

The only thing that changed was the torque, what is it?

How can I control it?

[jsisidore]

If I lower the throttle to 2/3 I can go further, but then at 60 I suddenly veer off to the right.

If to 1/3 then to 70.

Edited June 12, 2016 by jsisidore

[jsisidore]

It was the cargo, namely that part the scanner is attached to, it is not stock and must be bugged, sometimes you just can't win 

Edited June 12, 2016 by jsisidore
solved

[Val]

14 hours ago, jsisidore said:

And I'm currently having a weird glitch with my new craft, for some reason it veers off to the left and then I have to juggle A D to keep it straight until the end of the runway. I checked everything I could, everything is straight, at least no less than the old design, the only major difference is than the torque now is twice as big, from ~7kNm it grew to ~14kNm, with 1 tonne of added weight. I don't understand.

That kind of veering happens, when the main load bearing landing gear is too far behind CoM.

Just as having the CoL far from CoM is bad, having the landing gear far from CoM is also bad for stability.

[jsisidore]

Sadly after many attempts I was unable to land the craft since I was too heavy to achieve a vertical landing, i.e. I sank like a stone. I guess it is as hard as learning to drive in real life, brutal. Conclusion, rcs ports to maneuver the damn thing in Duna's atmosphere... or better yet a possibility for the engines to rotate e.g. Prometheus, Heliades.

[panzer1b]

Defenetely possible considering ive sent a SSTO to laythe, bop, and pol (landing too) and returned to kerbin with 0 refueling using some god awful monstrosity that was 40t pushed by a single rapier+nuke combo (1.0.5 game back when).  it had near 7000dV and considering i barely used any grav assists and those that i did use were only at jool, so yeah, possible byfar to get a nuke into LKO (rapier is efficient, turboramjet is quite doable but forget 7000dV after attaining a 100km circular orbit!).

Really teh trick with the new aero is to cut down on drag and manage heat effectively.  that means that nothing but absolutely essential things (wheels, wings, engines, fuselage) should be exposed to the air, all else should be inside a mk2 bomb bay or those tiny mk1 fuselage things that ive yet to manage to fit a weapon of any sorts into...  If you can do that and make the craft capable of ~1500m/s before switching to nuke from whatever jets you choose, you can and will get very high dV in LKO...

[AeroGav]

22 hours ago, jsisidore said:

Sadly after many attempts I was unable to land the craft since I was too heavy to achieve a vertical landing, i.e. I sank like a stone. I guess it is as hard as learning to drive in real life, brutal. Conclusion, rcs ports to maneuver the damn thing in Duna's atmosphere... or better yet a possibility for the engines to rotate e.g. Prometheus, Heliades.

I appreciate you'd like to try find your own way of doing things, but I think this does vindicate my approach to the problem - generous wing area + vernier lift engines.   

Raising the nose vertical, stopping in mid air then coming down in a rocket style tail sitter landing would require a higher TWR than nukes, so you'd be bringing oxidizer for the RAPIER - and lots of it because you'd be thrust borne for a relatively long time.  You'd need lots of reaction wheels or thrusters (drag penalty) because you'd be fighting the aerodynamics of the aircraft all the way trying to make it go tail first).

Alternatively you could make a VTOL  with a load of Spark engines, like Scott Manley tried a few times - massive drag penalty, quite a bit of weight and fuel required, still not easy to control.

Or just fill the cargo bay with radial mount chutes.   If you land on your gear you can withstand quite a high descent rate. Probably the best way if not using my method.

As far as my method goes, the extra wings have mass, which cuts into your delta V.    

BUT,  they also store liquid fuel  

>BUT  their mass : fuel stored ratio is not as good as dedicated tanks

>> BUT , BUT,  they actually allow the plane to fly more efficient on the NERV powered part of the ascent above 25km.   You can get sufficient lift with the nose only 5 degrees above prograde, and still climb at a decent rate and not overheat.

>>> BUT, BUT, BUT  lower in the atmosphere they create too much drag !  I can't go supersonic at sea level !

>>>>True but is it really such a chore to climb subsonic to 10km first?   What's an extra 5 minutes taking screenies of runway island when the whole mission including multiple attempts at a transfer orbit/getting an encounter will set you back a day?

 

 JSI,  as I see it you have 4 ways to increase the lift rating of your craft

1) Do what I do, and just tile Big S wings.  I'd be interested to see if you can make more natural looking shapes than i can

2) Use the tiny, lightweight mod Tweakscale to up size your exisiting wings.  You can simply enlarge them to 200% of original size, though it can look a little odd because thickness also increases.

3) Download B9 Aerospace Procedural Wings - Fork 

this is also a really small download and  like Tweakscale it can install automatically with CKAN.    Given that you've managed to create something so aesthetically pleasing despite being new to spaceplanes i'd like to see what you can do with this.

4) finally, you could install a 3rd party parts mod that simply adds new wing parts to choose one.

This is one route i've not gone, because mods that add extra parts are much bigger downloads, take up more memory, add to the game loading time and level of clutter in the part selection menus of the VAB/SPH.    It also takes a fair bit of downloading only to find there's nothing you really want out of what a mod adds.

BTW in my mk cockpit DUNA SSTO the vernier engines are in service bays rotated so that the open doors face downwards.    This means they contribute no extra drag when the doors are shut (that's probably being a little over the top tbh)

[jsisidore]

I can't stop but thinking how boring is ksp without mods. What do the devs thinking? 

[jsisidore]

After a not so short hiatus I decided to give it another go and finally was able to land the craft. It was tricky and far from realistic.

Deploying the survey satellite...


Surveying...

Awkward landing...


Deploying the harvesters... I'm not sure how many months it took me to refuel... but it certainly was less than a year


And this is how the screens would like like if I had a faster laptop.

Edited July 10, 2016 by jsisidore

[Firemetal]

It is possible. I have seen a few people do it post 1.0 (tho I haven't done it myself) so it is definitely possible.

[jsisidore]

On 7/11/2016 at 2:17 AM, Firemetal said:

It is possible. I have seen a few people do it post 1.0 (tho I haven't done it myself) so it is definitely possible.

You should read the whole topic if you want to try it yourself, it is full of all kinds of practical and theoretical information to help you get there

[Firemetal]

3 hours ago, jsisidore said:

You should read the whole topic if you want to try it yourself, it is full of all kinds of practical and theoretical information to help you get there

I believe it wouldn't be very fuel efficient so I think I am sticking to rocket mode rapiers thanks!

[Edax]

On 6/7/2016 at 7:56 AM, Val said:

Yes, it is possible to make LF-only designs. Quite doable with Nerv/Whiplash and Nerv/RAPIER(Airbreathing only) comboes.

But in practice it's more efficient (dV vs. total craft mass) to include a little oxidizer for the RAPIERs.

A well designed craft around 36t with 2 air-breathers (Whiplash/RAPIER) and 2 Nervs should be able to get to orbit with at least a couple km dV or more depending on payload.

Barring issues with exploding landing gear, these 2 examples should still be valid for 1.1.

The first one is capable of getting to orbit even if you remove the oxidizer.

 

https://kerbalx.com/crafts/8018/

 

https://kerbalx.com/crafts/9737/

I haven't made an SSTO in a long while, mostly because the landing gears all explode, meaning all the planes I have an orbit can't land, and all the designs I spend weeks designing and testing can't take off.  I've never really bothered finding out why the landing gear has suddenly becoming a big problem, I used that as an excuse to take a break from KSP...but what is the problem with landing gear now?  I know it had to do with the big overhaul patch and I haven't built a plane since that could successfully take off and land without damage.

[Zamolxes77]

On 6/7/2016 at 2:23 PM, jsisidore said:

Can you reach ~70km using nerv engines?

edit:

Sorry. Let me reiterate myself more clearly, since I thought it was obvious that you have to use air breathing engines just to get off the ground.
Is it possible for nerv engines to give you enough thrust to leave the atmosphere past 20ish km to reach 70km orbit?

It is, however its extremely inefficient. During atmospheric flight, atomic engines get a penalty, I think they function at 33% efficiency or something like that, so instead of 60kN of thrust you get 15. Furthermore, you craft has to be very light, small, in order for the Nerv to have a chance to push your craft from 24 km to 70+. If for example you get a TWR or .64 or lower during this phase, its very likely your plane will crash (run out of fuel) instead of making orbit.

A much better option is to use rapiers and Nerv's on you design, while carrying a limited amount of oxidizer. Around 24 km, switch rapiers from air breathing mode to closed cycle and they should run out of oxidizer when you reach your target Apo, outside the athmosphere naturally. Then you coast and circularize using your Nerv.

Here is a micro SSTO, that uses the concept I stated above: 1 rapier, 1 Nerv, and enough dV to take off, go to Minmus, land and return to KSC, with 2 Kerbals.

This is my shuttle craft and rescue vehicle:

Using only rapiers its easily capable getting into orbit and still have a large amount of dV left for orbital ops, around 1400 dV or so. On the plus side, it has a docking port, so it can refuel in orbit and be used for interplanetary missions.

This following one is a concept I was working on, trying to determine what's more efficient: rapiers with Nerv's or Whiplashes with Aerospike. Whiplashes beat rapiers in athmo flight , Isp wise and aerospikes beat rapiers in weight, thrust and Isp in Vac. They're also one of the most efficient rocket engine in ASL flight.

Sadly I used a completely different plane for my experiment, so I'm stonewalled. If I had to draw a line though, 2nd design is best: higher payload, in form of crew cabin and nose with more dV left in LKO.

[Foxster]

1 hour ago, Zamolxes77 said:

It is, however its extremely inefficient. During atmospheric flight, atomic engines get a penalty, I think they function at 33% efficiency or something like that, so instead of 60kN of thrust you get 15.

Sorta. 

At sea level they are indeed inefficient, giving around 14kN. 

But get into thin air and they are fine. At 12km they make 54.3kN and by 20km they are making 58.8kN or most of their max thrust. 

This is why they work fine on a craft that uses other engines to get to the limit of air-breathing before handing over to nukes. 

[DrunkenKerbalnaut]

On June 12, 2016 at 2:26 PM, jsisidore said:

...weird glitch with my new craft, for some reason it veers off to the left and then I have to juggle A D to keep it straight until the end of the runway...

If anyone happens to reach this conversation by way of googling this particular issue, ensure that the rearmost landing gear (assuming tricycle layout) of your plane have their steering disabled. 

[Kerbal101]

On 08.06.2016 at 3:31 PM, AeroGav said:

 

Here's a NERV/Whiplash ship flying from 18km to orbit.   It carries 5 crew and is pretty loaded with extras like docking port, lifting engines etc.

You can see that the NERV easily produce enough power though those airliner wings aren't really suited to this kind of thing and get pretty hot.  It actually goes all the way to Minmus and back - although the fuel reserves look mostly depleted by the end of the video, the gauges only show what's in the tanks the NERV directly connect to.   On part 3 you can https://www.youtube.com/watch?v=3nI-n0J262I

 see the remaining fuel get transferred in for the interplanetary leg.   And yes, there is 55 units of oxidizer on this thing.   It's for the Vernier engines only , which enable horizontal touchdown like this - https://www.youtube.com/watch?v=w6ZxVtZswh0

.....

 

^^^

Brilliant Video, thank you!!!

The Rapier+NERV designs are easy, nothing fancy. Rapier can generate plenty of thrust in open cycle to push Apoapsis high outside of atmosphere; then burning NERV prograde eventually stretches Periapsis. No big deal.

But Whiplash+NERV are completely different. Whiplash thrust cuts earlier beyond speeds needed for Apoapsis. So one ends with 45-60km Apoapsis tops, and the path is too short for NERV to stretch Periapsis.
So you solved this with Periapsis stretching in prograde, when still high in atmosphere.... consumes lots of fuel and overheat, but - brilliant!!

[AeroGav]

51 minutes ago, Kerbal101 said:

 

^^^

Brilliant Video, thank you!!!

The Rapier+NERV designs are easy, nothing fancy. Rapier can generate plenty of thrust in open cycle to push Apoapsis high outside of atmosphere; then burning NERV prograde eventually stretches Periapsis. No big deal.

But Whiplash+NERV are completely different. Whiplash thrust cuts earlier beyond speeds needed for Apoapsis. So one ends with 45-60km Apoapsis tops, and the path is too short for NERV to stretch Periapsis.
So you solved this with Periapsis stretching in prograde, when still high in atmosphere.... consumes lots of fuel and overheat, but - brilliant!!

There's two approaches really -

• zoom climb on jet power. Get sub orbital, then try to circularise on NERV before falling back.  The NERV burn is conducted out of the atmosphere so does not make use of lift.
• Get as fast as possible in horizontal flight, then pitch for best lift/drag ratio, start the NERVs and just cruise up to orbit like an airliner. As lift fades, orbital freefall effect takes over.

Ships in this weight class typically have 3 engines, which forces you to choose between 2 jets and 1 nerv or 1 jet and 2 nervs.   If you went with 2 jets, you probably don't have thrust > drag in the upper atmosphere on one NERV, so you need to use the zoom climb method.  If you went with 2 nervs,  you don't have the jet TWR to get a good zoom climb going.  Zooming out of the atmosphere just deprives you of lift and makes you fall back.

Which is easier? If you started off flying rockets, probably the zoom method, because it's all about thrust and momentum and not dissimilar to some rocket launches.     If you're a flight sim fanatic, the "cruise method" feels more familiar.

This is a more recent example of the cruiser method.  37 ton takeoff weight,  big s wings (more heat tolerant, but still mk1 fuselage),  with a  rapier instead of whiplash  (i did also add two junos as it couldn't take off from water without a little bit of extra power, but  i don't think they count ).    I wouldn't say it uses "loads of fuel",  I think we lost about 1000dV  accelerating from 1400 @ 22km  to 2200 @ 70km (orbit), which isn't a big loss at all.  Once in orbit, having 2 nukes instead  of 1 means higher TWR and lower losses when making transfer burns,  you can also make thrust-supported vertical landings more places.

 

ps - this flight profile does have a stronger zoom climb aspect to it - by accident really as the airplane had a tendency to weave up and down when left on Prograde hold -

 

[Kerbal101]

@AeroGav Thanks for posting the videos and explaining this in very detail!! But you use Rapier in these two spaceplanes, which make problem lesser due to their TWR. I am really interested in Whiplash or even Panter/Wet... more basic technology and stuff..

I figured out that main problem with SSTO is again, TWR of each two stages...
For example, airbreathing stage of less  than 0.7 won't even manage to break sound barrier.. And 1.5 is needed to for efficient breathing climb.
The same thing hits second stage nerv, Mk2-based SSTO doesn't even manage to circularize if Nerv TWR is less than 0.8.

I am designing a mk2 ssto with Whiplash/Nerv currently, and best profile found so far was 1100m/s burn under 2km, then quite energetic 45 degree up with Nerv up at 7.5km...
Managed to grab AP of 85km, but still can't break the 1700m/s horizontal speed before falling back into atmosphere. I assume I need more Nerv less fuel or payload...

But having lots of fun

Edit: extra thanks for the second video, I gotta try that out myself... cant break through the 1300m/s barrier in the 18-20km, I thought diving down is a no-go, apparently its not.

Edited February 26, 2017 by Kerbal101

[ForScience6686]

I don't see a panther nerv combo working.  You don't have the power to get fast enough in the atmosphere.  Whiplash will actually take you faster than rapier in jet mode, so I'd start there.

[Kerbal101]

39 minutes ago, ForScience6686 said:

I don't see a panther nerv combo working.  You don't have the power to get fast enough in the atmosphere.  Whiplash will actually take you faster than rapier in jet mode, so I'd start there.

Whiplash and panther problem is not thrust, but thrust barrier. Once aircraft is moving at higher speed or altitude, they can't keep up and loose thrust. Rapier gives more thrust than whiplash at specific speeds, and this allows it to push out of atmosphere much easier.

[AeroGav]

10 hours ago, ForScience6686 said:

I don't see a panther nerv combo working.  You don't have the power to get fast enough in the atmosphere.  Whiplash will actually take you faster than rapier in jet mode, so I'd start there.

It's not as efficient as Rapier / NERV or Whiplash / NERV.          But it can be done -

https://kerbalx.com/AeroGav/Voodoo-Ray

Also - https://kerbalx.com/AeroGav/Firefox-I

The simple fact  is that orbital velocity is 2250 m/s.

Panther gets you to about 750 m/s,  but  thrust falls to zero very quickly when you start going faster. Your rocket stage needs to be capable of adding +1500 velocity, plus whatever losses.

Whiplash peaks at 900 m/s, but hangs on to a reasonable amount of power till 1250.   This reduces the amount of rocket delta V needed to make up to orbital velocity. 

Rapier is obviously best,  as it can give you 1600 m/s in air breathing.   That  means you need less than half the rocket mode delta v compared with Panther.

I was surprised at how much better the Panther/NERV worked than Panther/Terrier tbh.      The greater fuel efficiency of the NERV engines helps with the long rocket mode burn needed to add +1500 m/s speed after the jets flame out.   The Panther engine is relatively light which offsets the mass of the NERV, and it has good performance at low speed, helps with getting supersonic.

If the RAPIER is available to you, there's only two reasons to pick something else -

1. In a large ship with marginal jet TWR that struggles to get supersonic, you can get supersonic for less dry mass by adding Whiplashes instead of extra rapiers.   Once supersonic, they need no help.   Note, every little helps here and i toggle my nukes on, and go into a shallow dive, across the transonic region.
2. Panthers and Whiplash are very cheap.   Funds cost of 2000 vs 6000 for the rapier.   This means you can decouple them with a clear conscience when they flame out, which greatly reduces the mass that the nerv stage has to lift.

Building a liquid fuel only ship really helps you to learn your design and flying skills , but  designs with a small amount of oxidizer still perform slightly better, all things being equal.   Why?

You need to learn about gravity.

Gravity is pulling you down.   Orbital freefall effect, or if you prefer the other explanation ,  the centrifugal force of hurtling around a planet, cancels out some of the gravity.  The closer you get to orbital velocity, the more gravity this effect cancels out.    However the remaining gravity must be cancelled out by lift.

Once you get to mach 5.3 or so (1700m/s) orbital freefall is taking care of 2/3 of your plane's weight.   You don't need much lift to maintain level flight.   This means you can fly really, really high where the air is very thin and drag is less, while still getting enough lift.   You can still get enough lift even though you're not raising the nose more than 5 degrees above prograde, for low drag.       This means you need fewer NERVs to complete the task, which is a good thing because NERVs are very heavy.

So,  have a look at this partially re-usable space plane i built for the shuttle challenge.   Two Whiplash, they get us to about 1000m/s , then i bring the NERVs online.   We max out at 1300, then i start two Terriers.   At 1380 the Whiplash flame out, I decouple them, making us 4 tons lighter..   At 1800 ish, the Terriers run out of oxidizer , they get decoupled too.  This only saves 1 ton, but at 450 funds each , you might as well just drop 'em.    The oxidizer tanks are worth quite a bit of money though even when empty and don't weigh much when empty, so i hang on to those.     The twin nukes are able to boost us from that point on to our final  650km orbit, where the 6 ton payload is deployed.

ps. do not build an aircraft to look like a shuttle unless doing so for a challenge.  They have the aerodynamics of a brick !

[AeroGav]

11 hours ago, Kerbal101 said:

@AeroGav Thanks for posting the videos and explaining this in very detail!! But you use Rapier in these two spaceplanes, which make problem lesser due to their TWR. I am really interested in Whiplash or even Panter/Wet... more basic technology and stuff..

I figured out that main problem with SSTO is again, TWR of each two stages...

 

Most of my designs the TWR fluctuates in flight between 0.33 and 0.6.    0.33 when the rapier is subsonic,  gets to 0.6 or higher when the rapier is in its sweet spot.     The NERV stage starts with 0.33 TWR and this gets higher as fuel burns off.      In real money, that's about one rapier per 30 tons of takeoff weight, and one nuke per 20 tons.

Quote

I am designing a mk2 ssto

That is the problem !   MK2 fuselages have four times the drag of mk1.    My flight profile requires low drag and lots of lift.

The mk2 ssto in my video has the main wings angled up 3 degrees (https://en.wikipedia.org/wiki/Angle_of_incidence_(aerodynamics)),  as are all the strakes etc.       The canards are angled up a couple of degrees more than that, so they stall first.   This enables us to fly on prograde, with the fuselage at basically zero angle of attack.    The downside is that you're giving up control of your vertical flight path by just locking prograde.   Which is why you see it zooming and diving.  If i take it off prograde to steer,  i loose loads of speed.    

Also making a plane with angled wings balance right is a lot harder than straight wings.   I'd say start out with a mk1, you don't need to bother with such tricks.  Move on to mk2 when you've got mk1 cracked !

  

Quote

For example, airbreathing stage of less  than 0.7 won't even manage to break sound barrier.. And 1.5 is needed to for efficient breathing climb.
The same thing hits second stage nerv, Mk2-based SSTO doesn't even manage to circularize if Nerv TWR is less than 0.8

To quote from a post on Airliners.Net

Hi everybody, 

A question that has been playing around my cranium for a while. And since I signed up today I can finally ask some real experts; 

When looking at the thrust to weight ratio of Concorde: 

MTOW: 408 000lbs 
Total thrust: 128 000lbf 
Total thrust: 152 000lbf (with A/B) 

Thrust/Weight: 0,314 
Thrust/Weight: 0,373 (with A/B) 

The drag of your mk2 fuselage is probably hindering you.

If you watch this video form the beginning, you'll see me cross the sound barrier.

Note that i 

1. climbed up to higher altitude, where drag is less.

2. kept my speed below 240 until it was time to cross the sound barrier, since 240 to 440 is the unusually high drag zone.

3. kept the AoA low.   When you can't get any higher without either exceeding 240 m/s or pulling the nose much more than 5 degrees above prograde, it's time to transition supersonic.

4. When I'd decided to go supersonic, i reduced the nose up trim, halving my AoA , lift and drag.  We gently nose over into a shallow dive to go through the sound barrier.  Once over 400, we gently pull up (the plane will tend to do that by itself as higher speed and thicker air means more lift).

5.  something i did not do, because it wasn't needed - i could have toggled the nukes on between 240 and 440, to get us through the rough bit.

Quote

Edit: extra thanks for the second video, I gotta try that out myself... cant break through the 1300m/s barrier in the 18-20km

1300 is very fast for a Whiplash.

Mach 3 is peak.   That's about 900m/s at altitude.    After that power falls away quite quickly.    Each Whiplash is 1.8 tons the nerv stage must carry to orbit as a payload, unless you decouple it.  Fortunately Whiplash engines are quite cheap !

The "knee" in the atmosphere curve corresponds to about 17km.   Above 17km power falls away very quickly.