Linear Engine SSTO

[recursive_mouse]

Want to check how many delta vs I could squeeze from a Practical, Conventional, Non-Ion SSTO in stock game, thus modified one of my 1.31 multi-purpose side-engined lifter design to stock 1.43, maximal takeoff weight ~385 tons, hit 100km LKO with 8205 m/s delta v remaining (or 8200 m/s if nose fairing come back in one piece ). 

Although utilizing linear engine construction, dimensional fuel tank, multi-layer wings, and autostrut-everypiece kinda like using k-drive, this stock design is pretty convetional by Kerbal means... Anyway, for a 38.5 tons per RAPIER design to easily takeoff / landing from KSC runway,  and could theoretically ship 126 tons payload to rendevous @ 60km Mun orbit, with ~1000-2000 delta v remaining for return trip, it's satisfyling.  ( Subtracting 22,400 liquid fuel. Cargo space is limit here unless go dimensional with payload )

 

craft.zip

 

Side-mounted engine pair benefit from CoM moving foward the more liquid fuel depletes, with 4 Nerv engines no age-long ion burn neccessary. As Practical, it comes with pod for 3 Kerbals, fornt docking, lots torque wheels, 4 RTGs, 10 vernier ports. Agile and structutraly sound. Also with probe core + big antenna for non-pilot Kerbonauts. ( By eliminates the 'Practical' Burdens and a pair of wing it hits 70 LKO almost 9000 Nerv-dv remaining )

With ridiculous L/D ratio from linear engine, this brick could glide Kerbin indefinitely... Had to open rear ramp to proper descending. Another problem is it had to make orbit in one attempt, as any save-load while in dense atmosphere will surely break the wings are generating huge lift... When higher than ~25kms, or landed at surface, save/load works properly. 

 

 

The uglies minimal version set 8861 manned / 8896 unmanned @ 100 LKO... I built everything around central monopellent tank, if this useless 1400kg dead weight could be removed... and improve ascent profile (had too conservative because save/load broken in lower atomosphere), I'm sure it'll break 9000 barrier...

Edited June 5, 2018 by recursive_mouse

[recursive_mouse]

Turns out very difficult to adapt KSS Phallic for SSTEO.  a redesign use all Mk2 structure from start, and... with *mild* offsetting its a easy success.

 

This all-stock-parts prototype Eve shuttle obtain 90-140 Eve orbit stupidly easy, often with 100-200 dv excess. I'll start testing on stock game see if nothing change. 

 

Another Kerbal test pilot live to see the day  

 

Even discarded docking port / ISRU / proper pod, I could only squeeze 5523 dv off KS-25 alone; as inexprienced as I am, this is insuffcient to obtain Eve orbit. Thus the final version have pair of Nerv attached, while retaining all the functionalities.

With this setup the spacecraft ascend simply pointing constant 25 above horizon, and completely avoided destroying the fairing nose in such fashion. At 42km alt gently flatten trajectory to 15 degree and continue burn with Nervs to 140km AP and 3 minutes grace time. Then imidiately aim for circulation, which takes ~6 minutes to complete. The remaining liquid fuel then burned up to lower eccentricity( for easy rendevous with leftover ).  The real challenge though, is to get off from the mountain without sliding off, tilting over, immediate stall and disintegration, or simply forgot retract gears and close cargo bay.  

Now I really consider offset is cheating (even without any shift or clipping). 

 

 

Edited June 7, 2018 by recursive_mouse

[Rune]

21 hours ago, recursive_mouse said:

Turns out very difficult to adapt KSS Phallic for SSTEO.  a redesign use all Mk2 structure from start, and... with *mild* offsetting its a easy success.

This all-stock-parts prototype Eve shuttle obtain 90-140 Eve orbit stupidly easy, often with 100-200 dv excess. I'll start testing on stock game see if nothing change.

Another Kerbal test pilot live to see the day 

Even discarded docking port / ISRU / proper pod, I could only squeeze 5523 dv off KS-25 alone; as inexprienced as I am, this is insuffcient to obtain Eve orbit. Thus the final version have pair of Nerv attached, while retaining all the functionalities.

With this setup the spacecraft ascend simply pointing constant 25 above horizon, and completely avoided destroying the fairing nose in such fashion. At 42km alt gently flatten trajectory to 15 degree and continue burn with Nervs to 140km AP and 3 minutes grace time. Then imidiately aim for circulation, which takes ~6 minutes to complete. The remaining liquid fuel then burned up to lower eccentricity( for easy rendevous with leftover ).  The real challenge though, is to get off from the mountain without sliding off, tilting over, immediate stall and disintegration, or simply forgot retract gears and close cargo bay.  

Now I really consider offset is cheating (even without any shift or clipping). 

 

 

 

Is that really a stock Eve SSTO? One that can take off with less than 1 TWR on Eve? If so, color me impressed. It's tiny for that!

 

Rune. And the dV budget looks... beyond tight.

[recursive_mouse]

2 hours ago, Rune said:

Is that really a stock Eve SSTO? One that can take off with less than 1 TWR on Eve? If so, color me impressed. It's tiny for that!

 

Rune. And the dV budget looks... beyond tight.

Turns out structral integrity benefits a little from KJR... after double the hidden struts, now it could really park in 7500m, & takeoff in stock game, but dv fell short to 4974+630, which is just between can do or impossible... I'm yet to archieve orbit. 

The entire weight exclude propulsion been minimized to < 12 t, totally excluded possibility for Mk3 parts. Which they are 20 times stronger than Mk2 counterpart. 

In 1.43 offset is more fragile( looks like the further the offset weaker the joints ), the 'dimensional' zip fuel have to packed exact same place to minimize torque. The exploding of 424 ton flammable is spectacular.  

Edited June 8, 2018 by recursive_mouse

[recursive_mouse]

Test Mission Report: Prototype SSTEO, Sucess (Stock 1.43)

Game environment: stock 1.4.3.2152, only KER & TimeControl installed, and only KER is used for reference. The save was copied from a modded install (for put the plane on Eve of course), but loaded in stock everything behave totally stock, i.e. fragile and wacky. 

 

The spaceplane still needs work to be user-friendly, but nontheless it's 100% capable of SSTEO. 

 

Mission:

T+0, parked at 7482m height, Eve.

It's a feat itself to stop the plane from sliding off,  1. open all the cargo bays for additional drag, 2. pointing sideway to stall the wheels, 3. finetune. 

Parking atop 7500m sierra is easier, but little short for take off. 

Anyway, close bay doors, Vector engine full throttle, steering to the mountain top.

 

The following is most crucial. The landing gear is finetuned to minimize dry weight. It's about *just right* to stand Eve gravity, so take off from exact heading is needed. Aim for slight right off mountain apex, and no control input, certainly avoid 'S' key before off ground. Otherwise firework show guaranteed. 

 

Without pitch input, the plane glide off ground at stunning 77m/s, climbing and accelerating with 0.62 TWR. The test pilot, Hanrick Kermann, don't unterstand this phenomenon. 

He then checked plane is stable in air; and pointed nose at 20 degrees above horizon. 

 

at about ~9500m altitude, Hanrick pointed 25 degrees above horizon, checked full throttle, all the way until near 40km.

 

From 35km, Hanrick is prepared to flatten the trajectory,  aerodynamic force from this altitude is so strong, very careful with KS-25 vectoring.

The adiabatic heating of 2600k-withstanding nose faring is best reference for now, near thermal limit -- lose efficiency due to drag; too cool -- not enough horizontal velocity. 

The plane should be point to 20 degree when exiting 41~42km, and keeping pulling speed vector down, until speed vector ~14 degree, nose pointing 10 degree. 

 

Once KS-25 flameout, should have 3 minutes grace time to 118-122km AP.

Imediately, activate Nerv engine, Quick plot a manuver at AP for reference. Pointing to prograde, pointing to horizon, whence at ~75km the drag is low enough to safely point down to manuver node. 

 

Examine the manuver node. After actual AP, it should boost PE to 93km within 3 minutes, 30 seconds . Will Hanrick Kerman survive the test flight? 

 

 

And voila. 136 dv remaining. (in 'Slightly Modded Space Programm' aka the SMSP simulation,  the safe and optimal result is 176 dv remaining means 5,433 m/s SSTEO  )

 

Hanrick imagining the future Eve scout mission, The spaceplane will take off Kerbin with half internal fuel, allow it to take 100 tons zip-liquid fuel tanker, left tanker at 120 LEO for return trip. It'll be relies on Nerv for spacetravel. Also relies on years of ISRU atop Sierra Evas, or double or tripple the ISRU time if ESRM - Eve surface relocation manuver is needed.

 

Nontheless it will 100% percent self-sufficient, the entire size land on Eve surface, and every part include nose faring return to Kerbin.  Maybe even some 50kg payload off Eve surface!

It might be practical. If the final version spaceplane is less rear-heavy when empty on fuel.  Right now it takes a combination of tricks to safely land the glider brick on Eve surface.  Some trick, like cargo bay aerobrake, is too effective to control by a Kerbal. 

 

Right now, he just...  stranded at LEO, waiting for rescue, like any lucky surviving Kerbal test pilot. Fortunate for him, there's proper-spacious Mk.I command pod, not space chair inside a Kraken Co. service bay. 

 

This landing gear setup is strangely effective to support near 800 tons of gravity force, front medium attached to Mk2 long adapter, set to rigid ,slightly more spring rate and dampen.

Rear gears aligned slightly after CoM. Inner pair of large gear attached to cargo bay, set to rigid and slightly higher spring and damping. Outer pair attached to wing , non-rigid, normal spring and damping. 

 

Edited June 8, 2018 by recursive_mouse

[recursive_mouse]

The .craft and savegame of said prototype SSTEO. 

 

.craft file

sandbox savegane

 

Unfortunately to shield struts from drag, wings have to offset into cargo bay, and with huge incidence bay door clipping is unavoidable. 

Edited June 8, 2018 by recursive_mouse

[recursive_mouse]

CORRECTION: miscalculated bipropellent shell weight. *Zero* structure gives 1867-500=1367, not anywhere off Eve. A single piece is mostly impossible. 

The most I could squeeze from my plane is 4904 rocket, 1101 nuclear. Even switch 600dv-equivalent weight to ion will not make difference. 

 

"This prototype shows howewer, Nervs could be added without sacrifice SSTEO capability. 

In 1.18 Rocket TWR, 0.12 Nerv TWR fashion, it takes ~4970 rocket dv, 500 Nerv dv to Eve orbit. 

Bipropellent ratio is fixed, 80% from fuel and 10% shell. Engine weight ratio also fixed, as 5.77% for said TWR requirement. If entire structure include ISRU constitute < 1.61% of gross weight and still provide substantial property(Mk3 cargo bay, huge landing gear,Big S wing, both do wonders in weight-efficiency term), then it becomes a Single Piece Kerbin-Eve shuttle,  i.e. a SSTEO by every possible definition, totally stock. With 1,600 m/s dv left at Eve orbit. Whithout any needs for rendevous and separation. 

Right now, it's 2.3%, indeed very close. But unfortounately can't do anything more radical with weak, Mk2 parts.  Maybe *slightly* go larger and add ion engine perhaps? "

 

Edit again: Modified from the successful prototype, this is absolute limit I could get:  launch from Eve surface, 1425~1450m/s dv left at 90-120 Eve orbit. in form of 0.01 TWR only works when sunlit.

I'm confident not a single gram is wasted on this craft, and still she's shy of getting anywhere else than, Eve. And, although it retains good stall performance,  sincerely don't think ascending efficiency could match the prototype. Even takeoff became harder. 

 

Conclusion, (even with ISRU) the ultimate SSTEO without any separation or rendevous, is straight impossible in stock game. 

 

 

One the previous prototype, I made mistake adding liquid fuel tank while leave big s wings empty. With remedy now get 103 more dv from Nerv! 

 

Edited June 9, 2018 by recursive_mouse

[Rune]

You just proved something I've always had a gut feeling should also happen in stock KSP now that we have a more realistic aerodynamic model. Under certain conditions, aerodinamic lift can be used to offset gravity losses and make orbit with TWR under one, at an advantage in mass ratio over straight up rockets (you manage to carry stuff like nukes at an amazingly low takeoff mass, compsred to other SSTEO I've seen). At least in RL. The thing is, you had to abuse docking bay occlusion to minimize drag. Not that I have anything against that, of course. In fact, I'd tell you to explore the wonders of stock autostrut. For parallel connections it's a bit explodey at times, but on linear ones it's rock solid. And the new 'rigid connection' setting should actually do just what KJR does, right? You activate those through the advanced tweakables option in the settings menu, BTW, just in case they slipped by you. Should shave some kgs on struts.

 

 

Rune. Oh, and it goes without saying... Amazing cratf dude!

Edited June 9, 2018 by Rune

[recursive_mouse]

59 minutes ago, Rune said:

You just proved something I've always had a gut feeling should also happen in stock KSP now that we have a more realistic aerodynamic model. Under certain conditions, aerodinamic lift can be used to offset gravity losses and make orbit with TWR under one, at an advantage in mass ratio over straight up rockets (you manage to carry stuff like nukes at an amazingly low takeoff mass, compsred to other SSTEO I've seen). At least in RL. The thing is, you had to abuse docking bay occlusion to minimize drag. Not that I have anything against that, of course. In fact, I'd tell you to explore the wonders of stock autostrut. For parallel connections it's a bit explodey at times, but on linear ones it's rock solid. And the new 'rigid connection' setting should actually do just what KJR does, right? You activate those through the advanced tweakables option in the settings menu, BTW, just in case they slipped by you. Should shave some kgs on struts.

 

 

Rune. Oh, and it goes without saying... Amazing cratf dude!

Thanks! In fact, several autostruts and rigid configuration have been experimented... like the rear landing gears, set inner pair to rigid and oouter pair to non-rigid, somehow cancelled overstress wobbling. And is the only way in stock game I could reliably parking, takeoff, move the craft to 7482m launch point (after final 7500m ISRU operation), and do a pendulum brake to brief hold. 

 

Right now, cockpit optimized to consist only sky-facing kerbal( inside nose faring ), a minimal jr. port to set direction, a wheel to enhance control and absorb heat.

In theory it could set Eve orbit and left with 1425dv(1dv per minute), but is so difficult to achieve. Overall chance will be sub zero. 

I guess it shows: even exploited all 'legitimate' extreme, build a SS Eve-Kerbin is still impossible. Infact it's easier to build a legitimate perpetual motor than convetional SSEK, as there are plenty with Kerbals.  

 

 

Edited June 9, 2018 by recursive_mouse

[Rune]

34 minutes ago, recursive_mouse said:

I guess it shows: even exploiting all 'legitimate' extreme, building a SS Eve-Kerbin is still impossible. Infact it's easier to build a legitimate perpetual motor than SSEK, as there are plenty with Kerbals. 

Well, yeah, but having wings (and therefore a potential high-drag precision landing) is one step closer to a practical Eve reusable architecture, and with some scaling left to go yet. There are bigger cargo bays and tanks, after all. So what if you need a LEO depot/transfer station supplied from Gilly to close the loop? Those are just more cool missions to fly, and payloads to design.

 

Edit: Gut feeling also says the trajectory could be optimized a bit. That Nerv should be burning as soon as its ISP is higer than the Vector, in order to increase TWR and minimize gravity losses when the wings stop working as well as they did at lower altitudes. Should. In theory. But it's an informed opinion.

 

Rune. I'm getting the itch to reverse-engineer something a tad bigger from those nice numbers you gracefully gave us.

Edited June 9, 2018 by Rune

[recursive_mouse]

45 minutes ago, Rune said:

Well, yeah, but having wings (and therefore a potential high-drag precision landing) is one step closer to a practical Eve reusable architecture, and with some scaling left to go yet. There are bigger cargo bays and tanks, after all. So what if you need a LEO depot/transfer station supplied from Gilly to close the loop? Those are just more cool missions to fly, and payloads to design.

 

Rune. I'm getting the itch to reverse-engineer something a tad bigger.

Well the formula is simple, for similar L/D/Weight ratio, rocket part have to be at least 94.6% for KS-25 or 94.3% for Mamoth( potential improvement! ), to left enough orbit window for more efficient low thrust engines. If low thrust engines onboard, then don't need full tanks except Eve takeoff, with potential huge dv/payload cap for daily. 

Fuselage weight and L/D ratio is variable here. And practical single-piece SSEK should keep a minimal potion of Nerv-level thrust for Eve ejection. 

It will not be as useful as SS left orbiter / drop lander, but fly strange wacky planes to a seemingly impossible mission is fun and thrilling 

 

Edit: and because Mk3 is ~20 times stronger, 6~8 times heavier, and quite big drag cube and payload space, the optimal weight for Mk3 SSTEO would be huge, aim for more than 5000t / at least 15 Mammoths

Edited June 9, 2018 by recursive_mouse

[recursive_mouse]

On 6/9/2018 at 8:08 PM, Rune said:

Well, yeah, but having wings (and therefore a potential high-drag precision landing) is one step closer to a practical Eve reusable architecture, and with some scaling left to go yet. There are bigger cargo bays and tanks, after all. So what if you need a LEO depot/transfer station supplied from Gilly to close the loop? Those are just more cool missions to fly, and payloads to design.

 

Edit: Gut feeling also says the trajectory could be optimized a bit. That Nerv should be burning as soon as its ISP is higer than the Vector, in order to increase TWR and minimize gravity losses when the wings stop working as well as they did at lower altitudes. Should. In theory. But it's an informed opinion.

 

Rune. I'm getting the itch to reverse-engineer something a tad bigger from those nice numbers you gracefully gave us.

As your suggestion, by burn low-thrust engines early (and sadly ditched ISRU) now get ~47 Nerv and ~1475 Ion delta-vs after Eve orbit. Quite a improvement and still lot of potential . You could tell by  AP @ 150km+  that I messed up quite a bit.

Should be capable picking up 150 LV-N, 1500 Ion total at 90-110 LEO,  capable of transfer one patient Kerbal for Kerbin aerocapture.

 

A minimal KJR setting still somehow reinforce joints, and is of utmost importance during take off (L/D/Weight now focusing on high-speed / stability). But that could be easily remedied for stock game. 

 

Without ISRU, that's technically a SSE->K or SSK->E only (possible but not very practical though). Straight roundtrip is still out of question. 

 

Edited June 11, 2018 by recursive_mouse

[recursive_mouse]

Something learned for daily KSP, wonder someone already noticed, a foward facing buttspike design is far more superior at dissipating heat, than nosecone use exact same parts. 

And in many configuration this can be done with minimal offsetting, in aesthetically pleasing way.