Heavy rockets

[GoSlash27]

Gorebello,

 If I read your OP correctly, you only want enough DV to get you a Jool intercept.

I call it 3500 m/sec to Kerbin orbit and 1,935 m/sec to transfer to Jool (plus another 300 for midcourse correction and reserve), which would bring up the transfer stage to 2235 m/sec.

If I were to design a lifter to do this in one shot (I wouldn't... but if I did) it would be a 3 stage design.

Stage 3: 36t payload, minimum t/w 0.5, DV= 2,235 m/sec. It could be jettisoned on a collision course with Jool.
7 LV-N, 82t total mass

Stage 2: 82t payload, minimum t/w 0.7, DV= 1,700 m/sec. This would get you to just barely short of LKO, so the stage could be jettisoned to deorbit on it's own.
5 Poodles, 170t total mass

Stage 1: 170t payload, Minimum t/w 1.2, DV= 1,800 m/sec.This would have all the aero fins and get you to 30km in a gravity turn.
3 Twin Boars, 425t total mass

If it were me, First off I'd try to reduce the tonnage of equipment I'm bringing to Jool. This would include ISRU, which would weigh a lot less than bringing fuel along for the return trip. If you're going to Jool, you'll be waiting there for 3 1/2 years for the return window. Might as well make use of it.

  Next, I'd plan to lift my ship in sections, assemble it in orbit, and fuel it in orbit. More launches, but much easier to manage.

Good luck!

-Slashy

 

 

[Temstar]

54 minutes ago, Geschosskopf said:

This loss of efficiency is due to vectors.  The object of the burn is to increase your "forward" velocity, but if you're not pointing forward (aka prograde) when you burn, you're only adding the component of the burn that's parallel to the prograde direction to your velocity.

Just always burn prograde and start your burn earlier, so that when your burn completes you hit the correct ejection angle in either prograde or retrograde Kerbin-Kerbol orbit. Either that or do multiple periapsis kicks.

Yes you lose some delta-V because some of the burn is done at higher orbit, but you also reduce your TWR requirement. Taken to the extreme you can get an orbit like this with electric propulsion in real life were the ejection burn takes months:


That said I personally favour TWR of between 0.3-0.5G. Any lower and the early burn time compensation becomes too much of a hit and miss.

 

Edited January 3, 2016 by Temstar

[Geschosskopf]

1 hour ago, Temstar said:

Just always burn prograde and start your burn earlier, so that when your burn completes you hit the correct ejection angle in either prograde or retrograde Kerbin-Kerbol orbit. Either that or do multiple periapsis kicks.

Yes you lose some delta-V because some of the burn is done at higher orbit, but you also reduce your TWR requirement. Taken to the extreme you can get an orbit like this with electric propulsion in real life were the ejection burn takes months:

.........

Actually, doing it this way is still inefficient for reasons besides Oberth.  You're still burning a significant amount of fuel in what are effectively the radial in/out and perhaps even retrograde directions, when compared to the ultimate direction of your departure vector.  It's just that ion probes have so little thrust they don't have a choice, and have so much dV they don't care.

Consider....  Assuming you start in a circular orbit, as soon as you start burning prograde at an arbitrary point, you create a new Pe where you are and a new Ap on the opposite side of the orbit.  As you continue to burn prograde, the Ap not only gets higher but also rotates around the orbit to stay in line with the Pe, which itself is sliding around following your ship.  IOW, the burn is not only raising your Ap, it's also changing your APe.  This rotation of the Ap/Pe is exactly the same effect you get with radial burns, and it happens here because you're directing some of your thrust in what is effectively the radial direction..

Now obviously, because all transfer burns take a measurable amount of time, you always have some of this APe change.  But changing APe does not in itself contribute to the desired departure vector, so any fuel used to move your Pe/Ap around the orbit is wasted.  Therefore, efficient burns need to minimize this effect.  You do that by having short burns at the desired location of your Pe so your Ap goes straight to the destination.  But an early burn maximizes the change of APe.

[AeroGav]

Since this sounded like a fun challenge, and I've never been to Jool, I decided to try build something to go there.      The Jool vehicle would have a Hitchiker Module, Mk2 cockpit, Science Lab, and every science instrument in the stock game that can be used from orbit.  Actually I wasted half the cargo bay on an orbital survey scanner which is useless on Jool, but perhaps could be used on the moons.

First attempt revealed that I was going to need the Kerbal Joint Reinforcement mod 

Undaunted,  Bill and Jeb try again,  5 Panthers, 4 Rapiers and 5 NERV.  

The Panthers are attached behind the NERV, and operate in mostly DRY mode, to get some of that sweet 9000 ISP.

 After takeoff, most pilots select GEAR UP.     We select GEAR DOWN,   all the way below ground level actually -

Shortly before 8km / mach 1, the pre-coolers that the RAPIERs and PANTHER are mounted to become empty.   This is not a problem as they are jets and drain evenly from every tank.  However I lock them at this point, to stop me refilling tanks that are going to get jettisoned when i turn on TAC FUEL BALANCER.

Afterburner is used from 280 to 430 m/s , along with a little nose down,  to get through sound barrier -

After that, I go back to dry mode, but by 14km they quit anyway.  Attempts at jettisoning spent engine stages revealed the need for separatrons -  otherwise you end up with holes in your airframe and going to the pool instead of Jool :

After working out how to shed the Panthers safely, it's time to activate the fuel balancer.  The NERVs which sit behind them, aren't as smart as jet engines and only pull fuel from the tank they're directly mounted to.  Without balancing they'll be out of gas immediately.

The RAPIERs are starting to fade, time to bring in the NERVs.

When they finally quit at 29.5KM,  I jettisoned them, and immediately discovered a flaw in my design.  Mounting them to the front wing caused my plane to immediately execute Pugachev's Cobra as soon as I punched them off,  somehow it got back under control again when I hit ALT - X to cancel all the pitch trim,  but i lost 200 m/s in the post-stall maneuver.     Still, we did gain a bunch of altitude and a huge upward vector

After zooming to 40km, we sag back down.  Optimal L/D ratio at subsonic is 2deg and 5 deg AoA at hypersonic, but right now i'm facing "pull up or blow up".   Used 7 - 10 in places..

"Finished with Engines".   Notice how our delta V immediately jumps from 1077 to 2036 from the loss of mass..

We make some burns raising our Apoapsis till the wing tanks are all empty, then they are discarded.      

Years later, we finally reach Jool.  Apparently it's only 2000 dv but i use over 3000 getting there from a highly elliptical orbit on Kerbin.  I'm no good at orbital mechanics.    Also, it appears I forgot the scientists...

[Temstar]

14 minutes ago, Geschosskopf said:

You're still burning a significant amount of fuel in what are effectively the radial in/out and perhaps even retrograde directions, when compared to the ultimate direction of your departure vector.

No I don't think that's correct, consider Newton's cannon firing on the Mun. If you fire the cannon ball out at horizontal to the surface at a mountain top at orbital velocity the cannonball with have an AP of X meters, and it suffered zero gravity loss (since the acceleration is instantaneous) and zero aerodynamic loss (since there's no atmosphere) to get that orbit.

Now instead of a cannonball out of a cannon, we launch a spacecraft horizontally from the surface from the same mountain top and follow the prograde. This spacecraft will take some non-zero time to reach the same orbital velocity as the cannonball, yet because it was launched horizontally and it always followed the prograde it also didn't suffer any gravity loss. Yes sure the spacecraft's AP is in a different location to the cannonball's AP, but that change in AP is not because of radial burn but rather because of gravity curving the trajectory differently under different flight perimeters.

It's the same with spiral burns - you don't suffer any gravity loss because you are always burning prograde and so unlike radial burn you never actually thrust against gravity.

[Snark]

1 hour ago, Temstar said:

No I don't think that's correct, consider Newton's cannon firing on the Mun. If you fire the cannon ball out at horizontal to the surface at a mountain top at orbital velocity the cannonball with have an AP of X meters, and it suffered zero gravity loss (since the acceleration is instantaneous) and zero aerodynamic loss (since there's no atmosphere) to get that orbit.

Now instead of a cannonball out of a cannon, we launch a spacecraft horizontally from the surface from the same mountain top and follow the prograde. This spacecraft will take some non-zero time to reach the same orbital velocity as the cannonball, yet because it was launched horizontally and it always followed the prograde it also didn't suffer any gravity loss. Yes sure the spacecraft's AP is in a different location to the cannonball's AP, but that change in AP is not because of radial burn but rather because of gravity curving the trajectory differently under different flight perimeters.

It's the same with spiral burns - you don't suffer any gravity loss because you are always burning prograde and so unlike radial burn you never actually thrust against gravity.

You're conflating "gravity loss" with "cosine loss" (added emphasis in quote above for clarity).  Those are different things.

• Cosine loss happens when you're burning off prograde (regardless of what direction you're traveling).
• Gravity loss happens when you're burning off horizontal.

Ideally, you'd like to avoid both.  The only way to do that is to burn prograde while traveling perfectly horizontally.  If you're not traveling horizontally, it's possible to avoid cosine loss (by pointing prograde) or gravity loss (by pointing horizontally), but not both at the same time.

Yes, prograde is good, to avoid cosine loss.  But even when you're firing perfectly prograde:  any time you are thrusting when you are not traveling perfectly horizontally, there is a gravity loss.  The amount of gravity loss is proportional to the sine of your thrust angle above horizontal.

Edited January 3, 2016 by Snark

[Temstar]

@Snark, I thought about what you've said and I have to agree.

Still, that means for a burn over an arc you have:

1. Always follow prograde - zero cosine loss, some gravity loss
2. Always follow horizon - zero gravity loss, some cosine loss
3. Always follow a point in between (eg node marker) - combination of cosine and gravity loss

I wonder if it's possible to do analysis on each of these, to see if the loss in all cases are equal or is there some local optimum solution.

Regardless for this particular case it may be easier to put the interplanetary ship into Low Mun Orbit. Fully fuel the stack in Mun orbit (via ISRU or otherwise) then when the transfer window comes up eject back into Kerbin with a low PE, reach the PE then complete the interplanetary burn. That way the outbound burn would only be around 1400m/s of delta-V divided into two relatively small burns, instead of the monster 1930m/s single burn directly from LKO.

[technion]

10 hours ago, Wanderfound said:

Actually, you just use aircraft LF tanks. Mk3 if you're going big, 1.25m if not.

The problem there is the Mk3 is a really dumb fit if the rest of your rocket is perfectly built around things like orange tanks. Adapters and such can add a lot of pointless weight, along with the aesthetics of it.

I used to be able to use one large grey tank and two LV-N's and get enough DV to capture at Eeloo and easily return. Just put it on an ejection stage and you're away. Options now are a lot more limited.

The game could really do with some mid-size jet fuel tanks.

[Geschosskopf]

1 hour ago, Temstar said:

I wonder if it's possible to do analysis on each of these, to see if the loss in all cases are equal or is there some local optimum solution.

Regardless for this particular case it may be easier to put the interplanetary ship into Low Mun Orbit. Fully fuel the stack in Mun orbit (via ISRU or otherwise) then when the transfer window comes up eject back into Kerbin with a low PE, reach the PE then complete the interplanetary burn. That way the outbound burn would only be around 1400m/s of delta-V divided into two relatively small burns, instead of the monster 1930m/s single burn directly from LKO.

The loss is minimal with the shortest possible burn at Pe.  That's why that's the recommended and standard proceedure :D.  Every other method is a makeshift to muddle through despite having insufficient thrust to do it right.

As to gong to Mun 1st, to refuel, this is just a makeshift for muddling through not having enough fuel in the transfer stage to start with.  If the ship has enough fuel for the transfer burn, then going to Mun to top up partway through the transfer brun simply means you carry a little extra fuel all the way to the destination, which you don't need because you had enough to start with.  And it doesn't save enough fuel on the transfer burn to make a significant reduction in the size of the original lifter.  While you will reach your Kerbin Pe at a higher speed so get more Oberth, and have less of a burn left so have a shorter burn time from that, you also have more mass than you'd otherwise have at that point in the burn so have a lower TWR.  The balance does show a slight efficiency advantage but it's only a few hundred m/s at most according to those who have tried it, so no big reduction in lifter size.

And for this slight advantage, you have to spend the time and money setting up and then using the whole Munar refueling system, which will never pay for itself given the hassle of using it and the small savings it provides.  So in the end, it's just not worth messing with.  Now in real life, there might be a point to refuelling off the Moon, but at the small scale of the KSP universe just makes it unremunerative.

Edited January 4, 2016 by Geschosskopf

[Temstar]

6 minutes ago, Geschosskopf said:

As to gong to Mun 1st, to refuel, this is just a makeshift for muddling through not having enough fuel in the transfer stage to start with. 

But you can look at that problem the other way around though, doing it this way cuts 500m/s off the delta-v budget. Since our payload is fairly big this cuts down the size of the transfer vehicle by a fair bit which further escalate down to the launch vehicle and so on.

ISRU setup doesn't have to mean a gigantic surface refinery. @Rune for example is working on an impressive tiny package where the tanker rocket double as delivery vehicle for the rover. You could very well have this package as part of the payload to Jool anyway and have it pull double duty landing on the Mun first to refuel.

That said the point is moot for me because I do have both a gigantic surface refinery and huge orbital propellant depot because I find watching the drills working and hording ore therapeutic.

[Wanderfound]

Something like this:

 

 

Craft file at https://www.dropbox.com/s/e01jsfkraj6ubql/Mothership.craft?dl=0

[Wanderfound]

And a little something to hang on the towbar:

 

Craft file at https://www.dropbox.com/s/38quq2zxwuzyry4/Miner.craft?dl=0

[Warzouz]

Ah, going to Jool : probably the nicest system to visit, you can se several moons mostly any time with the naked eye ! Exploring this system takes a long time and provide a lot of fun ! I woiuld recommend to set a permanent base there.

Your rocket is too big and you might have some issue with going to LKO. Your ship seems also too big and have redundant parts. As others said, "send empty tank at LKO and refuel there"

I would prefer to say "send empty tanks to Jool an refuel there" This is the whole concept of my Salamander Exploration Space Station : Use ISRU ! Burn all you station fuel during the interplanetary transfer and arrive there (mostly) dry. Then refuel at your target (or at Pol, it's easy)

As for TWR and dV. Sure, dV is the most important, but TWR is about comfort. Planning interplanetary transfer with a reasonable TWR is much easier than with a low TWR. But to go to Jool, that's not really an issue because the SOI is very big, you can't miss it. It's more of an issue for Moho, Dres or Eeloo. Having a too low TWR may cost you dV.

Don't hesitate to send multiple missions instead of a big one. Sending 2 Tons into Jool will cost twice sending 1 ton crafts (except for electric an control parts).

 

 

[Geschosskopf]

20 hours ago, Temstar said:

But you can look at that problem the other way around though, doing it this way cuts 500m/s off the delta-v budget. Since our payload is fairly big this cuts down the size of the transfer vehicle by a fair bit which further escalate down to the launch vehicle and so on.

ISRU setup doesn't have to mean a gigantic surface refinery. @Rune for example is working on an impressive tiny package where the tanker rocket double as delivery vehicle for the rover. You could very well have this package as part of the payload to Jool anyway and have it pull double duty landing on the Mun first to refuel.

That said the point is moot for me because I do have both a gigantic surface refinery and huge orbital propellant depot because I find watching the drills working and hording ore therapeutic.

500m/s more or less (and that's being generous to the Mun refueling option) in the rransfer stage isn't going to make or break the transfer stage or the lifter under it.  It's chump change, barely half what it takes just to reach Mun.  So going to great lengths (which are unavoidable with any refueling option) to save that little bit by refueling at Mun is definitely worth neither the expensive of setting up for that nor the trouble of actually doing it.

Remember, a transfer stage with about  0.7 TWR going to Jool  is over and done with in 5-6 minutes of your own real, personal time.  So that's the standard to compare other options to when factoring in the intangible value you place on your own time.  And you only need to spend money on that 1 ship.  OTOH, refueling at Mun will take at least an hour of your own, real, personal time just to use once, plus the money to create a refueling system and the time to get it set up on Mun.  And if topping off the transfer stage takes more than 1 tanker trip from Mun, you've wasted even more time.  And after proving this to yoursefl once, you'll never use it again, writing off the in-game cost of building the system as worth less than your real life time.  Oh, and you only save like 500m/s (AT BEST--plan for rather less)  doing all this, so even if you force yourself to keep doing it a few more times, you'll still never pay it off.

The bottom line is, as has been the case since Kethane (which goes back before 0.20) and through Karbonite and now into Ore, setting up a refueling system anywhere within Kerbin's SOI just ain't worth it in any real, measurable terms, whether in-game or in real life.  The KSP universe is just too small.  Now, if it has roleplay value for you, or you do RSS or something, fine.  But in the stock-size KSP, measured in funds and your own real time, no way.

[Admac]

 

On 1/5/2016 at 4:44 PM, Geschosskopf said:

Setting up a refueling system anywhere within Kerbin's SOI just ain't worth it

There are lots of good use cases for in system fuel mining!

I use boost tankers for all my serious business missions. Granted I have USI Kolonization mods with that handy logistics hub to get the fuel from my surface base to the orbiting ships, saving time at a cost of more fuel.

But I sent a reusable ship to haul 200 tons from Kerbin to Moho and back a few times as a part of the Grand Orbital Station challenge, and in space refueling is a billion times easier than lugging it up from the surface.

 

Back to the Jool question, I agree with the consensus here, you are really building 3 ships;

1) Launch vehicle with at least 3400 DV and starting TWR of 1.2 or more

2) A transfer vehicle with 4000ish DV and a TWR of 0.2 or greater

3) A payload to have fun at Jool. as light as possible. Your satellites should be closer to 0.5 tons each, that should be plenty for experiments and an RTG and lots of fuel.

All these parts can be made reusable too, if you are feeling hardcore and don't mind making them all a bit bigger!

[Geschosskopf]

5 hours ago, Admac said:

There are lots of good use cases for in system fuel mining!

I use boost tankers for all my serious business missions. Granted I have USI Kolonization mods with that handy logistics hub to get the fuel from my surface base to the orbiting ships, saving time at a cost of more fuel.

If you have the Magic Star Trek Resource Transporter (MSTRT, aka MKS Logistics Hub), then you avoid the major issue of an Kerbin-system refueling operation, which is the vast amount of time required flying tankers around to use it.  Even before career mode existed so money wasn't an issue, this amount of time, compared to the tiny amount of transfer dV savings you get from it, made Kerbin-system refueling impractical.  But if you have the MSTRT so your time investment is much less, you've still only got a tiny dV savings.  And if you can afford to set up the MSTRT system and build enough other ships to use it frequently, you're already so rich that the tiny dV savings is worth even less of your time than it was without the MSTRT.  By this point in the game, you're spending millions on major projects so don't notice a few tens of thousands more or less.

[Temstar]

It doesn't take that long, I can get pretty much any craft up from the Mun surface to hard dock with in 5-10 minutes. For Mun TWR of something like 2 just wait for target is about 5 degrees over your head in orbit and launch. Get to the target AP, warp there and then match velocity. I've done so many up and down trips with landers now most of the time I can launch to my Mun station just via eye ball and end up with in 500m of each other right at the lander's AP.

The important thing is to get a feel for these things so you don't have to mess with setting up nodes and all that nonsense. Just aim and shoot, even if you miss by a little bit Mun orbit is so forgiving it's trivial to manually correct to get a closest approach - target in front of you at closest approach, 45 degree prograde radial in burn; target behind you at cloest approach, 45 degree retrograde radial out burn.

Edited January 6, 2016 by Temstar

[Geschosskopf]

10 hours ago, Temstar said:

It doesn't take that long, ......

Hey, if that's how you enjoy the game, knock yourself out.  But you can't make a rational argument that justifies doing it in terms of the in-game economy or compared to the real time required to NOT do it.  This is not an opinion, it's been hashed and rehashed for years by smarter people than me, concensus has been reached, and nothing in the equations have changed enough to make a difference despite ISRU fuel going from Kethane to Karbonite to Ore.  Any perceived benefits are illusory due to using the wrong frame of reference to measure from.  But that doesn't take the irrational enjoyment factor into account.  If you enjoy it, that's all the justification it needs.

[Plusck]

This topic sprung to mind last night as I completed a painful transfer burn to Jool with a 36t payload.

The whole ship was an ugly mess:

Spoiler

The Vectors were expensive overkill. They only got there because I had copied the basis of a cheap lifter cluster from another ship, added more fuel, and then found it was just slightly too weak with the original Reliants. Rather than completely change the lifter stage (which I should have done) I got fed up and just splashed out on Vectors to get the damn job done.

More pics at http://imgur.com/a/AJbkf

The transfer stage had a starting TWR of 0.10. Ugh. So the burn had to be done over two orbits, the first using the last of the lifter stage (plus some fuel from the payload, so the final payload is actually more like 33t) and the second an excruciating 13 minutes of mostly prograde burning. After a failed first try, the manouvre node for the first orbit burn was deliberately set too early in the orbit: that meant that the last 600 m/s or so of the second burn, which unavoidably continued well past Kerbin Pe, ended up giving a decent ejection trajectory.

In my defence, I launched the ship without KER installed so I gave the transfer stage too much fuel (to be safe) and not enough TWR (because I didn't think it was so low).

[Wanderfound]

Nuke and conventional rockets working well in combo...

Made the burn in a single not-too-painful push, dumping all of the drop tanks along the way. Didn't even have to start too high; just barely grazed the atmosphere at periapsis, despite beginning the burn pointed almost straight down.

[Snark]

14 hours ago, Wanderfound said:

Nuke and conventional rockets working well in combo...

<screenshots>

Nice use in tandem!

Are those Gigantor solar arrays on the mining ship?  Not much solar power out at Jool... might be more practical to give it a couple of RTGs to maintain charge in normal operations, then fuel cells to supply the heavy load once you sink your teeth into a juicy ore deposit.

[MaxL_1023]

For really large rockets you want your TWR to be closer to 1.25 at launch and 1 above 30km - air drag becomes a real issue if you are building radially. 

I have launched probes in this size range by mounting the transfer stage serially below the probe (usually using chemical engines - the weight and low thrust of LV-Ns requires a very high fuel fraction to take advantage of the ISP with absolutely terrible TWR), then using girders to attach 4 copies of a heavy-lift rocket forming a square. You have tons of control authority by the displaced thrust locations and can use cross-struts to reinforce the design. 

Usually, I would use something like:

36 ton payload

Transfer stage using a Rhino/Mainsail/Skipper depending on size required

Second stage using 4 Rhinos in 4 radial stacks (If your Transfer stage is large enough to need a Rhino, you might want to do 6 stacks instead of 4 in a hexagon configuration). TWR ~ 1.1

First Stage using 4 (or 6) Mammoths in sustainer configuration with radial SRBs (Liftoff TWR near 0.9, 1.5 including solids, sustainer TWR is ~1.1 when solids burn out.)

This type of configuration can get almost anything anywhere.

[Wanderfound]

8 hours ago, Snark said:

Nice use in tandem!

Are those Gigantor solar arrays on the mining ship?  Not much solar power out at Jool... might be more practical to give it a couple of RTGs to maintain charge in normal operations, then fuel cells to supply the heavy load once you sink your teeth into a juicy ore deposit.

I didn't even think about solar distance. And I normally use fuel cells, too, because most of my miners are spaceplanes. But this time I went for solar, because the miner isn't designed for atmospheric flight.

Dangit.