Payload Issues - Wobble of Doom

[Somtaaw]

So after getting help previously with thrust issues, and successfully launching the core of my new orbital station at almost exactly 0800 Zulu yesterday I've now spent almost 24 full hours trying to get the first of two modular fuel hubs into space and to successful docks. Pics posted below, but the design on payload is more or less solid, and I'd prefer not tampering with it if at all possible.

All launches I've tried so far have been generally high atmosphere firework shows, ranging from as low at 20km to as high up as 60km before nuking. I've adjusted my launch platform repeatedly trying to get more stability (although I have no screenshots of my numerous failures, just my latest failure). I have successfully made it into stable orbit, performed rendezvous burns, and performed a close 200m approach.... then I realized the front 60% or so had sheared off at some point while I was in map view, probably at some point between successful stable orbit and final approach. The more and more I fail with getting even a basic modular station into a nice easy 150km orbit of Kerbin, the less I want to make any more stations.

Latest launch pic isn't that great, but its 4am and I've just spent 12 hours straight trying to loft just 1 fully intact fuel hub into space successfully, so won't take a new one with KSP closed down, but the bottom of each -64 is a Mainsail engine. The engine shrouded by a decoupler I think is a Skipper, but I've also tried it with Poodles.

Attempts so far have been primarily with payload empty (excluding that massive rcs bunkerage), and I've even tried lowering it's CoM down into the main thruster section using a -32 central fuel stick, with all those -64's Asparagus staged around the central -32, with the payload also contributing fuel into the burn (that design almost made stable orbit, before it split in half and then exploded)

Best results to date, have been with the rough pic I did link below, although if it's not my rocket spontaneously ripping itself apart, it's the payload snapping like a twig.

Payload:

Lastest failure to make orbit

Edited April 24, 2014 by Somtaaw

[ScottyDoesKnow]

Put those struts from the orange tanks going right to the payload instead of the docked parts, it'll be much sturdier. While it looks like the ends of the struts stay on the payload, once you leave and come back they'll be gone. If you keep having trouble, post the craft file as well.

[MarvinKitFox]

RCS tanks are squishy. Best not put them in the middle of the stack like that, or if aesthetics demand such placement, use struts that connect to more solid elements both above and below them.

I assume your current strut placement is also for visuals... consider attaching the struts to cubic octagonals, that you have rotated *into* the bodywork, thus hiding all sign of the strut attachment points.

Strutting via structural members on junior ports is very wobbly indeed.

[Taki117]

This may sound crazy, but instead of putting those girders on the jr. ports, why not put them onto decouplers between the ports to make a much sturdier connection, and then run struts from there to the top of your payload.

[capi3101]

What's the TWR on your booster? I've had issues before with payloads being served up at IHOP, and I traced back the problem to too high of a TWR. In general, keep the gee meter at the top of the green zone after you've made your gravity turn and you should be good; throttle down if you need to.

The other thing is to ditch the big RCS tanks entirely; a quad of RCS cylinders holds just as much monoprop as a big tank and they're a hell of a lot sturdier. You can set them diagonal to the docking ports, or perhaps cluster them up on the end of that Rockomax Adapter you've got there.

[Somtaaw]

Removed 2/3 of the RCS bunkerage, and tweaked a few other parts. Lost 3/4 of the junior ports, and adjusted the strut placement. That let me get into stable orbit, but the suggestion of switching to a quad of the radials is good, I might swap that in and put the port count back up. Rather have lots of RCS fuel available for my tugs, than to spend half my gametime lugging more fuel in, whether from Kerbin or installing Kethane and producing it myself.

Just installed VOID, and the successfully orbitted design has a launch TWR of 3.13, and 3.4 km/s. I might go back and merge it with some of the suggestions I got this morning from Kit & Scotty, with what Taki & capi gave. Remove the last big RCS tank, and add in radial tanks. Maybe add a few more supportive struts.

Playing with the bigger engines, fuel tanks and whatnot has thrown off my gameplay. I was fine as long as I stuck with 1.25m parts; playing with bigboy toys and I'm making very pretty ballistic missiles until I ask for help.

[Dkmdlb]

3.13 is too high for a launch TWR. You want to shoot for something like 1.5. So ditch half your engines and save yourself a lot of weight. You want to keep the throttle at 100% during ascent, and hit 250m/s as you pass 10km in altitude, and 700m/s as you pass 20km in altitude. If you are going faster than that, you need fewer or weaker engines. No point in carrying up a set of massive engines only to use them at 50% throttle, is there?

[Redshift OTF]

Since I haven't had this problem lately I'm trying to see what you are doing differently to me and the most obvious difference is a lack of SAS on your payload. SAS on the top of that should greately improve your launch stability and allow you to use less RCS to turn the thing when you do get in orbit.

[RW-1]

IMHO, struts to the payload, add a SAS and you will be gold. I've launched longer thinner stuff for station modules, but don't yet have a pic for ya.

Also, if launching really large fuel units, I'll drain half (or all depending) to reduce weight to get it up there, you can always launch more single tankers to dock and refuel it, which you end up doing anyways.

[Somtaaw]

3.13 is too high for a launch TWR. You want to shoot for something like 1.5. So ditch half your engines and save yourself a lot of weight. You want to keep the throttle at 100% during ascent, and hit 250m/s as you pass 10km in altitude, and 700m/s as you pass 20km in altitude. If you are going faster than that, you need fewer or weaker engines. No point in carrying up a set of massive engines only to use them at 50% throttle, is there?

Hmm, I'll have to keep that in mind. When I tried using less engines, I could barely make it to 40km up, before I'd be onto my last stage. Those speed/altitude numbers seem low, not saying they're wrong but food for thought to me. I guess I over-engine, even with throttling heavily back to get my station tanks up. Most of my rockets hit 250m/s by 4km up, and are pushing 1km/s by 15km, usually as 2-stage. First stage, like for my last failed launch, gets ejected somewhere between 50-80km up depending on how much I coast before starting circularizing of my orbit. And it's anywhere from a 60 second to 4 minute coast phase, which I use as a rotational period to lower from 45 degree to 90 degree.

@Red, I have a single a-sas, but it's with the primary launch stage, gets detached when I reach orbit. With the amount of monoprop I have stored, and that it's part of the launch fuel I use that for rotational needs.

I'll post a picture of the model that successfully researched orbit, can always use suggestions as I imagine I'll only increased the number of stations I have the longer I play. So I have just one of two planned fuel hubs in orbit now, and I figure at the very least I'll want two, maybe three more stations later (Jool, Eeloo, and Dres) so thats upto 6 additional fuel hubs I need to effectively launch. Unless we get full-blown orbital assembly, and no longer need to launch all new stuff from Kerbin.

That's the launch setup that worked. I'm already starting to consider alterations, taking into account what everyone's saying.

[Pecan]

Now you have VOID make good use of its TWR and deltaV stats. Your whole launch design is too big but inefficient.

1) You need approx 4,500m/s to reach LKO

2) Launch TWR should be around 1.5 - 1.7

3) You do not want to exceeed terminal velocity during the ascent

4) Your payload is 15 tonnes and even a basic rocket design should give a 10% payload ratio; ~150t total at launch. Asparagus should enable you to reach 15% - 19%; 88t. Your design is at least twice as heavy as it should be.

5) Mainsails and skippers are awful engines - try clusters (bi-, tri-, quad- adapters or use octagonal struts) of aerospikes or T30s for launching

6) When I started KSP I used 1.5m parts with no problem and completely failed at 2.5m too. Trust that you will 'get it' sooner than you think - but first work on building small.

[Somtaaw]

Now you have VOID make good use of its TWR and deltaV stats. Your whole launch design is too big but inefficient.

1) You need approx 4,500m/s to reach LKO

2) Launch TWR should be around 1.5 - 1.7

3) You do not want to exceeed terminal velocity during the ascent

4) Your payload is 15 tonnes and even a basic rocket design should give a 10% payload ratio; ~150t total at launch. Asparagus should enable you to reach 15% - 19%; 88t. Your design is at least twice as heavy as it should be.

5) Mainsails and skippers are awful engines - try clusters (bi-, tri-, quad- adapters or use octagonal struts) of aerospikes or T30s for launching

6) When I started KSP I used 1.5m parts with no problem and completely failed at 2.5m too. Trust that you will 'get it' sooner than you think - but first work on building small.

Sounds like some very solid advice, probably gonna screenshot, print it, and tape it to the side of my monitor. (Edit: Just started poking through VOID, you mention terminal velocity.... where do I find that information so I can build around NOT exceeding it) Although the payload is only 15t if it's empty of everything but monoprop. If I fill those tanks, payload jumps to just shy of 50t. Still a bit of overkill and bad design, and I've been looking at redesigning around the adaptors. Don't got aerospikes though, or most spaceplane content. Went straight for ultra heavy rockets, composites and meta-mats, now I'm working on unmanned and science.

After I finish this first generation station, my next project is trying to make functional Armageddon X-71 shuttles, and Armadillo landers (complete with jumpjets), watching that video right now and I've gathered over 50 good screenshots and I'm not yet to halfway. After banging my head off trying to replicate those, I'll start drawing up station gen 2, starting on paper this time with highlights like you've mentioned and the new goals.

Edited April 24, 2014 by Somtaaw

[Pecan]

For designing launch vehicles work backwards from the payload. Temstar posted an excellent example a while ago of his 'zenith' series of launchers which ... I have now totally failed to find again.

My more trial-and-error 'easy' approach goes like this:

1) Start with you payload or equivalent mass. Add docking port or separator beneath it and a remote guidance unit beneath that (planning to control and de-orbit the last stage of the launch vehicle so no debris is left in space). This is the 'total' payload you now have to lift.

2) Stick a fuel tank and engine on the thing - note the TWR.

a) Move up through 1, 2, 3 and 4 engines (I really recommend T30s if you don't have aerospikes) to see what TWR you can get.

If you can get TWR over 1.8 then add more fuel (= more deltaV = getting further)

c) If you can't get at least TWR 1.2 then reduce fuel.

3) Check the total deltaV.

a) If over 4,500m/s you will go to space today. Get out there and try it.

Otherwise add an asparagus stage and repeat from 2.

For 20t (can't find a 15t one at the moment) my 'standard' launcher is 126.792t - core plus two asparagus stages (each stage being a tank either side for symmetry, of course), X-200 fuel tanks, 4 aerospikes under core, 2 under each of the (4) stage-tanks. It's ony just over 16% payload ratio but a simple build, optimised I tend to aim for 18%. Having said that I've recently switched to just using ARM parts for reduced part-count (lag) and really simple builds.

I almost never launch more than about 40t (orange tube and accessories) myself, preferring to build in orbit (with docking). Apart from just getting the TWR and deltaV I still find it difficult to balance and fly bigger loads during the ascent. I will try putting together T30- and ARM- based 50t launchers for you this evening.

[Somtaaw]

I haven't updated yet to ARM. Friend who hasn't played in a while emailed me his copy of the game so I'm not forced to deal with dem asteroids yet. On the side, I was browsing forums looking for basics, and although I'm normally against using someone else's .craft file, I found someone's MechWarrior named boosters, and couldn't resist. So now I have launchers for 7/15/50t payloads. On the plus side, they have awesome names, and perfect payload distribution. On the downside, I tried taking them apart to learn better build technique, and they are insanely based on clipping.

I'm going to copy both your suggestions for future reference and it's time to start my usual. Go to Usuer\My Documents, open new folder named after game, proceed to fill with notepad reference notes, to do lists, and all project related stuff (such as pictures from Armageddon movie to try to KSP clone the cool stuff). Rinse and repeat for every game you play; I think I've got about 2GB of that sort of game reference information.

[Pecan]

Ok, just had a quick play with T30s. Docking port/decoupler and RCU under 50t payload, as before. X200 tank, quad adapter and 4xT30s give final-stage TWR of 1.35 - 1.54. This is low but it's a small tank and will only be used for circularising the orbit once in space. TWR is less important at that point so it's "ok".

3 asparagus stages (symmetry 2, in other words), based on TT-70 radial decouplers (the wide ones - the quad adapters can make it tricky to use the narrow ones). Each booster stack has X200 and orange tube tanks, quad adapter and 4xT30s, same as the core but with a lot more fuel. TWR from first stage to last is then 1.67-2.14, 1.65-2.36, 1.59-3.08 (a bit high right at the end there!) and 1.35-1.54. Total vacuum deltaV (with 50t payload) is 5,070m/s. It's 366.4t, so only 13.65% payload ratio but since it's the first thing I put together it'll do as a place from which to start.

The idea of the RCU is that once the orbit has been circularised you can decouple the payload, switch to the remaining, last, stage of the launch vehicle, turn it around and use the residual fuel to de-orbit it so that it falls back to Kerbin and is destroyed - like all the previous stages - and no debris is left in orbit.