Best way to stabilize stages with multiple engines?

[Grant_]

Right now the rocket has 5 Mainsail, 4 Skipper and 3 Atomic Engine. Unfortunately the Atomic engine section is making it very unstable....

I tried some girders but it didn't fit perfectly and didn't seem to help. Is there a better way besides adding as many struts as possible?

Thanks

Edited December 18, 2013 by Grant_

[Fett2oo5]

I'll take a guess that the connection failure is at that 1.25m fuel tank? You've got a lot of weight balancing on a "toothpick" between your larger fuel tanks.

Baring a significant design change, try using the same size tanks throughout.

If you are set on using this fuel tank configuration connect the 2.5m tanks that are on top and bottom of the 1.25m tank with 6-8 symmetrical struts. That might solve the immediate problem.

Of the two above, I suggest using 2.5m tanks for the entire craft, as you have with most of your design.

Edited December 16, 2013 by Fett2oo5

[Smidge204]

Efficient strutting is what you need:

In your case, use 3-way symmetry so the trusses are between (not underneath) the second stage engines, and strut up diagonally in both directions.

Where is that probe going? Seems awfully overbuilt even for a tour of the Jool system.

=Smidge=

[running]

What worked for me a lot of times... (assuming your break is between the fuel tanks):

Use the small structural cube thingy. (name?) place it on 4 (or more sides) above and below where the 2 tanks meet (and break).

Then connect a strut from the lower right conrner of the top cube, to the upper left corner of the lower cube, creating a cross (do this on all sides ofcourse)

(You can do the same even if there is more stuff in between like a decoupler and engine.)

For me that makes the rocket a lot stronger. However as stated above, this design is quite extreme in it's large weight ontop.

- You could use smaller tanks (as in thickness) to which you attach the nukes (and add more tanks without a nuke) to keep the top mass closer to the center.

- You could attach a few small SRB's to the top tanks (the ones with the nukes) that pull on the rocket during lift of instead of the balancing act from the bottom. You might run into trouble when they run out and you are still doing the balancing act.

- For the 3 engines and tanks below the nukes I would also look into smaller engines and more slimmer tanks to keep the mass closer to the center.

[numerobis]

You don't even need trusses to truss the gap between the Skipper stage and the LV-N stage: just run struts from the bottom tank to the top tank. I'd recommend 4 struts going straight up, unless you have a lot of torsion.

Other than that, there's lots to critique about your rocket, if you're interested. You could build a much smaller one and still get as much range.

[Taki117]

1. If the girders aren't helping take them out.
   
2. More boosters, this will allow you to get rid of the tri-tanks on the second stage
   
3. More Struts, depending on where the failure is happening at every joint, and make a strut cage around that lone LV-N in the middle.
   
4. It could just be me, but it looks like your lower stage is going into a fuel loop try Asparagus staging.
   

[Mr Tegu]

As running stated: If you pull all of the booster stages closer to the principle axis you will reduce the bending moments caused and thus improve the stability and strength of the rocket considerably (as well as saving a little mass too). Locking the gimbals on the outer engines will also help.

Edit: You seem to have attempted to use asparagus staging on the lower boosters but symmetry mode has messed things up for you. You need to place each crossfeed individually.

Edited December 16, 2013 by Mr Tegu

[Grant_]

Efficient strutting is what you need:

In your case, use 3-way symmetry so the trusses are between (not underneath) the second stage engines, and strut up diagonally in both directions.

Where is that probe going? Seems awfully overbuilt even for a tour of the Jool system.

=Smidge=

Hopefully Moho and the Sun, but right now I'm experimenting with maximizing my dV (16 000 atm).

I'll take a guess that the connection failure is at that 1.25m fuel tank? You've got a lot of weight balancing on a "toothpick" between your larger fuel tanks.

Baring a significant design change, try using the same size tanks throughout.

If you are set on using this fuel tank configuration connect the 2.5m tanks that are on top and bottom of the 1.25m tank with 6-8 symmetrical struts. That might solve the immediate problem.

Of the two above, I suggest using 2.5m tanks for the entire craft, as you have with most of your design.

During failure the skipper fuel tanks (stage 2) seem to push inwards at the bottom and then everything seems to blow up at launch, which does seem to take place there. I'm not set on the fuel configuration, just on the amount of fuel and thrust.

You don't even need trusses to truss the gap between the Skipper stage and the LV-N stage: just run struts from the bottom tank to the top tank. I'd recommend 4 struts going straight up, unless you have a lot of torsion.

Other than that, there's lots to critique about your rocket, if you're interested. You could build a much smaller one and still get as much range.

How would I make it smaller? Right now my staging is just based on engine efficiency and achieving a 16 000 dV.

1. If the girders aren't helping take them out.
   
2. More boosters, this will allow you to get rid of the tri-tanks on the second stage
   
3. More Struts, depending on where the failure is happening at every joint, and make a strut cage around that lone LV-N in the middle.
   
4. It could just be me, but it looks like your lower stage is going into a fuel loop try Asparagus staging.
   

I definitely will consider boosters. Just haven't added it to my program yet. Yes All of my tanks are in fuel loops since I think I need all the thrust the entire time :/ ... I guess I should calculate at what point I can get away with less thrust.

As running stated: If you pull all of the booster stages closer to the principle axis you will reduce the bending moments caused and thus improve the stability and strength of the rocket considerably (as well as saving a little mass too). Locking the gimbals on the outer engines will also help.

Edit: You seem to have attempted to use asparagus staging on the lower boosters but symmetry mode has messed things up for you. You need to place each crossfeed individually.

Right now the thing just explodes on the launch pad and I can't even take off... I'm just not sure how to move things in closer, while keeping all the engines. I haven't used asparagus staging yet... right now I'm just making sure all the engines get fuel for the entire stage.

Thanks everyone. I definitely have a lot to think about.

[Taki117]

Hopefully Moho and the Sun, but right now I'm experimenting with maximizing my dV (16 000 atm).

you can get anywhere you want with that kind of dV, and still have more than enough to get back.

During failure the skipper fuel tanks (stage 2) seem to push inwards at the bottom and then everything seems to blow up at launch, which does seem to take place there. I'm not set on the fuel configuration, just on the amount of fuel and thrust.

How would I make it smaller? Right now my staging is just based on engine efficiency and achieving a 16 000 dV.

the engine thing is due to your skippers accelerating faster than the tank because the tank is now empty, but the skipper still has fuel

I definitely will consider boosters. Just haven't added it to my program yet. Yes All of my tanks are in fuel loops since I think I need all the thrust the entire time :/ ... I guess I should calculate at what point I can get away with less thrust.

fuel loops tend to do crazy things to rockets, take a look at asparagus staging. Another alternative is to make all of the tanks in your initial stage the same so that they all drop at the same time.

Right now the thing just explodes on the launch pad and I can't even take off... I'm just not sure how to move things in closer, while keeping all the engines. I haven't used asparagus staging yet... right now I'm just making sure all the engines get fuel for the entire stage.

you don't need your engine for the entire stage, just long enough for it to lift its share of the fuel. That's the beauty of asparagus staging, you only carry the mass you are currently using, then as it gets used you drop the excess (skippers are heavy!)

Thanks everyone. I definitely have a lot to think about.

My comments are in red, sorry for not doing it proper but I'm on my phone.

[Grant_]

After using a tri-coupler on the nuclear rockets, I made a gif to show how it's failing:

[numerobis]

OK, so general critiques from me: your TWR is quite high.

Remember that a mainsail is a huge amount of thrust: it can push 100t and still have good TWR on liftoff. You've only got 250t on your five mainsails, which suggests you can drop to just three mainsails and still climb quite well, or add a whole bunch of fuel. Otherwise you'll have to throttle back to avoid exceeding terminal velocity. Similarly, each Skipper can push just under 50t of payload and still get enough TWR for mid stages. So you can drop one from the second stage.

You have a mess of fuel lines on the bottom stage. If you want to burn all the engines for the same amount of time, just put the same amount of fuel on top of each engine, and get rid of the fuel lines. Simpler, fewer parts, no worrying about whether you'll end up with unbalanced fuel usage (which cycles in the fuel line graph can do).

After my proposed modifications so far you have 3 mainsails on the first stage burning a jumbo and a small tank each, 3 skippers on the second stage burning a jumbo each, 3 LV-N on the third stage sharing a half-jumbo between them, then 1 LV-N on the fourth stage burning another half-jumbo, and finally your probe on top. You've got the same amount of fuel, but less engine mass, and therefore you have higher total deltaV. And a simpler rocket.

Departing from simplicity, the Skipper is not that great. We could replace the three Skippers by a cluster of 8 LV-T30s and one LV-T45 to get almost the same thrust (1920 kN rather than 1950 kN), slightly less mass (11.5t rather than 12t), slightly better Isp, and still have some thrust vectoring. Not a huge change in itself, but that's more thrust than a mainsail. You can therefore use these clustered engines twice: the first stage will be two mainsails and the cluster, each burning a jumbo and a small tank. The second stage will involve dropping the two mainsails but keeping the cluster, now burning three more jumbos. Total engine mass is reduced by 6.5t since the cluster replaced three Skippers and a mainsail. And the T30/T45 have higher Isp than the engines they replaced. So you have better Isp and lower mass => better deltaV overall. You also have a bit more thrust on liftoff, which you could use by adding more fuel, almost a jumbo's worth shared between the first-stage engines.

Now you can also run fuel lines from the mainsail towers to the cluster tower. Add the cluster's first-stage fuel to the mainsail towers, which means your second stage won't be hauling as many empty tanks. You can also split the second-stage fuel in two, and drop the empty fuel tanks: the cluster is burning three jumbos, so if you ditch half of the empty tanks, that's 6t you're dumping for half the burn => more deltaV.

Last trick, of less consequence than the previous stuff: you've got three idle LV-Ns. Unless you're using the FAR mod, turn that stage 90 degrees from the lower stages, and run fuel lines from the cluster's fuel tanks to the outboard LV-Ns. This way, two of your LV-Ns will be firing all the way up, allowing you to add about 8t of fuel to the second stage yet keep the same TWR. As a bonus, it also improves Isp (as soon as you get above 600m). If you're using FAR, this is probably bad advice since it means you'll have a bigger cross-section area, and thus more drag.

[numerobis]

As for explosions on the pad: MOAR STRUTS.

[Grant_]

The thrust has been calculated to get 2.2 TWR for the first stage, 1.5 the second (@6300m-low orbit) and the third 0.2. What TWR values should I aim for?

Edited December 17, 2013 by Grant_

[SRV Ron]

790+tons, 108 tons in orbit. KW Rocketry and Nova Punch mods used. Payload to Jool is stock.

[Vanamonde]

Just as a suggestion: the length of the LV-Ns seems to make them wobbily no matter how well strutted they are, so I simply place them on side mounts where they do not bear any weight.

[Taki117]

790+tons, 108 tons in orbit. KW Rocketry and Nova Punch mods used. Payload to Jool is stock.

pics

I recognize that craft! but seriously, if you want even fuel usage, you need the same amount of fuel for each engine, with enough weight to make them controllable.

[SRV Ron]

Older all stock design.

[SRV Ron]

I recognize that craft! but seriously, if you want even fuel usage, you need the same amount of fuel for each engine, with enough weight to make them controllable.

It flies quite well. The center tank feeds the twin LV-N till it runs out. Bracing properly is the key to stability.

When rockets fall apart on the pad, there are serious stability and bracing issues to be sorted out.

[numerobis]

The thrust has been calculated to get 2.2 TWR for the first stage, 1.5 the second (@6300m-low orbit) and the third 0.2. What TWR values should I aim for?

1.5 is plenty at launch -- remember that TWR grows during your stage as you burn fuel. Then 1.3 above 10km is a good start. That said, you might need more engines than I described; I was assuming the first stage took you to 10km or so. It might now, having reduced engine mass by 16.5t and increased the Isp.

[LethalDose]

Man... Just stack the tanks you have radially mounted. I do that all the time.

I typically try to have anything on the upper stage supported with lower booster stages to keep things stable and safe. It works moderately well.

And, No, I don't use asparagus staging, I prefer classic-ish (combo of serial and parallel) staging. My rationale is that if asparagus staging were feasible/practical, then NASA would have been doing it for the last 50 years. I'm willing to take that hit to efficiency on the chin for RP purposes, and I've done fine so far.

[Grant_]

Just updated based on your guys suggestions I did employ Asparagus staging to drop 2 of the 8 engines for the first stage due to excess thrust... but I got it to orbit nicely!

The only thing I dislike with having this much fuel is that is takes almost two minutes to create a proper orbit once I cut engines at 70 km and I doubt adding more nuclear engines will help. If I cut back from 16k dV to 15k dV I could use LV-T45 engines instead of skipper.... which is probably better.

Edited December 17, 2013 by Grant_

[The Ideal Gas Lawyer]

...

2.More boosters, this will allow you to get rid of the tri-tanks on the second stage

3.More Struts, depending on where the failure is happening at every joint, and make a strut cage around that lone LV-N in the middle.

...

More struts, more boosters. Most generic (and effective) advice in KSP, ever.

For the record, I agree.

[Taki117]

Just updated based on your guys suggestions I did employ Asparagus staging to drop 2 of the 8 engines for the first stage due to excess thrust... but I got it to orbit nicely!

PIC

The only thing I dislike with having this much fuel is that is takes almost two minutes to create a proper orbit once I cut engines at 70 km and I doubt adding more nuclear engines will help. If I cut back from 16k dV to 15k dV I could use LV-T45 engines instead of skipper.... which is probably better.

now for some optimization. 2 of those X200-32's is equal to 1 orange tank, and two of the X200-8's cans is equivalent to 1 X200-16.

Using this in your first stage, if you put an orange tank and an X200-16 and an X200-8 fuel tank on top of the mainsail you will have the same amount of fuel as you do now (Without having to run all the lines) If you then asparagus this you should get about the same dV. (But don't quote me on that, I'm guessing here)