How to tune SRB

[seaces]

Hello,

I will admit it. I run away from those SRBs simply because I go too fast with them and I used liquid fuel rockets simply because I can control them all the way up. I know I can limit them in VAB but often I do not know how to really optimize them so that I do not get red hot on the way up. Any tips?

[Sir_Robert]

43 minutes ago, seaces said:

Hello,

I will admit it. I run away from those SRBs simply because I go too fast with them and I used liquid fuel rockets simply because I can control them all the way up. I know I can limit them in VAB but often I do not know how to really optimize them so that I do not get red hot on the way up. Any tips?

They are most optimal with bigger ships, where the burning out SRB doesn't increase the ship's TWR by a factor 5.

Ideally, I use them on those ships that lack a few hundred m/s deltaV, or lack a little intial TWR. As the SRB burns out (and TWR increases), you can throttle back the main engine to keep the total TWR constant

[Evanitis]

You can limit the thrust on SRBs in the VAB/SPH. I usually use them to get a 1.6-1.8 TWR - sometimes along with the main engines, other times SRB only. But it entirely depends on the craft in question, so there is no universal answer here. KER or MJ readouts (or calculating TWR on paper) can help you find the sweet spot for solid boosters.

[Chaos_Klaus]

Most people use Kerbal Engineer Redux (KER) or other  mods to view statistics while building rockets.

SRBs are very inefficient because their specific impulse (ISP) is low. That means that they are very heavy for the amouut of delta v they give you. That pretty much rules SRBs out for anything but initial lift off.

The one thing SRBs are good for is producing lots of thrust! And they are comparatively cheap. 

So what most people will do, is build a liquid fueled center stack that has almost enough delta v but can barely lift off, because the thurst to weight ratio (TWR) on launch is about 1. Now, just add a few SRBs to help the rocket clearing the pad and maybe even accelerate through the sound barrier. By the time the SRBs burn out, the center stack has burned enough fuel so that TWR is a little higher and it will continue to lift the rocket.

Since providing lots of thrust is the only thing SRBs are good at, it makes no sense to throttle them down at all. If you have to do that, just go with less SRBs instead ... or less liquid fueled engines.

You really only need a TWR of about 1.5 on launch. If you have SRBs that burn out quickly, you can go with an initial TWR that is way higher (above 2).

Edited February 17, 2016 by Chaos_Klaus

[seaces]

Well yes that is what I planned them for to help me move the rocket. Anyway thnx for the answer now it is little bit clearer.

[Streetwind]

While it's more efficient to go full burn, sometimes a more reined-in TWR can mean an easier, cleaner ascent that saves dV in its own right.

For that purpose, i sometimes make SRB stages with staggered burnouts. For example, imagine I want to attach three BACC boosters. I'll use one stack-mounted center booster which I leave at 100%, and radially attached to it, two more boosters which I set to something like 75% thrust. That way, about three quarters of the way into the SRB stage's burn, the center SRB flames out, and your TWR drops even while the two outer ones continue burning for another while.

[Tex_NL]

12 minutes ago, Streetwind said:

While it's more efficient to go full burn, sometimes a more reined-in TWR can mean an easier, cleaner ascent that saves dV in its own right.

For that purpose, i sometimes make SRB stages with staggered burnouts. For example, imagine I want to attach three BACC boosters. I'll use one stack-mounted center booster which I leave at 100%, and radially attached to it, two more boosters which I set to something like 75% thrust. That way, about three quarters of the way into the SRB stage's burn, the center SRB flames out, and your TWR drops even while the two outer ones continue burning for another while.

While it is true you want to leave the thickest parts of the atmosphere as quickly as possible doing it as fast as possible is NOT the best solution.
Drag increases exponentially, drag multiplies by four if you double your speed. As a result you will be spending fuel fighting drag that can be better used to climb higher. You should only go as fast as possible AFTER you have left the lower atmosphere.

[Evanitis]

19 minutes ago, Streetwind said:

it's more efficient to go full burn

That's incorrect. Sometimes full burn is the most efficient, but that can only be a coincidence based on the rocket's weight and TWR. Feel free to test it in a new career: a single pod on a flea SRB reaches a higher altitude and velocity if you limit the thrust to 40%.

EDIT: Ahh, just saw you mentioned that it can save dV. Disregard this post.

Edited February 17, 2016 by Evanitis

[Streetwind]

31 minutes ago, Tex_NL said:

While it is true you want to leave the thickest parts of the atmosphere as quickly as possible doing it as fast as possible is NOT the best solution.
Drag increases exponentially, drag multiplies by four if you double your speed. As a result you will be spending fuel fighting drag that can be better used to climb higher. You should only go as fast as possible AFTER you have left the lower atmosphere.

In the absence of other forces, that's true. However, your rockets don't fly in the absence of other forces. Notably, there is gravity. During a typical rocket ascent, gravity losses outnumber drag losses between 4:1 and 5:1. If you boost harder, you spend less time burning upwards (suffering full gravity losses) and more time burning sideways (not suffering gravity losses) because you can fly a flatter ascent with a sharper pitchover without falling back down.

In trading gravity losses for drag losses, you almost always win. There is of course always the point of overdoing it, like @Evanitis example of a single pod on a Flea (which isn't even really comparable to a normal rocket), but as long as the question is "is my SRB stage's ending TWR of 3 too much?", the answer is "keep burning, you're nowhere near terminal velocity with that".

The challenge subforum has some really interesting entries in various "least dV cost to orbit" challenges, by the way. You'd be shocked at the TWRs of the winning craft.

[Snark]

2 hours ago, Streetwind said:

While it's more efficient to go full burn, sometimes a more reined-in TWR can mean an easier, cleaner ascent that saves dV in its own right.

For that purpose, i sometimes make SRB stages with staggered burnouts. For example, imagine I want to attach three BACC boosters. I'll use one stack-mounted center booster which I leave at 100%, and radially attached to it, two more boosters which I set to something like 75% thrust. That way, about three quarters of the way into the SRB stage's burn, the center SRB flames out, and your TWR drops even while the two outer ones continue burning for another while.

+1 to this; I love this technique and use it all the time.

My own take on it is tweaked slightly.  I tend to use radial SRBs a lot.  So any time I have 4, 6, or 8 of them, I do this:  instead of attaching them all as a single symmetry group, I attach them as two symmetry groups (i.e. two groups of 2, 3, or 4).  One group is set to a higher thrust than the other.  So, for example, if I work out that I need an overall thrust of 85% to get me my desired launchpad TWR of 1.5, I'll set one group to 100% and the other group to 70%.  The higher-thrust group burns out and is jettisoned first.

It works great.  Smooths out the ride (less of a TWR spike at the end; longer period of more uniform acceleration).  Ditches the mass and drag of the higher-thrust rockets sooner.  Basically, just adds "legs" to the SRBs.

I refer to it as "poor man's asparagus." 

(It also gives a bit of design flexibility.  For example, suppose I have 8 of the SRBs, and I need some AV-R8 winglets down at the bottom for stability.  I can put 'em on one of the 4-groups, but not on the other, since I don't need eight of the darn things.)

[wumpus]

Note:  "getting red hot" is only a problem for SRBs if they explode on you (which does happen when they get sufficiently hot).  A bigger issue is the difficulty in controlling a rocket traveling at transonic speeds.  If your rocket is capable of TWR>2, then dialing down the thrust to about 2.0 tends to cause the least total losses due to gravity and aero losses.  Of course,  if your rocket happens to have less than 2 TWR, it doesn't mean need you need more (I prefer closer to 1.7, but obviously many have success with ~1.5.  Once you have enough TWR you should only be adding more SRBs for delta-v, and that is typically a losing cause.)

Launching a rocket with TWR>>1.5 will make it pretty clear that you now have steering issues.  My favorite way to "cheap out" and get strong steering for low cost is to use a pair of AV-R8 (the control type) winglets attached on the North and South side of the rocket.  This gives plenty of control for the roll over (aka "gravity turn"), and hopefully the rest of the controls and/or stability can deal with all other movement.

Probably the best guide to SRBs in this forum:

 

[FancyMouse]

12 hours ago, Tex_NL said:

Drag increases exponentially

I just couldn't resist my instinct of being picky and wanted to say it's quadratic, not exponential.

[FullMetalMachinist]

7 minutes ago, FancyMouse said:

I just couldn't resist my instinct of being picky and wanted to say it's quadratic, not exponential.

Well, technically any number raised to a number, whether that's a 2 or 3 or 4 or even a variable, is still an exponential number. 

[Snark]

52 minutes ago, FullMetalMachinist said:

Well, technically any number raised to a number, whether that's a 2 or 3 or 4 or even a variable, is still an exponential number. 

<pointlessly_pedantic>
Not quite.  If you have some variable x, "exponential with x" means k^x, whereas "polynomial with x" means x^k ("quadratic" being the special case of "polynomial of power 2").  Those are very different things.  Doesn't matter what k is, the exponential is much faster increase when x gets large enough.
</pointlessly_pedantic>

Saying "drag increases exponentially with speed" would imply that it's basically some multiple of e^kv, which is not the case.  Drag goes up with speed, but not that much.

As distinguishd from, say, atmospheric pressure as a function of altitude, which does decrease exponentially.  (Not a simple uniform height constant anymore, since 1.0, but still it's pretty much exponential.)

Edited February 18, 2016 by Snark

[FullMetalMachinist]

6 minutes ago, Snark said:

<pointlessly_pedantic>
Not quite.  If you have some variable x, "exponential with x" means k^x, whereas "polynomial with x" means x^k ("quadratic" being the special case of "polynomial of power 2").  Those are very different things.  Doesn't matter what k is, the exponential is much faster increase when x gets large enough.
</pointlessly_pedantic>

We're getting slightly off topic now, I know. But now I'm a little confused. Does this not agree with what I said? I was trying to say that x^2 or x^3 or x^(whatever) are all exponential. Is that not correct? 

If I'm missing a point somewhere, please forgive my ignorance. 

[Snark]

1 minute ago, FullMetalMachinist said:

We're getting slightly off topic now, I know. But now I'm a little confused. Does this not agree with what I said? I was trying to say that x^2 or x^3 or x^(whatever) are all exponential. Is that not correct? 

If I'm missing a point somewhere, please forgive my ignorance. 

Yes, that's exactly what I'm saying (and what FancyMouse was saying), i.e. your understanding is not correct.  x², x³, and xⁿ are not exponential; they're polynomial.  2^x, 3^x, and n^x are exponential.  There's a big difference.

But yeah, hopelessly off-topic, I'll shut up about this now. 

[numerobis]

What I sometimes do for simplicity:

1. I tweak the SRB to have half its fuel.
2. I look at the engineer's report to find the mass.
3. I calculate: what thrust do I need to move that mass at a TWR of 2, or a bit less. Remember F = ma. You want a = 2 * 9.8 and you got m in step 2, so now you know F.
4. I calculate: alpha = F / T, where F is from step 3 and T is the thrust of the booster at full throttle (I use the sea-level thrust).
5. I tweak the SRB to have a thrust limiter of alpha. And I give it back all its fuel.

If you have two boosters, you use T as the sum of the the boosters. It's not too complicated to work out other cases, like implementing @Streetwind's staging (which I often do myself) or handling a central engine (e.g. if you need it for gimbaling).

Edited February 18, 2016 by numerobis

[Vegetal]

You could also do it like me and install Crzyrndm's SRB thrust plugin, which gives you 2 thrust sliders, one for when it's full, and the other for when it's empty.

Thrust gradually decreases (or increases if you want) until SRB shut off. It's beautiful.

[wumpus]

In career mode (or while designing rockets with an eye on cost), the only time I reduce the amount of fuel in a SRB is to meet launch pad requirements (obviously changing the thrust is another story).  Otherwise you are basically better off filling it up and taking whatever delta-v it can give you.  The reason it seems to work this way is the cost of the decouplers:  they tend to dominate the costs of lateral SRBs until you are using kickers (and by the time you are talking kickers, you are losing measurable delta-v with the fuel your rocket) [if you aren't using lateral decouplers, wouldn't you get more delta-v/kerbuck by filling the SRB and then just waiting for it to burn down to 2g while racing to the heavens?]

I've all but given up on the smallest and tend to only use the biggest two (thumpers and kickers).  For sufficiently small rockets, these can act first stage boosters on their own, larger rockets can either use a mass of SRBs (control can be tricky) or mass them into two groups attached latterly to your rocket (no need to waste more than two decouplers).