Single Stage to Laythe and Back

[PB666]

FYI, the best high bypass turbofans on airliners (which are subsonic) have effective ISPs in the range of about 6,500.

Thats not because their exhaust velocity is ~8x higher than a NERVA rocket.... but because it is *effective* ISP and they do not count the mass of the air used as reaction mass.

KSP does count that mass in their ISP equations... in a 15:1 ratio.... which means they should be using the *real* ISP(ie counting all the reaction mass, not just fuel mass), but 2500 as a value for *real* ISP is ridiculous.

I just had a thought, though, you could use one of the transportable landing platforms, make it superstrong, and place jets around the periphery and use that to raise the launch pad to say 20,000 meters, then launch spacecraft, the problem is you would need an autorecycler that would land the pad. This would avoid all those problems of drag and cut the altitude of gain orbit but 20K. A launch of 20K would have virtually no drag and craft could immediately turn to 45'.

As for the most unrealistic way to get to lathe. Use RCS thruster (which have no mass) and set the ISP to some insanely high number. This way you are only wasting mass on fuel, you could DL a fuel regenerator and get the total mass of fuel, tank and engine below 1.

- - - Updated - - -

I think Linear has got a good point. If you are going to complain about how something implemented into a singleplayer sandbox game in a fictional universe is "cheaty" or "unrealistic", then you might as well not play the game at all. I'm not one to tell other people how to play said type of game, but you want to create an SSTO that has ridiculous design criteria and you add silly constraints on top of that... so to answer your question, it is impossible to make a Laythe SSTO in the manner you want to.

Didn't I make this point the other day in the 'teachings of the Kerballah' post, KSP purism is self-contradictory....you have to create an artificial or contrived logic to justify it.

[KerikBalm]

It's about right for a turbojet (or a turbofan with afterburner on). The J-58 engine used in the SR-71 had a specific impulse of 1900 s at Mach 3.2, while the Olympus 593 engine used in the Concorde had around 3000 s at Mach 2. The basic jet engine should have higher nominal Isp values than it currently has, but the high-performance engines used for supersonic flight are much less efficient.

Yes, the number 2,500 is a reasonable number for *effective* ISP.

Its sort of a matter of how you define ISP... is it impulse per unit *carried* fuel, or impulse per unit *propellant*

Those numbers you see for jet engines in real life do not count the mass of the air expelled.

If you calculate it based upon propellant mass, then specific impulse perfectly correlates with exhaust velocity. This is not so with the way they calculate ISP for air breathing jet engines-> exhaust velocity is very low, and the mass flow rate is very high. Since they don't count the mass of the air, they get high numbers.

You can count the mass of the air, and then set ISP proportional to exhaust velocity

or

You can count only the mass of the fuel you carry, and get a high effective ISP.

KSP gives the engines the high ISP, but then counts the mass of the air... so that the effective ISP, as you would calculate for a jet engine, can be as high as 40,000.

TL:DR version:

Jets and rockets have their ISPs calculated differently

They use ok numbers for ISP when calculated the Jet way, but then do the calculations the rocket way.

Result is an ISP that is 16x too high.

[Jouni]

TL:DR version:

Jets and rockets have their ISPs calculated differently

They use ok numbers for ISP when calculated the Jet way, but then do the calculations the rocket way.

Result is an ISP that is 16x too high.

You're making things too complicated. The definition is just Isp = -F / (g0 * dm/dt), where Isp is the specific impulse in seconds, F is the thrust, and dm/dt is the rate at which the mass of the vehicle changes. There's no need to assume that any kind of propellant is even involved.

[KerikBalm]

Well, dm/dt is where the problem is.

Does the mass include the air or not?

If not... take your 2,500 number, but then don't use the mass of the air in the equations - KSP does include it and that is the problem.

Sure the craft mass only changes relative to the fuel.... but thats one way to think of it... another is to consider the intake air in the craft in any given instant, and that is expelled by the engine... you would see a dm if it flew into a vacuum (like imagine passing through some forcefield/plasma window/whatever and your craft instantaneously enters a hard vacuum).

In this case you would consider the intakes suppying more mass seperately from the engine ISP calculations.

After all, we are calculating ISP for the engine, not the whol craft.

An engine with no fuel supply has no dm/dt... but we can still calculate the rate it would use fuel... and the rate it would use air.

Yes... I am making it more complicated than it needs to be... but I'm trying to get the point across that while 2,500 ISP is a reasonable ISP for an air breathing engine... it is NOT reasonable the way KSP uses it.

[Jouni]

Well, dm/dt is where the problem is.

Does the mass include the air or not?

There is no air, no propellant, and no engine. The mass of the vehicle just changes for a completely irrelevant reason, and there happens to be a force that is proportional to the rate the mass changes. Further details, interpretations, and thinking only make simple things difficult.

Yes... I am making it more complicated than it needs to be... but I'm trying to get the point across that while 2,500 ISP is a reasonable ISP for an air breathing engine... it is NOT reasonable the way KSP uses it.

Is that what you've been trying to say? I already said the same thing on page 2. I thought you were arguing for something completely different.

[Rune]

There is a bug in the game that makes all airbreathing engines burn 16x less fuel than they should. This makes the effective Isp 16x higher than the listed values.

The peak Isp of the turbojet is nominally 2500 s, while the peak thrust is 225 kN. Under these ideal conditions, a single turbojet should burn 225 kN / g0 ≈ 22.9 tonnes of fuel in 2500 seconds, or around 1.83 units of fuel per second. A single Mk1 fuselage (150 units of fuel) should last for around 82 seconds.

That is actually quite reasonable, if you have a decent bypass rate. You are not counting all the air in there! Sure, KSP engines don't overheat and take you to Mach 6, that would be like having the bypass ratio of low-velocity turbofans (which are as fuel efficient as "isp>3.000", which is a wrong statement, but let's go with it), all while having the heat tolerance of the very theoretical SABRE with the fancy precoolers... or a Ramjet with its 'awful' isp (no bypass and stoichiometric burn, but still way better than any rocket by a factor of two!). Sure, completely unreasonable. But you will notice that both numbers are within one order of magnitude of real-life designs. Compare that with the three orders of magnitude or more of the thrust in ion engines which, I agree, makes sense for gameplay reasons.

Rune. So yeah, you might be shocked, but turbojets are more "realistically modelled" than ions. Kerbin has low orbital velocity, that's all.

Edited February 9, 2015 by Rune

[Renegrade]

That is actually quite reasonable, if you have a decent bypass rate. You are not counting all the air in there! Sure, KSP engines don't overheat and take you to Mach 6, that would be like having the bypass ratio of low-velocity turbofans (which are as fuel efficient as "isp>3.000", which is a wrong statement, but let's go with it), all while having the heat tolerance of the very theoretical SABRE with the fancy precoolers... or a Ramjet with its 'awful' isp (no bypass and stoichiometric burn, but still way better than any rocket by a factor of two!). Sure, completely unreasonable. But you will notice that both numbers are within one order of magnitude of real-life designs. Compare that with the three orders of magnitude or more of the thrust in ion engines which, I agree, makes sense for gameplay reasons.

A high-bypass turbofan has as much in relationship with a ramjet as an ion engine has with the Rocketdyne F-1.

*cough*

6.77 meganewton ion engine confirmed for 1.0?

[Starwhip]

Zokesia Skunkworks Project Kiri was about a Single Stage to Laythe and Back, just like you described, that could carry some kind of cargo there.

We're waiting on the larger MK3 parts, and then we'll see what we can do.

[KerikBalm]

That is actually quite reasonable, if you have a decent bypass rate. You are not counting all the air in there! Sure, KSP engines don't overheat and take you to Mach 6, that would be like having the bypass ratio of low-velocity turbofans (which are as fuel efficient as "isp>3.000", which is a wrong statement, but let's go with it), all while having the heat tolerance of the very theoretical SABRE with the fancy precoolers... or a Ramjet with its 'awful' isp (no bypass and stoichiometric burn, but still way better than any rocket by a factor of two!). Sure, completely unreasonable. But you will notice that both numbers are within one order of magnitude of real-life designs.

*sigh* You seem to be missing the point....

#1) Regarding what he said... you seem to be missing the point that 1 mk1 jet fusalage lasts a lot longer than 83 seconds with 1 turbojet running

#2) That value of 2,500 is only reasonable if you don't count all the air in there... because the mass of the air is not counted wen figuring ISP for airbreathing engines.

KSP does count the mass of the air, and that is the problem.

The part in bold is the part I have been trying to get accross over and over again (according to Jouni, not very well, but apparently his explanation didn't reach you either).

KSP jet engines are an order of magnitude too efficient, and propel craft to twice the speed they should, which happens to be orbital velocity.

On Earth, you can get about 1,000 m/s of the needed 8,000 m/s from turbojets (or rather, a turbo/ramjet hybrid like on the SR-71).

On KSP, you can get 2,300 m/s out of the needed 2,300 m/s for orbit.

Talking about orders of magnitude... what is 7,000/0 ?

KSP jets are not realistic. Neither are the ion engines, but thats due to a time-warp game engine limitation. As long as you don't use them for high-ish thrust manuevers (ie, landing on Ike), its not a big deal.

[Red Iron Crown]

Well, dm/dt is where the problem is.

Does the mass include the air or not?

If not... take your 2,500 number, but then don't use the mass of the air in the equations - KSP does include it and that is the problem.

Sure the craft mass only changes relative to the fuel.... but thats one way to think of it... another is to consider the intake air in the craft in any given instant, and that is expelled by the engine... you would see a dm if it flew into a vacuum (like imagine passing through some forcefield/plasma window/whatever and your craft instantaneously enters a hard vacuum).

In this case you would consider the intakes suppying more mass seperately from the engine ISP calculations.

After all, we are calculating ISP for the engine, not the whol craft.

An engine with no fuel supply has no dm/dt... but we can still calculate the rate it would use fuel... and the rate it would use air.

Yes... I am making it more complicated than it needs to be... but I'm trying to get the point across that while 2,500 ISP is a reasonable ISP for an air breathing engine... it is NOT reasonable the way KSP uses it.

Properly speaking, Isp only considers mass carried by the vessel (i.e. fuel but not intake air). It's the difference between exhaust velocity and effective exhaust velocity.

Definitely agree that KSP is not calculating it properly, it's taking advantage of the intake air twice.

[KerikBalm]

Well we need to agree on a definition.

According to wikipedia "Specific impulse (usually abbreviated Isp) is a measure of the efficiency of rocket and jet engines. By definition, it is the impulse delivered per unit of propellant consumed, and is dimensionally equivalent to the thrust generated per unit propellant flow rate"

In this case, air is propellant, and its the flow rate, not change in mass of the craft... using this definition, you'd never get a jet with a 2,500 ISP

You can even argue that there is air with mass inside the craft, and that air is being evacuated... its simply being replaced at the same rate.

If you had a rocket with fuel hoses hooked up to it, you wouldn't say the Isp is infinite while it is being fed ue trouh the hoses (obviously, a situation that wouldn't last long before one runs out of hose... but... you get the idea)

http://en.wikipedia.org/wiki/Specific_impulse#Specific_impulse_as_a_speed_.28effective_exhaust_velocity.29

"For a rocket, the specific impulse defined in this way is simply the effective exhaust velocity relative to the rocket, ve. The two definitions of specific impulse are proportional to one another, and related to each other by: v_e = g_0 I_sp, "

...

Actual exhaust speed versus effective exhaust speed

...

For air-breathing jet engines, particularly turbofans, the actual exhaust velocity and the effective exhaust velocity are different by orders of magnitude. This is because a good deal of additional momentum is obtained by using air as reaction mass. This allows a better match between the airspeed and the exhaust speed, which saves energy/propellant and enormously increases the effective exhaust velocity while reducing the actual exhaust velocity"

Because exhaust velocity and Isp are proportional (for rockets... everyone should agree here), and they use the term "effective exhaust velocity" I was using the term "effective Isp"

Last night, I made a "realistic" turbojet engine (or rather air turboramjet, or hybrid ramjet-turbojet like on the SR-71)

I cut the Isp value in KSPs game files from 2,500 to 1250 -> thereby doubling the mass flow of the engine

I then turned the intake air: fuel ratio from 15:1 to 1:1 ->

At this point, the engine was using 16x more fuel, and 8/15*2 = 1.06666x more intake air

The engine's TWR was also more than double that of the modern military F110

http://en.wikipedia.org/wiki/General_Electric_F110

So I cut the thrust in half

IIRC, jet ISP doesn't really decline so sharply with altitude, and in KSP, the effective Isp is modified by the velocity curve (at 0 m/s the turbojet produces half the thrust, but fuel use is unchanged). So keeping in mind the velocities one typically would be going at various altitudes, I came up with the following stats:

atmosphereCurve

{

key = 0 1200

key = 0.3 1250

key = 1 1800

}

velocityCurve

{

key = 0 0.5 0 0

key = 1000 1 0 0

key = 1500 0.5 0 0

key = 1700 0 0 0

}

Maybe I should nerf the max speed even more, but due to drag (even using NEAR), the first craft I tested didn't get much over 1,100 m/s anyway... which is realistic enough.

But then the other problem is KSPs soup-o-sphere.... The current atmosphere is so bad, that jet engines need to be a little OPd to be viable.

Given that the aerodynamics model will be updated the next version, I used NEAR as a stand in approximation (what it does, and what they describe the next version is, sounds very similar).

I had to delete this part of the .cfg to stop it mucking with my engines:

@PART[*]:HAS[@MODULE[ModuleEngines*]:HAS[@PROPELLANT[intakeAir]]]:BEFORE[NEAR]:NEEDS[!AJE]

{

@MODULE[ModuleEngines*]

{

@maxThrust *= 0.5

}

}

The result:

Now its still quite unrealistic... but there's some that can't be helped... The fuel consumption, TWR, and velocity I achieved before switching to rockets were all reasonable. Its simply with a 1/10th planetary radius... 1,100 m/s in the atmosphere (orbital speed even higher due to planetary rotation) is quite a lot.

For reference, I removed the wings (only used winglets for control), intakes (slapped on aero nose cones), jets, switched the LF tank for a LF+O tank, and tried a SSTO rocket, the result:

Both into 100km x 100km orbits.

Realistic jet performance, combined with a better aero model, allow SSTOs just fine on this 1/10th scale planet... so lets fix jet engines.

FWIW, I also changed the basic jet, aiming for a mostly subsonic performance, with a an effective exhaust velocity corresponding to an Isp of 6,500

atmosphereCurve

{

key = 0 2000

key = 0.3 2750

key = 1 3250

}

velocityCurve

{

key = 600 0 0 0

key = 350 0.9 0 0

key = 0 1 0 0

}

1:1 intake ratio, same TWR as the previous turbojet.

Endurance was quite good when packing more jet fuel, I could fly (again... NEAR making the atmosphere much less soupy) throttled pretty far back:

I'm not sure kerbal engineer was calculating thrings right, but it said 1 "realistic" basic jet, with 3 mk1 jet fuel tanks, at maximum throttle, flying at sea level, had an endurance of 24 minutes...

And if you can fly around at 30% throttle... that's quite a bit of flight time...

So... lets fix these Isps when the aero model gets fixed!

[Red Iron Crown]

Well we need to agree on a definition.

According to wikipedia "Specific impulse (usually abbreviated Isp) is a measure of the efficiency of rocket and jet engines. By definition, it is the impulse delivered per unit of propellant consumed, and is dimensionally equivalent to the thrust generated per unit propellant flow rate"

In this case, air is propellant, and its the flow rate, not change in mass of the craft... using this definition, you'd never get a jet with a 2,500 ISP

I looked into this a bit deeper as I always thought of air as a propellant, turns out I was wrong. "Technically, the word propellant is the general name for chemicals used to create thrust. For vehicles, the term propellant refers only to chemicals that are stored within the vehicle prior to use, and excludes atmospheric gas or other material that may be collected in operation."

Great post, btw, I love backing up theory with real experimentation. Science!

[timsfitz]

I've never managed to build a SSTO that can get back from Laythe. It's a challenge for sure, but I have a feeling it will be easier with the new parts.

[KerikBalm]

Yea... note the speed in that last picture though.... 365 m/s.... cruising on 25kN of thrust....

You can't do that in stock. We need a fixed aerodynamics model.

A fixed aerodynamics model will allow use to use jets that produce much less thrust, and still fly just fine. So the planes will still fly like planes, but vertical launch turbojet lifters won't work so well anymore.

The reduced TWR means that fuel consumption is reduced... which makes fixing the air intake ratio/Isp of jet engines more viable.

I used nearly 500 units of liquid fuel to get to about 30km and 1,110 m/s on the equivalent of less than 1 stock turbojet's worth of thrust (since max thrust was 100kN, while a stock turbojet is 225kN).

Try that in stock... you'll use much much less.

Right now we have no need for these large LF tanks, except for ridiculous amounts of spare fuel... like on laythe...

Fix the stock engines, and people will actually have to consider adding significant amounts of LF tanks to their planes.... which, IMO is good and renders a part non-pointless

[cantab]

Let's get back to the core question. You want to build a realistic plane to go to Laythe and back.

Step 1: Install FAR and DRE, or wait for the 1.0 KSP release. FAR isn't perfect and it has its oddities, but it's a LOT better than stock aerodynamics.

Step 2: Install a mod for realistic jets such as AJE. Or if you must use the stock engines, use the Basic Jet. You can certainly build an LKO spaceplane with it.

Step 3: Sort your in space engine out. The LV-N has realistic ISP if it were using hydrogen, but hydrogen is much bulkier than KSP fuel and boils off over time. The ions have reasonable Isp but excessive thrust, so I would say use them to transfer where it doesn't matter much but not to circularise since real ions wouldn't have the thrust for that.

[AbacusWizard]

I've built a spaceplane that can take off from Kerbin and land on Laythe... but the "able to go to Laythe" part was taken care of by docking with an interplanetary atomic freighter that I already had in orbit from previous missions. Infrastructure is awesome.

[Overfloater]

Sure it's possible, here's one i quickly slapped together

This is my most "legitimate" craft that flew to Laythe and back.

- No wing/flap spamming;

- No SAS spamming;

- No intake spamming.

What you see is all it has!

Its very similar to Warmblood's plane.

Edited February 9, 2015 by Overfloater

[AbacusWizard]

I'm sure the pilot enjoys the cozy warm glow from the nearby nuclear engine.

[Starhawk]

I just took this one on a round trip to Laythe.

This is the Snowbird.

I have to feel sorry for Jeb, though. I was just looking at the cockpit view in the Mk1 inline cockpit and thinking about how horribly bored Jeb must feel after 7 years and 386 days in there.

[EDIT] Total cost of round trip 1609 kurrency. Because I went direct to aerocapture at Laythe, Jeb had to endure something like 29g. Good thing he's g-proof in this career.

Edited February 10, 2015 by Starhawk

[AbacusWizard]

I have to feel sorry for Jeb, though. I was just looking at the cockpit view in the Mk1 inline cockpit and thinking about how horribly bored Jeb must feel after 7 years and 386 days in there.

Gotta be worth it for that view, though; wow!

[Wanderfound]

Refuelled, but single stage and carrying 26 Kerbals, to Laythe (able to return, but I didn't show that), flying with FAR and DRE:

http://s1378.photobucket.com/user/craigmotbey/Kerbal/Kerbodyne%2025/Kerbodyne%20Kosciuszko/story

&

http://s1378.photobucket.com/user/craigmotbey/Kerbal/Kerbodyne%2025/Kerbodyne%20Kosciuszko/Laythe%20Landing/story

[panzer1b]

Hey,

did anyone of you design a spaceplane which is able to go to laythe, land there, and then return to kerbin? I set myself the rule to build as realistic as possible, so i don't spam wings, control surfaces, SAS, and especially intakes.

I would like to know if it is possible at all, with these restrictions. For now my best spaceplane is able to escape Laythe.

My HK-303A "Tri-Fighter III" can pull off a laythe landing roundtrip back to KSC. The one in this pic is launched with a booster and guided kinetic missiles, and it gets like 7K dV with it, but the pure SSTO ground launched variant can make laythe and back.

Its not exactly 100% to your requirements, and it does use some clipping to maintain aesthetics. Still, it gets above 4K dV in orbit, and can land on laythe reliably provided you make an efficient approach, then return to kerbin (ive even managed to get it near KSC one time but ran outta LF to land on runway).

It weighs roughly 20t, and has 15 intakes. Ive found that roughly 2/3 intake area/ton to be the minimum to be able to get a craft out of atmosphere on pure jets. You can get away with less, but it becomes very hard to make it efficient per say. There are wings internal to the craft, and yes its unrealistic, but i build for looks above all else, my craft are more or less external 1st, and then cram the inside with whatever i need to make it work. Then again, as i like sci-fi designs above another F22 clone, i do take advantage of issues with the game engine to pull off some rather impossible stuff. Power consists of 3 rapiers (mostly used in jet mode), and 1 nuke, which provides atmospheric efficiency, long range interplanetary with the nuke, and TWR for landing in very high gravity and combat maneuvers.

As for your requirements, im not sure it can be pulled off.

The biggest issues is your requirement to be realistic, realistically, if you are 100% against clipping, and using multiple intakes per engine, i highly doubt you can obtain enough dV to get to laythe, land, and rtb. If you are also unwilling to use what id consider not realistic parts (ions, jets, nuke), then you can completely forget about making that much dV. Even a 100% fuel tank with a super light 390isp engine, will provide nowhere near 10000dV, and it takes 4500ish to get out of kerbin, and a bit under 3K to get out of laythe. This doesnt count the roughly 4K needed for going to laythe, and back. This is assuming you dont use jets and or nukes/ions.

Edited February 11, 2015 by panzer1b

[Jouni]

The biggest issues is your requirement to be realistic, realistically, if you are 100% against clipping, and using multiple intakes per engine, i highly doubt you can obtain enough dV to get to laythe, land, and rtb.

I quickly built the plane I outlined on page 2, and it has 5 km/s remaining at a 120 km orbit.

The plane is just like you see it. There are no hidden surprises. It could probably go to Laythe and back, but I'm too lazy to try.

[cerberusti]

It was a few patches ago (the first science patch I believe):

http://steamcommunity.com/sharedfiles/filedetails/?id=239385512

It takes a full set of science gear plus a lab to Laythe and back to the runway without refueling, and is completely stock other than mechjeb (also, there is a kerbal in a seat on the front gear to easily collect samples.) I used a Tylo assist to get back and hit Laythe directly on the way in (involving an over 70G aerobrake at Laythe as I came down straight on top of my landing site from an interplanetary approach.)