The CPU challenge

[Synapse]

Hey forum,

I designed the following craft for a one-way trip to Laythe:

However I’ve ran into a huge FPS problem trying to launch the thing. I’m getting 1 frame per second on my laptop and am wasting loads of fuel because I can’t get the throttle right etc. I’ve essentially given up launching this thing, so here’s where you come in!

I thought I’d post the .craft file here, and see what you can do with it. This is an open challenge to do the most impressive thing you can think of with this craft. You’ve got a nuclear engine with loads of fuel, 2 aerospikes, 5 mainsails and 12 SRBs at your disposal, as well as 2 detachable landing craft. (I haven’t rigged the staging up correctly for these, so you’ll need to detach them manually.

So, show me what this thing can do!

File: https://www.dropbox.com/s/8q65s3p4p7az9ym/Synapse%20Eden%201.craft

[Synapse]

Here is what happens when I try to launch:

[Tw1]

I say you need a better computer. Flying it now.

[Synapse]

Bonus points if you send me a save file either in LKO or near Jool so I can visit Laythe myself.

My CPU can handle anything up to Duna type missions and a small space station, but nothing like this.

[Scarecrow88]

I had a quick go and have a couple of observations.

Firing the solid boosters only at launch means you are carrying the dead weight of all those mainsails - why not add them to fire at launch rather than carry them until the solid boosters expire.

The RCS on your upper stage is slightly unbalanced, and the control thrusters on the booster tanks are not required. Once you have ditched the lower stages, the command capsule is pretty agile on torque alone, so you may not need any RCS, but if you must, change the tank on top for a couple of small round tanks attached to the main fuel tank.

Once your aerospikes have used up all their fuel and you jettison the tanks, you also jettison a load of unused RCS fuel.

Rather than have an RCS tank and fairing on top of your command capsule, I would rather have a docking port. I would also add parachutes and a decoupler above your ASAS. This at least gives your Kerbals a chance of landing or being rescued if it goes wrong. Currently you only have the two single man capsules capable of landing anywhere which would leave one Kerbal with no options of going anywhere other than to EVA over to a rescue ship.

The command craft can get a bit unstable after you have dropped one of the landers, due to the unbalanced mass, but it's not impossible.

The one man lander suffers a bit more instability, due to the remains of the decoupler leaving it unbalanced, and I struggled to keep it under control during landing.

Other than that, I think the craft has a fair bit of potential, so I hope you find a way round your lag issues so you can try it yourself.

[Synapse]

The one man lander suffers a bit more instability, due to the remains of the decoupler leaving it unbalanced, and I struggled to keep it under control during landing.

I've noticed this before when I did a similar Mun-Minmus mission, and I'm wondering what the solution is to balancing those mini vessels?

I really don't think I'll be able to launch this, even if I remove the RCS and some other parts.

Edited July 16, 2013 by Synapse

[Tw1]

I'd also add that without separatrons on your boosters, the second pair will crash into the core stage, should you forget to spin.

I'd also add a grey tank below each of the orange ones, to reduce the chance of overheating.

The lopsided parachute placement on your landers is not going to help if you want them to land straight, but it should be survivable. A lack of ladders may be inconvenient.

Still testing, but it looks like you have the delta v to get to laythe. I'm just going to do a Mun and Munmus tour with it though.

Edited July 16, 2013 by Tw1

[Darren9]

I'd probably get Hyperedit, remove the ascent package and put it in LKO - you should be able to fly the mission yourself from there.

[Synapse]

Can't you just send me a save file?

[Tw1]

Here it is, in a 70 x 71 Kerbin parking orbit: https://www.dropbox.com/s/5scoxujaktojud5/persistent.sfs

I made one small change before launch, to make better use of the aerospike boosters.

No idea how long it is until a Jool window though.

Edited July 16, 2013 by Tw1

[Synapse]

Thanks so much Tw1!

[Tw1]

Seeing as this was technically a challenge, this is what I did with it:

1: Launch it from KSC.

Jeb, Bob and Bill, the most famous kerbalnaughts ever, were tasked with testing a rocket from another space program, which had been having a few problems with their launch site.

Jeb: "Is this thing even moving? I can hardly feel it. Where's my acceleration?"

Bob: "This is how spaceflight should be."

I had added two parts- kerbal engineer for some delta V stats, and an RTG for power, I use a life support mod that needs it.

2. Go to the Mun.

Got it into a nice low orbit for landing. Then detached a lander which had Jeb in it.

3. Land on the Mun and return to orbit.

As Scarecrow88 said, the landers are hard to keep steady. But Jeb likes a challenge. Wreasling the controls to keep the lander pointing the right way, Jeb slowly decreased the horizontal velocity, then the vertical velocity.. And landed.

...In a rock. Yeah.

Jeb hoped out to confirm that he had in fact landed.

16 hours later, a Minmus transfer window would open.

At that time, Jeb returned to space, rendezvoused with the ship, then Eva'd across.

That lander would be left in Mun orbit.

4. Land on Minmus.

Despite looking very lopsided, the vessel had enough torque to hold its course during its transfer, correction, and capture burns.

They ended up in a low orbit around Minmus. This would be Bill's turn. As it requires very little thrust to land on Minmus, controlling the lander was easier. Bill set it down in a place he later named Synapse Saddle, after the ship.

18 hours of science and skiing (mostly skiing) later, it was time to return.

5. Return to Kerbin!?

Seeing as all the parachutes were on the landers, which had been left behind, and there definitely wasn't the TWR to attempt a powered landing, the three had a slight problem.

But this problem had a solution. A day or so later, they rendezvoused with a shuttle KSC had put in orbit remotely.

After EVAing across with the ship's memory banks and the rock samples, it was time to make the deorbit burn. The Synapse-Eden 1 was to be left in Kerbin orbit indefinitely.

Jeb: "Wasn't too bad a ship"

Said Jeb, as they set up the burn with the shuttle's advanced telemetry.

20 minutes later, they would be back at KSC.

The landing overshot the space centre slightly, but there was enough fuel to turn around and head back. Bumping over the runway at 12m/s, they parked next to the launch pad.

There, they posed next to the launch clamps that had held the rocket that took them into the sky 5 days, 1 hour, and 45 minutes earlier.

As I said earlier, it's a pretty capable craft, and should get you to Jool, and if you areobrake right, getting to Laythe should go fine. Accurate landing may be tricky, but kerbals can swim, right?

That's your delta V carrying both landers. Should be plenty.

Edited July 16, 2013 by Tw1

[koshelenkovv]

Hey forum,

I designed the following craft for a one-way trip to Laythe:

It can do almost anything if you throw away excess parts.

[Tw1]

....Do it win?

[codepants]

A better alternative to your current decoupler setup would be to reverse the couplers. EG, right now they detach from the main unit and are left on the landers. Try having them detach from the landers and remain on the main unit. Then the landers would have less off-center mass and the main unit would have more (counterbalancing the undetached lander).

You could also consider mounting radial parachutes on them which would allow docking ports on the heads, then docking them tangentially to the main unit. It wouldn't be pretty, but they would keep their balance after detachment. To balance the main unit in this scenario, add an additional docking port on top. When one lander leaves, move the other one from the side to the top. Although, at this point you can re-dock the detached lander before moving the main unit... assuming you use whichever one has RCS, as you'll need to translate.

....Do it win?

[Tw1]

-snip-

Good enough for me

[Eric S]

A better alternative to your current decoupler setup would be to reverse the couplers.

Radial decouplers aren't easy to reverse, if that's what you're saying. The structural pylon, however, lets go of both objects. Or at least did last time I used one.

[Nao]

@OP If you have CPU limitations (like me) designing for low engine (and part) numbers would be top priority. For example you can get that craft to orbit on just 3 mainsails and a lot less parts, like this:

Your ship dropped me to 15-20 FPS at launch and after this modification i had 25-30. It has more Dv, uses 1/3 less parts and has only 3 engines burning at the same time. (Most of the FPS loss come from SRB's)

My (actually they are not mine, i'm a dirty thief, i will not steal the shark thou! Tw1 well earned his hi5).

[Dem Wot Goz Ub]

Halfway through hunting for the shark:

Check it out!

I'm in Laythe orbit now. Waiting until tomorrow to try to drop a can or two.

Edited July 19, 2013 by Dem Wot Goz Ub

[Synapse]

I'm on my way to Laythe in a new ship, you were right that SRBs for some reason lag everything up. I used 5 mainsails and 4 aerospikes this time to get it into orbit.

Dat 30 minute burn...

[Dem Wot Goz Ub]

I'm on my way to Laythe in a new ship, you were right that SRBs for some reason lag everything up. I used 5 mainsails and 4 aerospikes this time to get it into orbit.

Dat 30 minute burn...

alt+> three times will make it a 2.5 minute burn. Just don't try to turn while you're doing it.

[Tw1]

Dat 30 minute burn...

Tip: Break long escape burns up for efficiency, and to releve boredom. This is called the periapsis kicks technique.

I described my method in another thread:

It takes a little practise. I'll describe my technique.

I use protractor and Kerbal alarm clock, which makes timing a little easier.

I launch the whole thing, and complete any docking several days before the window. This gives a good margin for error. Then, I do a burn at the ideal point, stating about 5 degrees before, and ending about 5 degrees after, give or take.

Before long, that point becomes the periapsis.

The trick is to use the time until periapsis to time when you do the final burn. Watch that, and the time to the ideal alignment carefully. Normally, the perfect transfer time and the time you arrive at the burn point are of by a bit, but it's still close enough to be accurate.

Then, as long as you timed it well, and don't accidentally escape before you meant to, you can use a node to make a smaller burn out to your target.

Edited July 20, 2013 by Tw1