Why did my rocket not launch???

[Loren Pechtel]

28 minutes ago, Snark said:

Just to take a step back, here... rather than simply trying to fix the rocket, is it worth trying to fix the problem?  i.e. what's the purpose of this vehicle?

Reading between the lines (apologies if I've missed something along the way, the thread has gotten pretty long at this point), I gather that what you're after is a vehicle that can schlep four tourists to orbit and back home, and that's pretty much it, yes?  Maybe some science along the way, as a low-hanging fruit, but mainly what the ship is about is "get four tourists to LKO and then home again", yes?

So, my suggestion:  rather than trying to fix a particular design, start with the payload and design up from there.

Your center core is a great idea.  Command pod, couple of crew cabins.  Decoupler so that it's the only thing coming back, good.  I would suggest putting a pair of AV-R8 winglets left and right on the rear of that vehicle-- it will give you steering ability on reentry.  If you're only going to LKO, you can get by without a heat shield, as long as you have good steering ability so you can make maximum use of body lift.  You could even stick a couple of fixed fins on the middle of the craft, around the CoM.

So that's your reentry vehicle.  What next?

Below the decoupler, you put your upper orbital-insertion stage.  Suppose this is two of the 2-ton LFO tanks with a Terrier.  (I'm guessing you have to stack 'em because you haven't unlocked the tall skinny 4-ton tank yet.)  So at this point you've got a craft that's probably somewhere around 8.5 tons, carrying 4 tons of fuel.  With a Terrier, that's over 2100 m/s of dV, which is 2/3 of the needed dV to get to orbit, right there.  So all you need to do is give that thing a boost of a bit over 1000 km/s and you're good to go.

Let's say, for example, that you put 3 more of the 2-ton tanks under it, with a Swivel.  That'll raise total mass up to around 17 tons, of which 6 are fuel in that stage, so you'll get on the order of 1300 m/s dV.  When you build that, disable the top two tanks so that the bottom tank will drain first.  Your craft is really going to have some aero stability issues with that light, winged thing up at the front, so give this stage four of the AV-R8 winglets, put just as far to the rear as you can manage.

Theoretically that's enough dV, but it's pretty low TWR and we'd like to have some safety margin.  So now comes the MOAR BOOSTERS part of the design.  Strap on four radial Thumpers around that bottom stage.  With decouplers, nose cones, and maybe some more AV-R8's at the back of the boosters, that brings your total ship mass up to something like 48 tons.  If you want launchpad TWR of 1.5, you'd set the thrust limiters on the Thumpers to around 70%.

That ought to get you to orbit just fine, with a much lower part count.  Take off on the SRBs; by the time they burn out, you're already pitched over at 45 degrees or more, and the relatively low TWR of the Swivel stage won't matter as much.  Just remember to enable the tanks on that stage as the bottom one drains, so you can keep continuous power.

(Actually, you could give the SRBs a bit more legs by setting up the four radial boosters in two pairs.  Instead of setting all four of them  to 70%, you can set one pair to 85% and the other pair to 55%, so that the hi-power pair ditches runs out of fuel and ditches first.  Sort of a poor man's asparagus.)

Anyway, just a thought. 

It wasn't meant to be the prototype of a class, I was just building a rocket to gather as much science and contracts as practical.  The only craft I've made more than one of is the little science rover I sent around the launch complex--I built a second one after unlocking more instruments to put on it.  (And then I build a third similar one because I got a contract to test the Juno engine while splashed down.  I drove it slowly into the water and tested the engine.  Since I now had a radial intake on top the engine still worked--and somehow so did the steering.  I was expecting to recover it from the ocean but I was able to drive it back to the runway.)

31 minutes ago, FullMetalMachinist said:

Well, one option is the Kerbal Joint Reinforcement mod.  It takes the stock joint reinforcement that happened in 0.23.5 and pushes it even further, making all your joints between parts even more rigid.  Now, there are some very well respected users on the forum that would say that KJR is more of a crutch, and that if you need it to function, then it's really a sign that your basic rocket design needs improvement.  I'm not refuting that opinion, just pointing out that differences between players exist.

Another option is careful, strategic use of connector struts.  The main thing to keep in mind when using struts is that they add a lot of drag, so use as few as you can get away with while still making your rocket more stable.

The last option is perhaps the least attractive.  Basically redesign your rocket from top to bottom, keeping in mind the whole time that it should look and behave like a real life rocket.  This is fairly difficult, and at times counter-intuitive.  If you have a design that you're mostly happy with, but is still wobbly, you can upload the .craft file somewhere (dropbox, google drive, kerbalx.com) and let myself or other players try it out, and modify it if necessary, to give you some ideas on what works and what doesn't.  

Other than that, useful screenshots and very descriptive problems usually get the most useful advice. If all else fails, keep trying. This game is not for the faint of heart, and can seem a bit daunting to begin with.  But it is so worth the effort when you do something awesome for the first time.  I'll never forget the accomplishment I felt when I first landed on the Mun.  So just keep at it, you'll get better with practice, and there are plenty of helpful players on the forum when you have problems.

I strutted between the outer engines, it didn't help.  Where should I strut?

[FullMetalMachinist]

4 minutes ago, Loren Pechtel said:

I strutted between the outer engines, it didn't help.  Where should I strut?

I would get rid of those struts and instead strut from to tops of the radial boosters to the central stack.  However, that type of instability that needs fixing with struts is one of the hallmarks of an inherently unstable rocket, and maybe should be redesigned using the tech that was unlocked by the previous mission.  But if you're set on using this rocket again, then struts from the radial boosters to the center should help. Be sure you go from the outside in, and not the center out with the struts.  The part that you first attach a strut to gets all the weight and drag of the struts applied to it, and the little nub that remains after separation continues to add weight and drag to the parent part.  So if you go from the center out, when you decouple the boosters, you won't see the struts any more, but the little nub will still be there, taunting you with increased drag and mass.  Better to attach them to the boosters first, where those pesky nubs will fall away with the spent stages.   

[Loren Pechtel]

11 minutes ago, FullMetalMachinist said:

I would get rid of those struts and instead strut from to tops of the radial boosters to the central stack.  However, that type of instability that needs fixing with struts is one of the hallmarks of an inherently unstable rocket, and maybe should be redesigned using the tech that was unlocked by the previous mission.  But if you're set on using this rocket again, then struts from the radial boosters to the center should help. Be sure you go from the outside in, and not the center out with the struts.  The part that you first attach a strut to gets all the weight and drag of the struts applied to it, and the little nub that remains after separation continues to add weight and drag to the parent part.  So if you go from the center out, when you decouple the boosters, you won't see the struts any more, but the little nub will still be there, taunting you with increased drag and mass.  Better to attach them to the boosters first, where those pesky nubs will fall away with the spent stages.   

Thank you, although that will do do nothing about the wiggling I saw in the central stack.

[FullMetalMachinist]

7 minutes ago, Loren Pechtel said:

Thank you, although that will do do nothing about the wiggling I saw in the central stack.

Well honestly, sorry I couldn't help more, that was about the best I could do. I would suggest following Snark's earlier suggestion. Maybe not word for word if that's not the exact mission plan, but his general though process was spot on. Start with what you want your final payload to do, then build your rocket backwards from there. 

If you get stuck with any other problems, maybe think about opening a new thread. This one has meandered off it's original question enough, I think. And don't forget to give specific details about what you're having trouble with, and post screenshots. Good luck, and remember that a big part of the fun of this game is trial and error, and learning from your mistakes. 

[Plusck]

28 minutes ago, Loren Pechtel said:

Thank you, although that will do do nothing about the wiggling I saw in the central stack.

Oh, but it will. The boosters are already connected low down on the central stack. Adding a strut from the top of the boosters to higher up on the stack will make triangles radially around the rocket which distribute the forces exerted by your gimballing engine and your torque wheels. It will most certainly reduce the wobble that is visible in the pic you posted.

Once you are out of the thick atmosphere and lose your boosters, you should need less control input anyway. You can right-click your central rocket engine and toggle gimbal on and off as required. You can also switch SAS on and off to dampen the wobble.

One big reason for wobble is that the part doing the controlling is at the very top of the stack, together with the torque wheels. To turn, the torque wheels push one way and the engine vector changes to push the other way. However, due to the large number of weak joints the top will start moving quite quickly, while the middle of the rocket is still pointing in the original direction. The controlling part "feels" this movement first, and reverses the movement to stabilise at the new heading. However the springiness of the joints makes it snap back too quickly, so it again reverses the movement... and meanwhile it is telling the rocket engine at the other end to do the opposite. This sort of oscillation is unavoidable with SAS on the best of rockets, and on a noodly one it can rip the whole thing apart.

[Loren Pechtel]

1 minute ago, Plusck said:

Oh, but it will. The boosters are already connected low down on the central stack. Adding a strut from the top of the boosters to higher up on the stack will make triangles radially around the rocket which distribute the forces exerted by your gimballing engine and your torque wheels. It will most certainly reduce the wobble that is visible in the pic you posted.

Once you are out of the thick atmosphere and lose your boosters, you should need less control input anyway. You can right-click your central rocket engine and toggle gimbal on and off as required. You can also switch SAS on and off to dampen the wobble.

One big reason for wobble is that the part doing the controlling is at the very top of the stack, together with the torque wheels. To turn, the torque wheels push one way and the engine vector changes to push the other way. However, due to the large number of weak joints the top will start moving quite quickly, while the middle of the rocket is still pointing in the original direction. The controlling part "feels" this movement first, and reverses the movement to stabilise at the new heading. However the springiness of the joints makes it snap back too quickly, so it again reverses the movement... and meanwhile it is telling the rocket engine at the other end to do the opposite. This sort of oscillation is unavoidable with SAS on the best of rockets, and on a noodly one it can rip the whole thing apart.

Next bird I will try strutting that way.  Is there any other fix beyond unlocking fatter rockets?

[Plusck]

1 minute ago, Loren Pechtel said:

Next bird I will try strutting that way.  Is there any other fix beyond unlocking fatter rockets?

Longer parts help a lot. Each joint adds a possible bend, so it's just a question of reducing joints by using taller tanks instead of two or four shorter ones.

Also, building outwards rather than upwards can help, but then you do lose more energy to drag. Unlocking fuel lines to build outwards with asparagus fuel routing helps both efficiency (dropping empty tank mass and unnecessary engines ASAP) and stability (especially if you arrange fuel lines so that the bottom tanks empty first).

And fins with control surfaces help too - they don't suffer so much from the inherent difficulty of torque wheels, while limiting the amount of active control needed anyway simply by being there.

[Loren Pechtel]

8 hours ago, Plusck said:

Longer parts help a lot. Each joint adds a possible bend, so it's just a question of reducing joints by using taller tanks instead of two or four shorter ones.

Also, building outwards rather than upwards can help, but then you do lose more energy to drag. Unlocking fuel lines to build outwards with asparagus fuel routing helps both efficiency (dropping empty tank mass and unnecessary engines ASAP) and stability (especially if you arrange fuel lines so that the bottom tanks empty first).

And fins with control surfaces help too - they don't suffer so much from the inherent difficulty of torque wheels, while limiting the amount of active control needed anyway simply by being there.

I guess I need to concentrate more on the top part of the tree.  I was more interested in the stuff on the bottom that would get me more science.  Why are the joints so weak?  Don't Kerbels have good welding torches??

[Snark]

53 minutes ago, Loren Pechtel said:

Why are the joints so weak?  Don't Kerbels have good welding torches??

To make the game more interesting, by giving you challenges to overcome.    Building a successful rocket in KSP is a balancing act between many different requirements:  structural rigidity, sufficient dV, launchpad TWR, aerodynamic stability, rocket costs, building upgrade costs, tech requirements, payload considerations... it's what makes the game endlessly entertaining because there are endless combinations of challenges.

I know it may not seem like it, when you're relatively new to the game and lots of things seem frustratingly hard.  Once you get used to it, though, it's very manageable.  It's very easy to build rockets that don't flop around, once you understand the technique.  It'll come with practice, don't worry!

Also:  weak joints are an important message to you.  Listen to this message.  It means "you're building it wrong."    This is why I advise people against installing KJR out of frustration before they really learn how to build rockets:  it means they won't learn that skillset.  A floppy rocket is KSP's way of telling you "maybe you should rethink your design."

In your case, what it's telling you is:  you're making your rocket too big.  You only have beginner parts unlocked, so build beginner rockets with them.  Specifically, smaller rockets.

One very common mistake I've seen that leads to lots of newbie frustration:  Don't try to build a big rocket when you're still low on the tech tree.  Doing so forces you to do things that will require, for example, stacking lots of little tanks, which is a recipe for disaster.  Be a little more patient and save the big rockets for a little further down the tree, when you've got larger tanks to play with.  The 4-ton tall-skinny LFO tank, for example, is a great workhorse and will enable you to build all kinds of nice craft; unlock it as soon as you can.

In the meantime, while you're working up the tree:  instead of trying to build big ships, figure out how to get more done with smaller ships.  For example, do you really have to send four kerbals up at once?  If you build a ship that only carries two tourists instead of four, then your engineering job is enormously simpler (especially if you haven't unlocked the bigger fuel tanks yet).

And yeah, that means you'll have to do two launches instead of one.  But on the other hand... how much frustration and grief has your current 4-seater brought you?  How many repeated, failed launch attempts?  I kinda suspect that a small, simple 2-seater that just runs like clockwork and succeeds on the first (or maybe second) try will go so quickly that you could have launched and retrieved at least three of them in the time you've spent on this one.

And furthermore, it'll be better for you, too, in terms of developing your skills:

• By making the rocket small and simple so you don't have to contend with engineering problems so much, it'll allow you to focus on flight technique.  Learning's a lot easier when you can focus on one thing at a time.  Later on, you'll be building bigger and more complex ships where the engineering is more of a challenge, and it would be good to have your piloting skills rock-solid by then so you don't have to worry about it.
• The skills you learn in flying those simple little rockets are directly transferrable to your later career.  You're always going to need piloting technique, and it takes practice.  The fact that you're doing multiple little launches will give you more practice at this very important fundamental skill.
• Whereas, if you spend a lot of time trying to figure out "how do I make this monstrosity fly" ... that really is wasted time, because even if you learn how, it's not a skill you'll be using much.  Because pretty soon, you'll have unlocked the bigger fuel tanks, stiffness just won't be a problem, and you won't be flying any limp-noodle rockets anymore.
• Finally:  The tyranny of the rocket equation means that heavy is bad, and less is more.  If you find yourself trying to do a really challenging mission that takes lots of dV... and you find that your design seems to keep ballooning and ballooning and pretty soon your Moho mission is thousands of tons and costing millions of funds... you would be astonished at how much smaller a rocket you could build that would deliver the same payload at a tiny fraction of the cost, if it's designed efficiently.  So you need to develop a keenly-honed sense of "mass economy" in KSP, and get in the habit of ruthlessly paring down your designs so that you don't waste mass.  Forcing yourself to stick to smaller ships for a little while longer will help you learn to get more done with smaller rockets.

Anyway, just my two cents.    (Okay, maybe three.  Four, tops.)

Edited February 17, 2016 by Snark

[FullMetalMachinist]

50 minutes ago, Loren Pechtel said:

I guess I need to concentrate more on the top part of the tree.  I was more interested in the stuff on the bottom that would get me more science.  Why are the joints so weak?  Don't Kerbels have good welding torches??

This is the main reason I use Kerbal Joint Reinforcement. The floppy joints add an artificial engineering challenge. Granted that if you do design a successful rocket around that challenge it will perform better overall. But I really think that the current stock joints are far too weak. 

As a side note, the Realism Overhaul suite of mods lists KJR as a prerequisite. Meaning that the authors of the mod whose purpose is to make KSP as close to real life as possible, think that the stock joints are too weak. 

[Gojira1000]

KJR is not a must-have on a rocket (it's nice, but I just lofted a 1300-ton refueller last night without KJR and it wasn't too noodle-y) but on space-planes it very nearly is, unless you like your wings with a lot of flapping. (Also external strutting looks awful to me on a space plane)

[FullMetalMachinist]

17 minutes ago, Gojira1000 said:

KJR is not a must-have on a rocket

I'm not saying that it is, just pointing out that without it there are design difficulties that may be tough for some players to deal with.