Theoretical optimum of cost efficiency to orbit (1.2)

[Curveball Anders]

21 minutes ago, Leafbaron said:

I would say an even more relevant figure would be the cost of the launch vehicle not including the cost of the payload then divide that figure by weight of the payload. 

Yeah, that's what I meant.

[Reusables]

13 hours ago, Leafbaron said:

I imagine the theoretical limit would be close to 0. if  you use a spaceplane with ISRU and had an asteroid in LKO. Once you reach space a release payload you could then restock on fuel from the asteroid and return to KSC as though you hadn't used any fuel at all, less the 30-40 m/s required to deorbit. 

I think refueling should count as profit, so should not included in the lifter cost.

By the way, is it practical(profitable) to get an asteroid to LKO and mine it? I was told that it costs too much, so lifting fuel to orbit is cheaper.

13 hours ago, Norcalplanner said:

Back when I ran a similar challenge in 1.0.5, it was a major accomplishment to get below 700 funds/ton, and only a few got below 600 funds/ton.  It might be a good idea to take a look at those previous entries for design ideas.

 

Thanks, I like the challenge and the impressive results!

[KerikBalm]

12 hours ago, Curveball Anders said:

I assume that a more relevant figure would be launcher cost divided by payload tonnage.

 

13 hours ago, TimePeriod said:

I do not really understand what you are trying to get at? I already said it was a disposable rocket, nor an SSTO.

 I took the total cost of the entire craft+cargo and divided it by the total tonnage, is that not you are supposed to do when figuring out how much it would cost pr ton?

I said "even if" - it was clearly not 275/ton for a disposable... its not even 275/ton if it was a SSTO

12 hours ago, TimePeriod said:

In that case, we get a lot higher number; 19345 / 4.5 = 4298/t (I think)

That's what I was getting at. When calculating the cost of getting a payload into orbit, you don't care about the tonnage of SRBs you discard before orbit. The way you calculate it, one just wants to use the parts that have the most mass per fund. If you included an Ore tank, your number would be lower if the Ore tank was full, even though it  has nothing to do with getting payload to orbit and holds back the entire rocket.

275/ton is ridiculously good, only achievable with SSTOs, and I think even then its hard to achieve without air breathing engines(I suppose we shall need to do some math like I did for the Rapiers to see).

39 minutes ago, Abastro said:

I think refueling should count as profit, so should not included in the lifter cost.

By the way, is it practical(profitable) to get an asteroid to LKO and mine it? I was told that it costs too much, so lifting fuel to orbit is cheaper.

Thanks, I like the challenge and the impressive results!

I wouldn't include refueling: that requires elements independent of the launcher. It doesn't have to be an asteroid. You could have a large fuel depot in LKO, and supply it from Mun/Minmus mining. (something like this)

Spoiler

Then its simply a matter of launching an SSTO with only partially full tanks (the payload here is essentially empty tanks), refueling everything, and bringing it back down. Given the small % of total craft cost that goes to fuel, this probably won't net a profit unless you get 100% recovery, so this heavily favors space planes, or at least winged rockets (a vertical launch, horizontal landing rocket vaguely resembling the shuttle orbiter could work)

[Leafbaron]

6 hours ago, Abastro said:

I think refueling should count as profit, so should not included in the lifter cost.

By the way, is it practical(profitable) to get an asteroid to LKO and mine it? I was told that it costs too much, so lifting fuel to orbit is cheaper.

Thanks, I like the challenge and the impressive results!

I think so. If the right asteroid comes along and is easy enough to capture. The plus side of doing this versus launching fuel to a depot is the fuel you get from the asteroid cost nothing, excluding the cost to get it to LKO. and I assume that the asteroid will supply more fuel that it took to get it there by quite a margin. So even if we do factor the cost of fuel getting the asteroid to LKO doing so will result in a net positive. 

[Whisky Tango Foxtrot]

13 hours ago, Abastro said:

By the way, is it practical(profitable) to get an asteroid to LKO and mine it? I was told that it costs too much, so lifting fuel to orbit is cheaper.

The trick is to intercept it far from Kerbin, change its orbit to be equatorial with a PE of around 60km or so and use multiple aerobraking passes to lower its AP. You can get a sizable asteroid into LKO that way with minimal fuel expenditure. Of course, it helps if its starting trajectory is close to what you want.

[Aegolius13]

On 2/26/2017 at 9:54 AM, Cunjo Carl said:

A practical disposable whiplash launcher has been a white whale of mine over the years.

 

On 2/27/2017 at 5:35 AM, MR L A said:

I have two TSTO disposable whiplash launchers! One of them (J.A.M.S - Jet Assisted Micro Shuttle) is a drone... I wouldn't really call it practical tbh.. it can take a tiny payload to LEO, but I've never actually used it on preference to my 1.25m rockets which can lift more for less. The other design is a crew transport which is insanely fun to fly! Take it up to a space station, change crew and then return, much more fun than a rocket! This second design (J.A.R.O.D.S - Jet Assisted Rapid Orbital Deployment Shuttle) is one that I actually use, though launch timing is pretty essential.

As far as I can tell both designs have 3 flaws:

1) "disposable" whiplash systems are less cost effective than traditional rockets (though I guess this only matters in Career mode)

2) I use ram intakes rather than shockcones, so that sees a reduction in performance. Cost trade off (see above)

3) both are fairly small designs and end up with not a lot of fuel to play with once in orbit
 

On the other hand they are SOOO0o0o0o cool.

I'll upload some photos tonight if I remember

I like using a disposable Whiplash launcher for smallish projects.  I've found that one Whiplash can booster around 4 tons of upper stage and payload for just a tiny bit more money than a Swivel-based booster.  Plus it gets going quite a bit faster than a Swivel by the time it burns out. 

I've had the best luck with the divertless intake.  It's dirt cheap, does not need a front node, one holds more than enough fuel for the whole flight, and it seems to suck in enough air.  

This system does seem to only work with a fairly narrow payload range.  If it gets too big, the craft won't accelerate fast enough to overcome the thrust loss as you climb.  Might be possible to build a booster with multiple Whiplashes, but that probably starts losing out to things like Kickbacks and Skippers.  

[rcp27]

I haven't done any detailed study, but for small payloads to LKO my feeling is that for minimizing cost with a disposable launch system, a system based on an SRB is the way to go. I've made serious money on LKO career contracts using a QBE core, a couple of solar panels, a communotron 16, 0.975 inline battery and reaction wheel as the payload, with an ant and 3 oscar Bs as an orbital engine on a 0.975 m decoupled, on top of a 1.25 m fairing, Thumper SRB with three basic fins and a launch clamp. For that king of payload, the cost to orbit isn't that far off the fuel burn alone for an SSTO capable of delivering the same payload.

[Reusables]

 

20 hours ago, Leafbaron said:

I think so. If the right asteroid comes along and is easy enough to capture. The plus side of doing this versus launching fuel to a depot is the fuel you get from the asteroid cost nothing, excluding the cost to get it to LKO. and I assume that the asteroid will supply more fuel that it took to get it there by quite a margin. So even if we do factor the cost of fuel getting the asteroid to LKO doing so will result in a net positive. 

So it varies per orbit of the asteroid? I didn't know that, I've never played with asteroids.

13 hours ago, Whisky Tango Foxtrot said:

The trick is to intercept it far from Kerbin, change its orbit to be equatorial with a PE of around 60km or so and use multiple aerobraking passes to lower its AP. You can get a sizable asteroid into LKO that way with minimal fuel expenditure. Of course, it helps if its starting trajectory is close to what you want.

Wow, aerocapture? That looks great solution for this one!

5 hours ago, Aegolius13 said:

I like using a disposable Whiplash launcher for smallish projects.  I've found that one Whiplash can booster around 4 tons of upper stage and payload for just a tiny bit more money than a Swivel-based booster.  Plus it gets going quite a bit faster than a Swivel by the time it burns out. 

I've had the best luck with the divertless intake.  It's dirt cheap, does not need a front node, one holds more than enough fuel for the whole flight, and it seems to suck in enough air.  

This system does seem to only work with a fairly narrow payload range.  If it gets too big, the craft won't accelerate fast enough to overcome the thrust loss as you climb.  Might be possible to build a booster with multiple Whiplashes, but that probably starts losing out to things like Kickbacks and Skippers.  

If my calculation is right, Skipper is not very cost efficient for disposable lifter compared to Reliant.

On the other hand, kickback is great for first disposable stage compared to the other choices. Though I think you can get a bit more advantage with dv.

3 hours ago, rcp27 said:

I haven't done any detailed study, but for small payloads to LKO my feeling is that for minimizing cost with a disposable launch system, a system based on an SRB is the way to go. I've made serious money on LKO career contracts using a QBE core, a couple of solar panels, a communotron 16, 0.975 inline battery and reaction wheel as the payload, with an ant and 3 oscar Bs as an orbital engine on a 0.975 m decoupled, on top of a 1.25 m fairing, Thumper SRB with three basic fins and a launch clamp. For that king of payload, the cost to orbit isn't that far off the fuel burn alone for an SSTO capable of delivering the same payload.

Yeah, SRBs are cheap. May I know your cost efficiency in lifter cost per payload mass?

I think you would get TWR over 2~3, which will kill the efficiency (drag penalty will be heavy on transonic region). If you have the SRB with thrust limiter way down, you'd better have multiple SRBs, I think.

Also, AFAIK you can get better result with other SRBs like Hammer and Kickback. Thumper has slightly better Isp than Hammer, but poor wet-dry ratio of Thumper compared to Hammer makes up the dv difference.

 

By the way, experiments have shown that TWR of 1.5 is better than 1.4 for sufficiently aerodynamic rocket. I'll update the atmospheric stage information with that.

Edited March 1, 2017 by Reusables

[Leafbaron]

3 hours ago, Abastro said:

 

So it varies per orbit of the asteroid? I didn't know that, I've never played with asteroids.

 

What do you mean by this? Some asteroids pass very close to Kerbin in favorable inclinations. so that are easy to capture because with a very small push they can be captured in a stable orbit and then tweaked down to LKO <250km. Asteroids do contain a finite amount of ore to my understanding. 

Edited March 1, 2017 by Leafbaron

[Reusables]

14 minutes ago, Leafbaron said:

What do you mean by this? Some asteroids pass very close to Kerbin in favorable inclinations. so that are easy to capture because with a very small push they can be captured in a stable orbit and then tweaked down to LKO <250km. Asteroids do contain a finite amount of ore to my understanding. 

Yeah, that's exactly what I meant. I just realized that after reading the post.

Edited March 1, 2017 by Reusables

[Archgeek]

Sounds like for disposables the way to go might be a well-tuned, dense, payload of ore and a nosecone atop a carefully tweaked kickback.  Possibly with a flea-based PAM for circularization.  Heck, if you're gonna PAM, you could go partially disposable -- toss a couple of chutes on the kickback, get the apo a decent ways into the future, splash down the kickback, then pop the PAM near apo.  Perhaps the amusingly low-drag (and fairly cheap, I think?) juno intake could be leveraged nosecone-ways.

[rcp27]

I'm home from work now and have re-created this launcher and flown a test flight.

8 hours ago, Abastro said:

On the other hand, kickback is great for first disposable stage compared to the other choices. Though I think you can get a bit more advantage with dv.

There is so much wrong with this particular launcher that it is borderline un-flyable.  Anything approaching an efficient flight profile is just not happening.  The first issue is that, while the Thumper can get this load out of the atmosphere, you don't have enough margin to do a very effective gravity turn.  In the event, though, that's not much a problem.  The real problem comes from the Ant.  It gives a new definition to "feeble".  The burn time required to get any kind of useful dV is insane.  If you do make a nice gravity turn so you leave the atmosphere with an Ap of perhaps 85 km and a the rest of the energy in prograde orbital velocity, because the Ant is so feeble, you will go up to Ap, come down the other side and burn up on re-entry long before you are anywhere near orbital velocity. The only way I can make this thing fly at all is basically to go straight up.  The objective is to get as high an Ap as possible from the Thumper to maximise the time I have between leaving the atmosphere and re-entering in the hope that the Ant can burn its six little legs off and achieve a stable orbit.  The launch went something like this:  Go straight up.  The Thumper quits at about 40 km altitude, with an Ap of around 200 km.  Once I hit 60 km, I ditched the fairing and Thumper, pointed between 25 and 30 degrees above the horizon and burned that Ant on 100%.  All the way up.  And back down again.  I passed through Ap at about 220 km, and finally got the Pe above 70 km when I was at 140 km altitude coming down the other side.  I flew this with the Thumper at 50% thrust, but didn't put too much effort into optimising it.

8 hours ago, Abastro said:

Yeah, SRBs are cheap. May I know your cost efficiency in lifter cost per payload mass?

The rundown of the whole craft is as follows:

1 communotron 16: cost 300, mass 0.005

1 Probodobodyne QBE: cost 360, mass 0.07

2x OX 4L 1x6 solar panels: cost 2x380, mass 2x 0.0175 (yes, I could get by with the OX stat for lest cost and weight, but I prefer bigger panels)

1x 2HOT thermometer: cost 900, mass 0.005 (redundant, there to demonstrate the satellite can complete contracts and do science)

1x Z200 0.925 m inline battery: cost 360, mass 0.01 (a pair of Z100 batteries would be cheaper for the same charge/weight)

1x small inline reaction wheel: cost 600, mass 0.05

3x Oscar B fuel tanks: cost 3x 70, mass 3x 0.225

1x Ant engine: cost 110, mass 0.02

1x TR-2V stack decoupler: cost 300 mass 0.015

1x AE-FF1 Airstream protective shell containing all of the above, cost 300, mass 0.075

1x Thumper SRB: cost 850, mass 7.56

3x Basic Fin: cost 3x 80, mass 3x 0.01

1x TT18-A launch stability enhancer: cost 200 (recovered), mass irrelevant.

The orbiter itself has a dry mass of 0.27 and a cost of 3545 (dry), 3600 (wet).  Total cost of launch (inclusive of payload): 5392 (5192 after launch clamp recovery).  That's a total cost to orbit of 1647 plus a little over 2 Oscar B's worth of fuel (I had a little less than 1 tank left for establishing the target orbit, but that's plenty for completing most simple contracts).

8 hours ago, Abastro said:

I think you would get TWR over 2~3, which will kill the efficiency (drag penalty will be heavy on transonic region). If you have the SRB with thrust limiter way down, you'd better have multiple SRBs, I think.

Also, AFAIK you can get better result with other SRBs like Hammer and Kickback. Thumper has slightly better Isp than Hammer, but poor wet-dry ratio of Thumper compared to Hammer makes up the dv difference.

There are plenty of ways to make the craft more efficient if you think purely in terms of fuel burn, but they all add to the cost significantly more than they save.  A kickback alone costs 2700, which is more than the total cost to orbit of this launch, so that's a non-starter.  A Hammer lacks the dV to get the Ap out of the atmosphere, so would need to be combined with some other sustainer engine.  That means a TR-18A stack decoupler.  A Hammer plus a decoupler cost 800 combined before you add the cost of the sustainer engine.  That's only 50 cheaper than a Thumper, but even a sepatron costs 75, and that's the cheapest engine in the game.  In an earlier iteration of this launcher I used a Spark, which can achieve the same orbit with only 2 Oscar B's worth of fuel, and a much easier flight profile because it has something like 10 times the thrust.  But a Spark costs 240 (so 130 more than an Ant) and the third Oscar B only costs 70 wet, so where's the saving?

Edited March 1, 2017 by rcp27
Added SRB thrust detail

[Birdco_Space]

I have a question on how we define payload?  Is this weight of craft to orbit, or weight of just the things to complete the mission?  do we factor recovery cost into this?

I have a SSTO space plane that costs 32k funds, but if recovered on the runway only expends around 800 funds in fuel (plus a few hundred funds on the occasion I derp the landing and remove a wing).  The functional payload is only 3 kerbals, but I get the whole aircraft to orbit, which weighs around 5 tonnes when in orbit. 

So this could be considered tremendously efficient working the numbers right (800 funds for a 5 ton spacecraft to orbit, equating to 160 funds per ton), or not so efficient (32k fund spacecraft to get 3 80kg kerbals to orbit, for 200k funds per ton)

[MR L A]

On 27/02/2017 at 3:01 PM, Abastro said:

I think you should be using cheaper air intake (like diverterless) and doing more aggressive gravity turn. Ram intakes are just too expensive for a disposable craft. Also you need less initial TWR for the airbreathers.

AFAIK whiplash has OP Isp (nearly no need to bother fuel usage) and decent TWR on supersonic speed..

Gonna disagree, the system I have keeps part count low, provides the perfect amount of fuel for 4 whiplashes and keeps the ship aesthetic i.e. not cluttered with a dozen ugly intakes ruining aero profile. Using air intake graphs (found somewhere on the forums) the only intake better is the shock cone which is definitely too expensive. The rest simply don't allow the engines to function well enough. The system I use isn't that expensive anyway... one Ramp intake, one Nacelle (fuel adjusted and cheap) and one Whiplash. Works like a treat. But like I said, it's mostly about being fun to fly than practical (I'd just use a rocket otherwise), though it does get to orbit pretty damned quickly. 

Also, I didn't mention gravity turns? This is a plane with a rocket on the back, not a rocket with jets.
 

Edited March 2, 2017 by MR L A

[Reusables]

On 2017. 3. 2. at 4:53 AM, rcp27 said:

I'm home from work now and have re-created this launcher and flown a test flight.

There is so much wrong with this particular launcher that it is borderline un-flyable.  Anything approaching an efficient flight profile is just not happening.  The first issue is that, while the Thumper can get this load out of the atmosphere, you don't have enough margin to do a very effective gravity turn.  In the event, though, that's not much a problem.  The real problem comes from the Ant.  It gives a new definition to "feeble".  The burn time required to get any kind of useful dV is insane.  If you do make a nice gravity turn so you leave the atmosphere with an Ap of perhaps 85 km and a the rest of the energy in prograde orbital velocity, because the Ant is so feeble, you will go up to Ap, come down the other side and burn up on re-entry long before you are anywhere near orbital velocity. The only way I can make this thing fly at all is basically to go straight up.  The objective is to get as high an Ap as possible from the Thumper to maximise the time I have between leaving the atmosphere and re-entering in the hope that the Ant can burn its six little legs off and achieve a stable orbit.  The launch went something like this:  Go straight up.  The Thumper quits at about 40 km altitude, with an Ap of around 200 km.  Once I hit 60 km, I ditched the fairing and Thumper, pointed between 25 and 30 degrees above the horizon and burned that Ant on 100%.  All the way up.  And back down again.  I passed through Ap at about 220 km, and finally got the Pe above 70 km when I was at 140 km altitude coming down the other side.  I flew this with the Thumper at 50% thrust, but didn't put too much effort into optimising it.

So, you are going for highly inefficient trajectory due to the TWR of circularizing stage. If you don't have enough TWR to circularize(spends too long time), then you are doing it wrong.

In detail, You should be losing over 600m/s of dv doing inefficient gravity turn, and lose another 300m/s performing circularization burn with pitch of 30 degrees. That's nearly 1km/s of lose!

Also, 50% is way down for Thumper. You may as well use Hammer.

On 2017. 3. 2. at 4:53 AM, rcp27 said:

The rundown of the whole craft is as follows:

1 communotron 16: cost 300, mass 0.005

1 Probodobodyne QBE: cost 360, mass 0.07

2x OX 4L 1x6 solar panels: cost 2x380, mass 2x 0.0175 (yes, I could get by with the OX stat for lest cost and weight, but I prefer bigger panels)

1x 2HOT thermometer: cost 900, mass 0.005 (redundant, there to demonstrate the satellite can complete contracts and do science)

1x Z200 0.925 m inline battery: cost 360, mass 0.01 (a pair of Z100 batteries would be cheaper for the same charge/weight)

1x small inline reaction wheel: cost 600, mass 0.05

3x Oscar B fuel tanks: cost 3x 70, mass 3x 0.225

1x Ant engine: cost 110, mass 0.02

1x TR-2V stack decoupler: cost 300 mass 0.015

1x AE-FF1 Airstream protective shell containing all of the above, cost 300, mass 0.075

1x Thumper SRB: cost 850, mass 7.56

3x Basic Fin: cost 3x 80, mass 3x 0.01

1x TT18-A launch stability enhancer: cost 200 (recovered), mass irrelevant.

The orbiter itself has a dry mass of 0.27 and a cost of 3545 (dry), 3600 (wet).  Total cost of launch (inclusive of payload): 5392 (5192 after launch clamp recovery).  That's a total cost to orbit of 1647 plus a little over 2 Oscar B's worth of fuel (I had a little less than 1 tank left for establishing the target orbit, but that's plenty for completing most simple contracts).

Let's calculate payload mass and lifter cost here.

Payload: Antenna(0.005t)+Probe(0.07t)+2xSolar Panels(0.035t)+Thermometer(0.005t)+Battery(0.01t)+Reaction Wheel(0.05t)+1 OscarB fuel tank(0.225t) = 0.4t

Lifter: Ant(110)+2xOscarB(140)+Decoupler(300)+Fairing(300+)+Thumper(850)+3xBasic Fin(240) = 1940

So you are using 4850 per 1t of payload! AFAIK you can use less money to lift 1t.

I think these improvements will cut down your cost to target orbit decently:

1. Fairing is too expensive and heavy for such a small payload. It'll cost over 300 and heavier than 0.075t, since cost and mass of deployed parts is added to the base cost/mass. If you are worried about aerodynamic stabilization, just put small nose cone(180, 0.01t). (You won't need it)

 (If it is for aesthetis, let me know, as it is reasonable choice for the purpose)

   + If you think aerodynamics is a thing, I can say it's definitely not. Definitely, the loss is smaller than 100m/s.

2. If you are aiming to do circularization burn with single Ant engine, just use single OscarB tank to get more TWR. Here are some calculations to verify this:

   Now, your dry mass of second stage is 0.27t, and its wet mass is 0.87t, thus it has dv of 3.6km/s. When you are on LKO with an OscarB tank fuel to spare, the wet mass of the stage is 0.47t. So you have spare dV of 1.7km/s. AFAIK that's enough to get it to Duna and Eve! Dv of 1.2km/s will be enough. (1km/s lets you escape kerbin)

  When you have single OscarB tank, the second stage has dry mass of 0.22t and wet mass of 0.42t. Thus initial TWR of 0.485 and dV of 2km/s. It'll be enough to do efficient circularizing burn(TWR increases over time). Considering 1km/s you wasted on gravity turn, and 0.6km/s of unnecessary dv, this one is only 0.1km/s short! Also the reduced dry mass will allow more dV for the first stage.

3. You are fine with single Hammer, as its dv is not that short! With the payload above, you need to launch 0.435t (including decoupler).

  Single Hammer has atm dv of 170*9.81*ln(3.995/1.185) = 2km/s. If you are doing efficient gravity turn, 2.8~3km/s is sufficient while going eastward. Thus 1~1.2km/s of dV will be left. It could be a bit short with westward orbit. If you want some more dv, just put something like Flea under Hammer, and explode it away.

+ If you want more dV, 2 Hammer with explosive decoupling will work. The lower stage gives dV of 775m/s more. The so the two Hammer has total dV budget of 2.7~2.8km/s with the payload, which can make orbit themselves! It could be a bit unstable, so some dV will be lost. Though it'll be still enough.

+ Probodobodyne HECS costs 650 and weigh 0.1t. So it's cheaper and lighter than 1 small reaction wheel + 1 QBE probe! The torque should be enough to turn such a small craft. (Also this will give you more spare dV) Also, why do you need battery?

On 2017. 3. 2. at 4:53 AM, rcp27 said:

There are plenty of ways to make the craft more efficient if you think purely in terms of fuel burn, but they all add to the cost significantly more than they save.  A kickback alone costs 2700, which is more than the total cost to orbit of this launch, so that's a non-starter.  A Hammer lacks the dV to get the Ap out of the atmosphere, so would need to be combined with some other sustainer engine.  That means a TR-18A stack decoupler.  A Hammer plus a decoupler cost 800 combined before you add the cost of the sustainer engine.  That's only 50 cheaper than a Thumper, but even a sepatron costs 75, and that's the cheapest engine in the game.  In an earlier iteration of this launcher I used a Spark, which can achieve the same orbit with only 2 Oscar B's worth of fuel, and a much easier flight profile because it has something like 10 times the thrust.  But a Spark costs 240 (so 130 more than an Ant) and the third Oscar B only costs 70 wet, so where's the saving?

I know that Kickback is improper for this kind of thing. Also I think you're right about Spark. But, AFAIK, you can do explosive decoupling with Hammer.

15 hours ago, Birdco_Space said:

I have a question on how we define payload?  Is this weight of craft to orbit, or weight of just the things to complete the mission?  do we factor recovery cost into this?

AFAIK, payload is something that can be decoupled from the lifter, technically. In my opinion, it is acceptable if you can decouple it by adding some decouplers. About recovery cost, it's included in reusable case, and excluded in disposable case.

4 hours ago, MR L A said:

snip

So it is disposable spaceplane? I thought you are launching it vertically. In the case, single divertless intake will work decently. (As speed gained with airbreathing stage will be near 1km/s)

[rcp27]

16 hours ago, Abastro said:

So, you are going for highly inefficient trajectory due to the TWR of circularizing stage. If you don't have enough TWR to circularize(spends too long time), then you are doing it wrong.

In detail, You should be losing over 600m/s of dv doing inefficient gravity turn, and lose another 300m/s performing circularization burn with pitch of 30 degrees. That's nearly 1km/s of lose!

Also, 50% is way down for Thumper. You may as well use Hammer.

So, you are going for highly inefficient trajectory due to the TWR of circularizing stage. If you don't have enough TWR to circularize(spends too long time), then you are doing it wrong.

In detail, You should be losing over 600m/s of dv doing inefficient gravity turn, and lose another 300m/s performing circularization burn with pitch of 30 degrees. That's nearly 1km/s of lose!

Also, 50% is way down for Thumper. You may as well use Hammer.

 

The problem with this analysis is it doesn't consider actually getting out of the atmosphere.  If you go with a single SRB as a launch stage, you have to accept it's sub-optimal burn profile (constant thrust until the fuel is gone) and the need to get out of the thermally and aerodynamically challenging part of the atmosphere.  I haven't chosen the Thumper for either it's thrust or it's dV, but for the fact that it can get out of the lower atmosphere.  I've just had a go at a basic test: I put a HECS probe core on a Hammer, nothing else.  That was a total failure: it's aerodynamically unstable.  I added 3 fins.  After a reasonably involved test campaign with different thrust settings and attempted launch trajectories, I never managed more than 22 km.  I'll grant that I didn't find the optimum, but I reckon 25 km is the maximum you'll get out of a Hammer.

 

I don't approve of "explosive decoupling, but for The Sake of Science^TM I had a go with a Hammer + Flea combo.  Result: no decoupling  As a result of extensive testing, all I can presume is that this "technique" has been removed from the game.  Anyway, I moved on to Hammer, TR18-A decoupler and Flea as first and second stage, with a simple HECS in control.  I did manage to achieve sub-orbital flight with an Ap of 99 km as my best shot, but the Flea + Hammer + decoupler costs 1000 while a single Thumper is 850, so I'm not sure where the benefit comes from.

 

16 hours ago, Abastro said:

 

Let's calculate payload mass and lifter cost here.

Payload: Antenna(0.005t)+Probe(0.07t)+2xSolar Panels(0.035t)+Thermometer(0.005t)+Battery(0.01t)+Reaction Wheel(0.05t)+1 OscarB fuel tank(0.225t) = 0.4t

Lifter: Ant(110)+2xOscarB(140)+Decoupler(300)+Fairing(300+)+Thumper(850)+3xBasic Fin(240) = 1940

So you are using 4850 per 1t of payload! AFAIK you can use less money to lift 1t.

I think these improvements will cut down your cost to target orbit decently:

1. Fairing is too expensive and heavy for such a small payload. It'll cost over 300 and heavier than 0.075t, since cost and mass of deployed parts is added to the base cost/mass. If you are worried about aerodynamic stabilization, just put small nose cone(180, 0.01t). (You won't need it)

 (If it is for aesthetis, let me know, as it is reasonable choice for the purpose)

   + If you think aerodynamics is a thing, I can say it's definitely not. Definitely, the loss is smaller than 100m/s.

2. If you are aiming to do circularization burn with single Ant engine, just use single OscarB tank to get more TWR. Here are some calculations to verify this:

   Now, your dry mass of second stage is 0.27t, and its wet mass is 0.87t, thus it has dv of 3.6km/s. When you are on LKO with an OscarB tank fuel to spare, the wet mass of the stage is 0.47t. So you have spare dV of 1.7km/s. AFAIK that's enough to get it to Duna and Eve! Dv of 1.2km/s will be enough. (1km/s lets you escape kerbin)

  When you have single OscarB tank, the second stage has dry mass of 0.22t and wet mass of 0.42t. Thus initial TWR of 0.485 and dV of 2km/s. It'll be enough to do efficient circularizing burn(TWR increases over time). Considering 1km/s you wasted on gravity turn, and 0.6km/s of unnecessary dv, this one is only 0.1km/s short! Also the reduced dry mass will allow more dV for the first stage.

 

When I first created this concept, I didn't want to be ridiculously unrealistic, so I used a fairing, but experiments suggest it's important anyway (see below).

 

16 hours ago, Abastro said:

3. You are fine with single Hammer, as its dv is not that short! With the payload above, you need to launch 0.435t (including decoupler).

  Single Hammer has atm dv of 170*9.81*ln(3.995/1.185) = 2km/s. If you are doing efficient gravity turn, 2.8~3km/s is sufficient while going eastward. Thus 1~1.2km/s of dV will be left. It could be a bit short with westward orbit. If you want some more dv, just put something like Flea under Hammer, and explode it away.

+ If you want more dV, 2 Hammer with explosive decoupling will work. The lower stage gives dV of 775m/s more. The so the two Hammer has total dV budget of 2.7~2.8km/s with the payload, which can make orbit themselves! It could be a bit unstable, so some dV will be lost. Though it'll be still enough.

+ Probodobodyne HECS costs 650 and weigh 0.1t. So it's cheaper and lighter than 1 small reaction wheel + 1 QBE probe! The torque should be enough to turn such a small craft. (Also this will give you more spare dV) Also, why do you need battery?

 

Your points on optimising the orbiter are good ones.  So I have run some tests as follows: Communotron 16, HECS, 3x OX Stat, 1x Oscar B, Ant, TR-2V decoupler, Thumper, 3x Basic fins.  I found the result became unstable in the transonic regime.  I'm not sure if this is due to aerodynamic forces or mass distribution.  I added the 1.25 m fairing back in, and the result flew nicely. I found that if I flew a trajectory that resulted in a Thumper burn-out with an Ap of less than about 200 km (so more horizontal velocity), atmospheric heating becomes high enough to destroy parts (the fins all burn off, but the fairing protects the orbiter). The result is a successful orbit, but without the external battery the satellite has insufficient energy storage to actually transmit a temperature reading, so can't manage to bring any science home.  Anyway, that launch cost me 3748 total, so that is a new personal best for "cost to orbit" for a minimum orbiter.

[Reusables]

4 hours ago, rcp27 said:

The problem with this analysis is it doesn't consider actually getting out of the atmosphere.  If you go with a single SRB as a launch stage, you have to accept it's sub-optimal burn profile (constant thrust until the fuel is gone) and the need to get out of the thermally and aerodynamically challenging part of the atmosphere.  I haven't chosen the Thumper for either it's thrust or it's dV, but for the fact that it can get out of the lower atmosphere.  I've just had a go at a basic test: I put a HECS probe core on a Hammer, nothing else.  That was a total failure: it's aerodynamically unstable.  I added 3 fins.  After a reasonably involved test campaign with different thrust settings and attempted launch trajectories, I never managed more than 22 km.  I'll grant that I didn't find the optimum, but I reckon 25 km is the maximum you'll get out of a Hammer.

 

I don't approve of "explosive decoupling, but for The Sake of Science^TM I had a go with a Hammer + Flea combo.  Result: no decoupling  As a result of extensive testing, all I can presume is that this "technique" has been removed from the game.  Anyway, I moved on to Hammer, TR18-A decoupler and Flea as first and second stage, with a simple HECS in control.  I did manage to achieve sub-orbital flight with an Ap of 99 km as my best shot, but the Flea + Hammer + decoupler costs 1000 while a single Thumper is 850, so I'm not sure where the benefit comes from.

 

So aerodynamics seems to be a problem with that. AFAIK 2km/s is sufficient to get to upper atmosphere witha reasonable TWR. (I got out of atmosphere with 1.6km/s)

(Edited) I just recognized that Ant engine doesn't have aerodynamic cover, so it needs to be in the fairing. In this case, fairing is fairly better than using spark

About the gravity turn, it works with constant thrust. I think no one actively control thrust to get to orbit, usually.

Also didn't know that explosive decoupling doesn't work now. 

 

4 hours ago, rcp27 said:

When I first created this concept, I didn't want to be ridiculously unrealistic, so I used a fairing, but experiments suggest it's important anyway (see below).

 

Your points on optimising the orbiter are good ones.  So I have run some tests as follows: Communotron 16, HECS, 3x OX Stat, 1x Oscar B, Ant, TR-2V decoupler, Thumper, 3x Basic fins.  I found the result became unstable in the transonic regime.  I'm not sure if this is due to aerodynamic forces or mass distribution.  I added the 1.25 m fairing back in, and the result flew nicely. I found that if I flew a trajectory that resulted in a Thumper burn-out with an Ap of less than about 200 km (so more horizontal velocity), atmospheric heating becomes high enough to destroy parts (the fins all burn off, but the fairing protects the orbiter). The result is a successful orbit, but without the external battery the satellite has insufficient energy storage to actually transmit a temperature reading, so can't manage to bring any science home.  Anyway, that launch cost me 3748 total, so that is a new personal best for "cost to orbit" for a minimum orbiter.

 

Great, you got the cost reduction!

You are right about the fairing, I didn't know that Ant engine doesn't have any aerodynamic cover.

Aside from that, the basic fins have low maximum temperature of under 1000K, and the other has max. temp. over 1200K. So you'll be fine with thermal issues here. Try a bit more aggressive turn.

I think if you want to get the experiment, Z-200 is the best one without fairing. It's lightweight, so won't affect performance too much.

I want to see what your gravity turn is like. Would you post some pictures of it?

Edited March 4, 2017 by Reusables
Some realizations

[Reusables]

@rcp27, So I've made a craft which can get to orbit, with thermometer and mk0 battery to transmit the data.

The first stage is Hammer,

Total cost is 3646, and lifter cost is 1211.

Album: http://imgur.com/a/lVUAP

Craft file

Edited March 4, 2017 by Reusables
Added the pic.

[rcp27]

10 hours ago, Abastro said:

Aside from that, the basic fins have low maximum temperature of under 1000K, and the other has max. temp. over 1200K. So you'll be fine with thermal issues here. Try a bit more aggressive turn.

 

By the time the fins burned off the craft was flying in a straight line anyway, and as they were attached to the discarded booster, they aren't really that important.

 

5 hours ago, Abastro said:

So I've made a craft which can get to orbit, with thermometer and mk0 battery to transmit the data.

The first stage is Hammer,

 

Right, just built a HECS, 1 Oscar B, 1 Ant, 1 Z-100 Battery, 1 Communotron 16 and 2 OS-Stat solar panels craft and put it on a Hammer under a fairing with a TR-2V decoupler and achieved orbit.  Mine only cost 3298.

 

The experienced KSP players will be shouting at their computers "fix your staging!" around about now.

 

 

So the question is why did this work when my earlier HECS + Hammer tests went so badly wrong? The answer, it appears, is aerodynamics.  I did a few test flights  All had the Hammer at 40% and the flight trajectory was straight up.  Hammer + HECS: unstable.  Hammer + HECS + 3 fins reached a little over 20 km.  Hamm, HECS, 3 fins and a 0.625 m nose cone: a bit better, reached 26 km.  Hammer + HECS + 3 fins + 1.25 m nose cone.  Similar, under 30 km.  Hammer, Hecs, 3 fins, 1.25 m fairing: holy cow!  Not only does this thing go up much much faster than the others, with the engine not burning out till over 40 km, it leaves the atmosphere with an Ap of over 280 km.  Also, the drag reduction at the front means the fins can be ditched and a stable flight maintained.  So I conclude that we all need to forget the old days of "fairings are just for aesthetics".

[Reusables]

10 minutes ago, rcp27 said:

snip

So it works with radial battery! That's really cheap craft.

By the way, I think it could work with small nosecone & FL-A10 Adapter without fairing, if Ant had casing. But it doesn't, which results in the blunt end generating too much drag. If you are fine with offset, putting spider with the above configuration could work. There's no need for that, though.

Also, the drag matters on the transonic region, which is why TWR is critical. (Still, gravity loss tends to be way bigger)

Edited March 4, 2017 by Reusables

[rcp27]

6 minutes ago, Abastro said:

So it works with radial battery! That's really cheap craft.

My understanding is the radial battery, like the thermometer, just adds its mass to the part it's attached to, so doesn't unbalance the craft.  It's inside the fairing (just), so doesn't have any aerodynamic issues.  And it contains enough charge to transmit the all important science even on the night side of Kerbin.

I've had a bit of a play around, and I think the key issue is the step from 0.625 m parts to the 1.25 m Hammer.  I just launched a satellite using the 1.25 to 0.625 m structural part and a 0.625 m nose cone instead of a fairing.  While the radial battery is OK from a mass distribution point of view, it's no good aerodynamically, so I had to use an inline one instead.  I achieved orbit, but with a wafer thin fuel margin (less than 1 unit left in a less than 80 km orbit) and it cost more than the last one.

[Reusables]

5 minutes ago, rcp27 said:

My understanding is the radial battery, like the thermometer, just adds its mass to the part it's attached to, so doesn't unbalance the craft.  It's inside the fairing (just), so doesn't have any aerodynamic issues.  And it contains enough charge to transmit the all important science even on the night side of Kerbin.

I've had a bit of a play around, and I think the key issue is the step from 0.625 m parts to the 1.25 m Hammer.  I just launched a satellite using the 1.25 to 0.625 m structural part and a 0.625 m nose cone instead of a fairing.  While the radial battery is OK from a mass distribution point of view, it's no good aerodynamically, so I had to use an inline one instead.  I achieved orbit, but with a wafer thin fuel margin (less than 1 unit left in a less than 80 km orbit) and it cost more than the last one.

What engine did you use for the craft without fairing?

[LordCorwin]

The cheapest I've got is about 2000 $/Tm with a 2 stage rocket and a 14Tm payload, still far from expectations.

[rcp27]

2 hours ago, Abastro said:

What engine did you use for the craft without fairing?

Same old Ant.  I have a suspicion that the Ant may have its physics modelled as if it did have a fairing even if one doesn't show.

For the sake of completeness, I thought I'd see if I could do any better in the SPH.  I knocked together this little spaceplane

  

That thing in the middle is a diverterless intake that also holds liquid fuel for the whiplash, and there are a pair of Sparks each with 3 Oscar Bs on the sides.  The payload, at the front, is the same orbiter as before, still with its ant and single oscar B.  I couldn't get the plane to orbit, but I achieved a suborbital flight with an Ap a little over 85 km.  I ditched the fairing and launched the payload near the edge of space on the way up, took control of the payload and flew it to orbit while the plane began its re-entry.  I achieved orbit with he plane still in the 60's, went back to the plane and attempted to fly it back to base.  Because it was suborbital, re-entry heating was no problem, but on the return flight I ran out of fuel a little to the east of the islands with the island runway, so had to ditch in the ocean (which I did successfully with no damage).  I reckon with bit of optimising and practice I could get something for that kind of price that would make it back to the runway.  For such a light plane with a whiplash on the back it accelerates like you wouldn't believe.  I set the undercarriage so that it sits with a nose-up attitude on the runway, and it practically flies itself.  Obviously it only pays off if you get the plane recovery money.

[Reusables]

10 minutes ago, rcp27 said:

Same old Ant.  I have a suspicion that the Ant may have its physics modelled as if it did have a fairing even if one doesn't show.

I found that it makes the attached nodes draggy like a blunt edge. I used a debug menu(Alt+F12) option of displaying the drag of each part. Radial spider may work better with offset.

14 minutes ago, rcp27 said:

For the sake of completeness, I thought I'd see if I could do any better in the SPH.  I knocked together this little spaceplane.

That's great idea! I'm curious if spaceplane can still beat rockets with these small payload.

So how much does the fuel cost?