Mastering the Korolev Cross

[shadragon]

My heavy lift orbital rocket has four SRB's at the base of a larger SRB in the center. When the four SRB's burn out, I jettison them and they usually collide below my vehicle. 2 chances in 5 they'll damage something on the rest of the ship. 

I've tried TT-38K Radial Decouplers singly and in pairs. but nothing I do lets them peel away smoothly. I know there's a mod to do this, but can it be done stock? I'm early career so there may be a part coming up soon, but right now I can't see a way to jettison the SRB's so they fall clear safely. 

Any tips?

[Gaarst]

Use sepratrons. Best way to get simple clear separations is to orient them upwards and inwards.

[shadragon]

Looks like the answer, but takes 1100 to unlock. That'll be a while.  

http://wiki.kerbalspaceprogram.com/wiki/Sepratron_I

[Guest]

I think you can orient fins/wings and slanted nosecones so the airflow pushes them away without using sepratrons, but I don't know the specifics of how-to.

[FullMetalMachinist]

37 minutes ago, shadragon said:

I've tried TT-38K Radial Decouplers singly and in pairs.

Just to be clear, if you're saying that you tried attaching the side boosters to the center one using two Decouplers, this doesn't do anything. The side booster will only ever attach to one decoupler or the other, never both. 

To solve your problem without using extra parts, attach the booster to the radial decoupler higher up. Place the decoupler a little higher up than you normally would, and attach the booster to it. Then use the offset tool (press 2) and move the booster down so that it is attached more towards the front (of the side booster). That way, when you decouple it, some of that ejection force causes the front end to rotate away from your craft. Then aero forces will cause it to keep rotating away and move it away from the center. 

[shadragon]

@FullMetalMachinist I did try a single higher up, but not the offset. Will give that a go tonight. 

@String Witch I thought of that too, but felt it would cause unnecessary drag. I'm going to play around with various combos this evening. Thanks. 

[DD_bwest]

definitely use sepratrons,   Dont be afraid to use a bunch.    Im often building rockets that use 5m+ boosters around a larger core.  strap on a bunch of those and you will get them to peel away nicely.

[Signo]

It is an old cross but "checks out". In the pic you can see the sepratrons firing.

[eddiew]

Tbh, it does seem odd that you don't get separatrons with the first SRB unlock (Hammer?)  It's the same stuff in a smaller tube...

[Snark]

2 hours ago, shadragon said:

I did try a single higher up, but not the offset.

Yah, @FullMetalMachinist's advice is spot on:  you want the decouplers to be mounted as high up on the radial booster as possible (or, phrased another way, you want the radial boosters to be mounted as low as possible on the decoupler), so that the ejection force tends to kick the booster nose-out away from the main ship.  That will cause aero forces to make them peel away.

The other thing is to have the radial boosters mounted as low as possible on the central core.  Reason:  As soon as you eject them, your center core is going to start moving forward relative to them (since it's thrusting and they're not).  The farther back you can mount them, the less distance the core has to cover in order to "clear" them (i.e. get in front of them, so it doesn't matter if they come inwards; they'll just hit each other, not the core).

If your radial boosters are very big/massive (e.g. Kickbacks, or liquid fuel boosters on big tanks), you can consider using the bigger/heftier Hydraulic Detachment Manifold; it has more ejection force, that can help.

[MK3424]

How about using 2 sepatrons but the top ones burn longer, and the top one uses less thrust than the bottom ones.

[Kurld]

Without Sepratrons, one thing I used to do that worked ok was reduce the decoupler force to almost nothing. The idea is to impart as little torque to the booster as possible, while still pushing it away from the main rocket. That way they wont flop around so much when they jettison. 

Another thing to help is build a mockup of the booster as it's own ship and use the CoM tool in the VAB to give you the ideal location to mount the booster to the radial decoupler (or to mount your Sepratrons if you have those). You want the decoupling force to push out through the CoM of the booster. Remember to remove the fuel from your booster mockup when getting the CoM.

[SSgt Baloo]

I sometimes place a shallow "vee" of small fins at the back of the radial booster, and (as others have said) the decoupler as high on the radial booster as possible, with the back of the booster low enough to assure (probable) clean separation. When the decoupler kicks the nose away from the core booster, the fins help swing it away from it, and the added drag helps the spent boosters fall behind before they can collide with anything else.*

* After much practice. Sandbox mode is a great teaching tool. YMMV.

Edited September 1, 2016 by SSgt Baloo

[swjr-swis]

3 hours ago, shadragon said:

Any tips?

A few. Personally, I have a much better non-collision record by making sure the boosters don't rotate at all when staged. So a bit contrary to what others here are saying, my nr. 1 tip would be: remove/avoid any source of rotations as best you can.

• Know where the CoM is of your boosters, and place the SRB CoM centered on the radial decouplers - then offset the decouplers as needed to adjust the height of the SRBs. This way when they get the kick from the decouplers, there will be no rotation and they should 'fall' straight down without risking collisions.
• A single trut at either end of the booster is usually enough to stop them from wildly wobbling or swaying about when they flame out, which is another source of unwanted rotations. Even then and especially with high TWR SRBs, give the ship a few moments between flameout and decoupling to allow even small oscillations caused by the sudden drop in thrust to die out.
• Following the prograde indicator into a gravity turn is good practice for lots of reasons, but it also helps with safe release of SRBs: try to be as close on the prograde vector as you can when you release them, to avoid drag from causing (often asymmetrical) rotations. Around the time you release SRBs drag tends to be a major force and can easily nullify any other attempts at a controlled release.
• If boosters/SRBs appear to be gaining some kind of unwanted rotation no matter what you do, try to separate them more from the center stage - especially the engines. There are a number of parts/engines in KSP that have an invisible collision mesh bigger than the visible surfaces, which means that placing them in the default spot the editor places them has them already colliding, even though visibly there appears to be enough separation (Mammoth and Twin-Boar in particular are notorious for this). The result is that at decoupling, things immediately get a kick from that engine and often the induced rotation is exactly the one you don't want.
• If you feel you really need a bit of rotation to pull the SRB tip away from the center stage: a single fin on the bottom outside of the SRB, due to the induced drag, can be just enough to do this.
• Sepratrons should be a last resort, in my opinion. They add drag, mass and part count, it can take some tweaking to ensure the rotation they create doesn't overdo it, there's a risk of damaging other parts, and they won't always overcome the problems mentioned above. If you do use them, since you often have to tweak anyway, consider minimizing the solid fuel they pack and limit the thrust.

[Plusck]

5 hours ago, FullMetalMachinist said:

To solve your problem without using extra parts, attach the booster to the radial decoupler higher up. Place the decoupler a little higher up than you normally would, and attach the booster to it. Then use the offset tool (press 2) and move the booster down so that it is attached more towards the front (of the side booster). That way, when you decouple it, some of that ejection force causes the front end to rotate away from your craft. Then aero forces will cause it to keep rotating away and move it away from the center. 

100% agreed that this will solve the problem.

When I started playing KSP, it struck me that the most solid connection from SRB to main stack should be at the bottom. That way, the SRB's driving force would be held close to the centre stack and wouldn't wobble, the "push" would be directed through the middle of the ship and all connections would therefore suffer the least, etc.

But no. Each individual part in KSP is utterly solid and doesn't flex, and struts are not really bars of metal but are rigid in all axes. Therefore you only need to tether the bottom with a strut, and it won't move, while the bit that has an explosive separation (the TT-38K here) should be at the top, to force the SRBs to peel out. And it works every single time.

You've got to remember it's a game. I'm sure in real life the strongest bolts are at the bottom, but the most explosive ones are surely at the top. You can't do both in KSP, so concentrate on the explosive part and strut the rest.

[Magzimum]

Obviously you can also use the TT-70 Radial Decoupler, which creates a little more drag than the TT-38K, but is quite likely to solve the problem too. I find this the easiest solution. These are unlocked in "Advanced Construction", at a cost of 90 science points, I believe.

Personally, I had the most success with method similar to what @swjr-swis advised: placing the decouplers right on the center, and decoupling while moving nicely prograde.

I found in tests that puting the decoupler high up (away from the CoM of the booster) has a risk of pushing the bottom (engine) of the SRB into the center rocket. But as was advised, you should also place them low, perhaps with their nozzles below the nozzle of the central rocket. As long as they are placed low enough they will move inwards a bit before being pushed out by aero forces, but all that happens below your rocket.

If for whatever reason you need to decouple the boosters while going wildly off the prograde market, then sepatrons are the best. But I never use sepatrons on ordinary kickbacks... only on clusters of kickbacks or on much larger (2.5m or 3.75m) liquid fuel boosters with heavy engines at the bottom.

[Palaceviking]

12 hours ago, MK3424 said:

How about using 2 sepatrons but the top ones burn longer, and the top one uses less thrust than the bottom ones.

Omg people! He hasn't unlocked sepatrons!

[MK3424]

2 hours ago, Palaceviking said:

Omg people! He hasn't unlocked sepatrons!

Oh, my bad... it seem he then have to place a radial decoupler at the right place for a korolev cross..

[Vive_moi]

Hello,

I usually configure my booster (SRB or LF) as FullMetalMachinist do, but if generated separation forces are not enough (rarely), I stage the decouplers just before the booster cutoff (about 0,2s). This way the booster, lightly rotated by the decoupler, will get a small shoot away from its engine.

Note: the booster, when decoupled, will go out and forward largely faster than your rocket. so you cannot have any part directly on front of the boosters, else they will destroy it.

See you,

Vive_moi

Edited September 2, 2016 by Vive_moi

[shadragon]

I don't have the sepatrons yet, but am advancing fairly quickly. Still mastering getting into orbit. Well, when I say mastering, I haven't accomplished it yet. (Curse you Scott Manley for making it look so damned easy in your videos!!!)

@String Witch Tried pairs of offset fins just below the nose cones of the SRB's, tilted away from the main body. No effect. Still collided underneath my vehicle. 

@Snark Also tried decouplers as you suggested, but on stage separation, the SRB's stayed put attached to the central SRB. I think I may have screwed up the placement, but when I remounted them a second time in VAB, the snap-to looks to have placed the SRB's directly against the inner SRB and not the decoupler. Not sure if that is a bug, or lack of finesse on my part. Will work on it. 

 

[Snark]

32 minutes ago, shadragon said:

@SnarkAlso tried decouplers as you suggested, but on stage separation, the SRB's stayed put attached to the central SRB. I think I may have screwed up the placement, but when I remounted them a second time in VAB, the snap-to looks to have placed the SRB's directly against the inner SRB and not the decoupler. Not sure if that is a bug, or lack of finesse on my part. Will work on it. 

Yes, that's almost certainly the problem.  It's really easy to miss the decoupler and accidentally attach it directly to the central core, especially since the camera angle makes the part that you're placing get in the way of the decoupler that you're trying to attach it to.

A couple of ways you can guard against that:

First, there's a subtle editor effect you can watch for.  When you're placing a part, and it's semitransparent and glued to your mouse, and you move it around looking for an attachment point... the part you're going to attach it to will highlight in the editor.  So, let's say you have a central core, on which you've placed the decoupler, and now you're trying to place your radial booster on the decoupler (and not on the core).  The decoupler is small, so your booster is obstructing it... but the core is not obstructed.  So just carefully watch that core.  If it's highlighted, don't place because it'll be attached to the core.  Kind of grope around (i.e. slide it vertically up and down) until you see it un-highlight; that means that you're on the decoupler rather than the core.

Second, the decoupler does have some "depth" to it, so after you've placed the radial booster, turn the camera around and look at it from the side.  If you've successfully placed it on the decoupler, you should see it on the decoupler, spaced a little bit out from the center core, with an air gap between them.

[Chiron0224]

I typically mount my SRBs low on the rocket, wait a bit after flame out, and then stage with the rocket pointed prograde.  At this point I pray to the deep space Kraken (he must be appeased) that it works.  I'm glad someone posted about this though, I'm going to play with it tonight and try to master the Korolev Cross myself.

[Snark]

Incidentally, if you have 4-way radial boosters, there's another little technique that you can use to give them a bit more burn time, and also not stage so many things at once (which may help with the collisions problem).  I call it "poor man's asparagus".  It's usable any time you have an even 4-or-more number of SRBs.

Let's say you've decided what launchpad TWR you want, so you can set the thrust limiter on your SRBs accordingly.  For example, suppose you do the math and you work out that you need 85% thrust on your four SRBs in order to get the desired launchpad TWR.

Instead of adding them as a single 4-way symmetry group set to 85%, you add them as two 2-way symmetry groups at right angles to each other.  Call them Group A and Group B.  Both groups ignite together (i.e. you lift off the pad burning all four SRBs), but you give them different thrust levels.  In this case, instead of all four SRBs set to 85%, you set group A at 100% and group B at 70%.

That will give you the same launchpad TWR.  However, group A will burn out first (because of the higher thrust setting), and you stage them away as soon as they burn out.  Then you can continue flying on group B for a while, until it burns out, and then you stage those away and fire up your core stack.

Advantages of this technique:

• Drops dead weight and drag earlier in the process, so it's more efficient.
• Smooths out the brutal TWR spike that you tend to get with SRBs at the end of their burn time.
• Gets a longer burn time out of the SRBs, and you can tune your ascent profile so that you don't go too fast too soon.

Aside from those benefits, it'll also be the case that you're dropping them only two at a time rather than all four at once, so possibly that may help you if you've got collision issues.

Note:  If you do use this technique, set it up so that the boosters on the east/west sides of the rocket are jettisoned before the boosters on the north/south sides.  It's more aerodynamically stable that way, assuming that you're launching eastwards.

[Sharpy]

There are a few approaches that work for me.

1, Place them, where they won't harm anybody. They may require a bit of strutting though.

2. Tilted winglets. Better for lower altitudes, where air is still thick. Four per booster. They are dirt cheap, and very light, so the only real cost is effort.

3. Separatrons. Good for high altitude.

4. Increase the distance even more.

Regardless, before separation:

- center on prograde direction.

- if that's still not enough, throttle down.

Also, stop using these crappy decouplers. Structural pylons and small hardpoints beat them by strides in functionality; the decoupling power is so crappy in all of the radial decouplers the drop really makes no difference. Instead, you're getting extra separating distance and much better aerodynamics.

[dire]

@Sharpy I prefer a method that typically saves a bit on part count and improves stability compared to what you're doing. Instead of the delta-wings with elevons I'll use Type-A wings with either one decoupler on the end, or two decouplers paired on either side; and then a tail fin either on the edge of the wing or on the bottom of the booster that's on the edge of the wing. Here's an example of what I mean -- Note if you look closely the decoupler marks on the top stabilizer fin; these six boosters just successfully decoupled. Also notice I killed thrust in order to minimize the chances of a hit breaking something.

On 8/28/2016 at 7:04 PM, dire said:

 

 

Edited September 3, 2016 by dire