Intermediate Tip: Separation Anxiety 1.0.5 [FAR]

[Dr.LoveJoy]

Separating Booster Stages Gracefully

 I have seen this technique used in various craft designs, but I thought I would make an explicit post on separating booster stages, especially using FAR, with parts that tend to touch your spacecraft or each other during rotation away from the spacecraft while separating, due to aerodynamic effects.

For booster stages, you generally don't need sepatrons to get your boosters away from the craft. In fact, in some designs they simply do not work the way you would expect due to aero drag overcoming their power, or due to parts clipping into each other and "sticking" to each other while rotating away from your main vessel (I tested this with full size Flea SRBs as separators and even they didn't work in FAR. the stickiness was too great).

Take a tip from the Russians: mount your first stage boosters with a slight angle inwards: use the decoupler itself to rotate a few degrees inwards.. doing it with the tank alone will not work. When the boosters fall away, they will rotate around their center of mass, which is way down near the bottom of your boosters when the fuel tanks are empty. If you rotate them inwards a few degrees, a normal decoupler will provide enough force to rotate them away, and their rotation will never get near your spacecraft.

You will lose a small amount of delta V, but in your first stage boosters, this likely doesn't amount to too much.

For stages that are separated high up in the atmosphere, this isn't as much of a concern, and sepatrons will work just as you expect, but you can use this technique there as well to avoid the additional part count.

This technique used in stock KSP will also help avoid your boosters impacting each other on the way down, if you would like to use a recovery mod to save cash.

Basic Example of inwards rotation

The example is not elegant, but it shows the principle. :-) This design is very simple, and is spec'd to cheaply take a full orange tank into orbit at 80km with 500+ m/s to spare (using FAR). It succeeds in stock KSP without the fairings with 1000 m/s to spare. This was intended at a Munar or Minmus orbital refueler. It has loads of RCS as well. To use it as a Munar/Minmus refueler in FAR, you can add a few solids to increase the delta V left after orbital insertion. Mainsails and Poodles for the entire thing.

The shown craft game has FAR, StockBugFix and Procedural Fairings installed so that you can fire your rockets while they are protected by a boat-tail. If you don't have those mods installed, don't use fairings like this.. they will not work! The engines will not activate.

POST NOTE: Yes, I do know there are other ways of achieving this (with different nose cones, etc) but this is one technique that I haven't seen getting much attention, and it has saved me a lot of hassle in early games! Also, in FAR, your boosters, if ejected in a low enough atmosphere with enough speed, will have a good chance of self destructing regardless of how you tweak it. To avoid that, you have to use more advanced nose cones, add back in sepatrons to keep the stage upright after separation, and use any other technique you can to keep them upright until they slow down enough not to get destroyed by aero forces. But they will still not touch your main spacecraft.

Edited February 1, 2016 by Dr.LoveJoy
title change

[eddiew]

A screenshot would add a lot of clarity to what I suspect is a good post

[Dr.LoveJoy]

There is one in the link. :-) I suspect I do not have enough status here to post images as the "Insert Other Media' button tells me I do not have access.

Edited February 1, 2016 by Dr.LoveJoy

[Temstar]

You don't need to mount the boosters with an angle to achieve aerodynamic separation, you just need to have the decoupler attached to the booster above its empty CoM and when you stage the decoupler will give the booster a kick so they peel outwards like flower petals.

Aerodynamic separation works fine for something like two or four boosters around the core. When you have a lot of boosters things can get tricky once you pitch over. The boosters on the side and bottom of the rocket can still separate fine but the ones on the dorsal surface get dicey due to gravity. I still prefer sepatrons for staging.

[Dr.LoveJoy]

Thanks Temstar, I did know that :-) The issue is that in some designs, it is hard to achieve that balance. Also, using FAR changes things. In stock KSP, this isn't needed at all in my example design, but using FAR, and after putting in the boat-tail fairings, I needed a way to keep the ends away from the middle more definitively. I then realized there were probably a lot of people who may have made designs where their booster stages collided too rapidly due to other reasons.

If you have significant overhang on the bottom, with FAR the vessel will shield the bottom from aero but not the top, so the rotation rate will be much faster than you would expect and the boosters end up colliding, even if you balance the center of mass. Once they collide, they can "stick", which creates other cascade effects. Especially in my case, where the craft is just over supersonic when the first boosters separate. 

Mounting them with a slight angle inwards avoids all the potential problems associated with this. It moves the center of mass away from the vessel a bit so the rotation angle is subtended beyond the crafts trajectory. 

 

EDIT: By the way, all of this is so you can use the stock parts and so on with FAR more easily. If you had added needle tanks and/or fairings up high, the boosters would separate out and down, and you wouldn't need this method. Clearly, there is a great deal to talk about with regards to separating boosters well, but I wanted to add this as a very simple trick to getting a design the otherwise works well to work better without having to redesign from scratch.:-) It's also a way to keep part count down, since you don't need to include sepatrons if you don't want to. 

Edited February 1, 2016 by Dr.LoveJoy
clarification

[fourfa]

1 hour ago, Dr.LoveJoy said:

You will lose a small amount of delta V, but in your first stage boosters, this likely doesn't amount to too much.

I assume you mean due to the cosine loss of slightly off-angle engines?  You can angle them back to straight if you like - the "F" key while using the rotation widget ("3" key in the VAB/SPH) will help a lot.  But cosine loss really is negligible under a few degrees anyway.

[Snark]

Is it worth adding [FAR] to the thread title?  It sounds like the problem you're trying to solve is mainly a FAR one.

I'm just running stock aero, and have never had any trouble with radial boosters hitting the central core of the ship.  Just mount the decouplers above the CoM of the boosters, and they peel away nicely, just as @Temstar pointed out.

If this is a technique that's needed mainly because of FAR, would be handy to label it explicitly as such to avoid confusion.

[Dr.LoveJoy]

23 minutes ago, Snark said:

 

If this is a technique that's needed mainly because of FAR, would be handy to label it explicitly as such to avoid confusion.

Thanks! Done.

In stock it can also be useful in some situations, especially when using many small boosters around a larger stage that are packed closely together. You've probably seen the 'wedding cake' style method of doing this with Really Big Rockets.

40 minutes ago, fourfa said:

I assume you mean due to the cosine loss of slightly off-angle engines?  You can angle them back to straight if you like - the "F" key while using the rotation widget ("3" key in the VAB/SPH) will help a lot.  But cosine loss really is negligible under a few degrees anyway.

You're right about that. In this design, it was about 95 m/s difference in final delta V after orbital injection with upright boosters. However since this thing is supposed to be able to get to very high or elliptical orbits to do rendezvous, I wanted to maximize delta V after orbital injection. In the end, I did rotate the engines to point straight down, which improved things. However I lost the side benefit of less sensitivity to angular perturbation. Which was ok in this one, but in another incarnation I was lofting a very unwieldy payload that had a tendency to oscillate, so I had to put them back to their original angle.

Edited February 1, 2016 by Dr.LoveJoy

[Temstar]

Another method I've seen for safe booster separation is centrifugal separation. Just before booster burn out you start rolling the rocket to build up some rotation. Then once boosters are done you hit stage and they will all fly off at a tangent to the core. You can then de-spin the core.

The con is that it cost you delta-V in terms of steering loss, and your rocket becomes spin stabilised around the separation event instead of follow prograde which can cause aerodynamic drag loses.

Edited February 1, 2016 by Temstar

[Dr.LoveJoy]

2 minutes ago, Temstar said:

Another method I've seen for safe booster separation is centrifugal separation. Just before booster burn out you start rolling the rocket to build up some rotation. Then once boosters are done you hit stage and they will all fly off at a tangent to the core. You can then de-spin the core.

The con is that it cost you delta-V in terms of steering loss, and your rocket becomes spin stabilised around the separation event instead of follow prograde which can cause aerodynamic drag loses.

Really handy for perfect fairing separation though, so for an upper stage its a great technique :-)

[Snark]

Another method that can work in stock (dunno what this would do in FAR) is to add some small fins near the top of the boosters, out-angled slightly.  This causes aero forces to peel away the booster quite nicely, when you're down in the thicker part of the atmosphere.

I hardly ever use this technique myself; I find that simply attaching the decoupler above the booster's CoM does the trick for me.  However, the couple of times I've used it, it worked pretty well.

[Dr.LoveJoy]

11 minutes ago, Snark said:

Another method that can work in stock (dunno what this would do in FAR) is to add some small fins near the top of the boosters, out-angled slightly.  This causes aero forces to peel away the booster quite nicely, when you're down in the thicker part of the atmosphere.

I hardly ever use this technique myself; I find that simply attaching the decoupler above the booster's CoM does the trick for me.  However, the couple of times I've used it, it worked pretty well.

That works in FAR as well. But the problem solved isn't that the boosters are not rotating away, it is when the bottoms of the boosters collide with each other on separation, or the spacecraft, when they do rotate away. If there is any 'side slip' in the design, they can quickly slip downwards and inwards if their center of mass is close to the center of drag of the spacecraft.

Edited February 1, 2016 by Dr.LoveJoy

[forsaken1111]

I just spin my ship really fast before separation and the boosters fly away woosh

[RW-1]

I've on occasion had issues with sep's, however for the most part, depending on if that will be an iffy proposition, I find that if one awaits staging for about 3-5 seconds after separation, your boosters fall back without hitting the spacecraft, and you can then toggle the current stage to light off. (a small amount of Dv loss, but negligible imo in the grand scheme rathar than adding sepatrons to solve that puzzle).

Been good so far ... YMMV

[Dr.LoveJoy]

13 hours ago, RW-1 said:

I've on occasion had issues with sep's, however for the most part, depending on if that will be an iffy proposition, I find that if one awaits staging for about 3-5 seconds after separation, your boosters fall back without hitting the spacecraft, and you can then toggle the current stage to light off. (a small amount of Dv loss, but negligible imo in the grand scheme rathar than adding sepatrons to solve that puzzle).

Been good so far ... YMMV

This can work, except in my case where the second stage has a TWR of <1 and the first stage has a TWR >1.. so even if I wait, the boosters "hang around" for a long time and draft upwards in the air stream.. long enough so that side slip happens downwards, they collide, and often bounce off my spacecraft.

In my case, angling them away worked very well to avoid this.

That's why I said "in certain cases" instead of "in the general case" :-) In the general case, your second stage will have a TWR >1 and your spacecraft will accelerate away from any devastation that might happen below. BUT, if you are trying to save money, and your TWR on higher stages start out at TWR <1, you can get into this sort of problem depending on the design.

Basically, angling the boosters slightly away from your spacecraft costs nothing, and s a way to get out of that problem very easily, once you encounter it. You don't have to pile on sepatrons :-)

14 hours ago, forsaken1111 said:

I just spin my ship really fast before separation and the boosters fly away woosh

I wish I could do that on a supersonic first stage with no wings while below 15km ! :-)

THIS is a link to a more obvious screenshot which might make things clearer. It's fairly simple. The 4 poodles up above are just slapped on the 4 small fuel tanks, which are radially attached, plumbed with fuel lines to the main orange tank, and angled slightly to avoid cooking the nose cones below. I needed the TWR in that stage so I sacrificed the delta V to get it with more engines, and it turned out that I got more delta V with 4 poodles and the extra fuel tanks than with a skipper, plus it made the whole design shorter, which helped with aerodynamics. Three large reaction wheels. Asparagus staged. This can lift probably more than 50 tons to orbit, if you don't mind sacrificing the upper stage to space. And its pretty cheap.

Edited February 2, 2016 by Dr.LoveJoy