Self-correcting re-entry body shape - rules of thumb?

[Crown]

When I build re-entry vehicles that are supposed decelerate by air, I sometimes have problems keeping them with the heat shield towards the airflow.

Is there a way - kinda basic building rules - to determine whether they are stable or not?

I know that the Mk1 pod doesn't need input to stay retrograde to aerobrake. The try of making the shake look like a Mk1 pod or like the Curiosity EDL vehicle doesn't always work. Often, SAS helps me with pointing the retrograde but on some missions, battery life is tightly calculated. Disposable SAS modules are sometimes no option - geometrically and financially.

I tried to find something via internet search but the material I've found doesn't too much into detail. Haven't had a chance to hit the university library with this topic yet.

[Dafni]

Generally try to keep the COM as low as possible, and maybe increase drag at the top (by adding airbrakes or fins or something)

[Superfluous J]

@Dafni's got it. Take the best model airplane you can find and duct tape a brick to the tail, and drop it off a building. It's gonna fall brick-first, even though it's been built to fly forward.

Then run, because the plane's owner is going to be very mad at you.

[Crown]

I will try to attach wings to a fairing as soon as I get out of this building here.

But is there rule of thumb one can go by like
"The re-entry body including the payload has a COM_^real.An equally shaped but uniformly dense re-entry body has a COM_uniform. If COM_real is in front of COM_uniform, then the re-entry body will stay stable "front first", otherwise it will want to turn around. Where in front of describes a displacement towards the heat shield."?

 

[Superfluous J]

8 minutes ago, Crown said:

I will try to attach wings to a fairing as soon as I get out of this building here.

But is there rule of thumb one can go by like
"The re-entry body including the payload has a COM_^real.An equally shaped but uniformly dense re-entry body has a COM_uniform. If COM_real is in front of COM_uniform, then the re-entry body will stay stable "front first", otherwise it will want to turn around. Where in front of describes a displacement towards the heat shield."?

 

In short, yes.

The further forward your COM, the more stable your vessel in that direction. If your COM is in the back, you're going to flip around. If it's in the middle, you're going to need to hold your direction with fins or reaction wheels and power. If it's in the front, it's going down nose first.

For 2 ships of a given shape, it's pretty much directly related to COM placement. I am wary to say "if the COM is 75% of the way forward, you're 75% stable forward" because I don't even know what that means, but it's essentially true. The further to one end the COM, the more the ship will want that end to point forward, and at some point that desire will override everything else.

[numerobis]

One construction trick is to have the pod near the back, and fuel tanks above it. At launch, the CoM will be above the CoP and you'll go straight. During your orbital maneuvers, you'll use up the fuel, so CoM will drift backwards to the pod and engine (which don't lose mass), so that at re-entry, CoM will be below CoP and you'll tend to fall retrograde.

Another trick: put elevons up front, ahead of CoM. On launch, use them as canards -- be careful not to go too far from prograde! On re-entry, turn them off (turn off pitch/yaw/roll) and use them passively to shuttlecock.

Or even more fun: leave them active on re-entry, turn off pod torque, and set your SAS direction to retrograde. When you're pointing retrograde, SAS will try to correct for slight off-retrograde motions, but because you're going backwards it'll actually just increase the motion. Eventually you'll end up prograde, and SAS will correctly force you away from that to be retrograde. All in all you'll be tumbling wildly, mostly perpendicular to the air flow. That ensures high drag and it spreads the heat over your entire surface, so nowhere gets overly hot. Then, within limits, you can leave the heat shield at home. Doing this I'm normally subsonic by 15km altitude.

[Snark]

1 hour ago, Crown said:

But is there rule of thumb one can go by like

"The re-entry body including the payload has a COM_^real.An equally shaped but uniformly dense re-entry body has a COM_uniform. If COM_real is in front of COM_uniform, then the re-entry body will stay stable "front first", otherwise it will want to turn around. Where in front of describes a displacement towards the heat shield."?

 

Well, sort of but not quite.

For example, a cone is going to want to fly butt-first.  That's what the Mk1 pod does.  Pretty sure it is uniformly dense.  It'll be stable even though its "real" CoM is exactly congruent to its "uniform" CoM.

What the actual rule is, it will be stable if the real CoM is in front of the real center of drag.  The problem with the rule-of-thumb you propose above is that you're essentially assuming that the "uniform" CoM will be located in the same place as the center of drag.

For some shapes (such as cylinders) that will be true, but for others (such as cones) it will not.  Basically, the "uniform CoM = CoD" only applies to shapes that are symmetrical on their longitudinal axis-- i.e. the top and bottom halves of a cylinder are identical, but for a cone it isn't.

What makes it tricky is that there isn't any simple way to know where the center of drag is, other than visual guesstimation.  There's no in-game tool to do it (because it's hard).

[Geschosskopf]

In general, for both reentry vehicles and ascending rockets, try to build the thing like an arrow:  heavy stuff on the end you want to be in front, tailfeathers on the other end.  Insufficient weight at the front can be countered by more drag at the back and vice versa, and the further apart the weight and the drag are, the less you need of either.

This all means it's pretty easy to make unpowered, self-correcting reentry vehicles if they have some length to them.  But things that are more discoidal like that Curiosity thing violate all these rules of thumb so pretty much have to be powered.  Which is in fact shown in that pic, with the RCS firing.  For them, you need either RCS or lots of torque and the juice to run the torque long enough to matter.  RCS is harder to do because it needs tanks and the exposed thrusters have a good chance of burning off the ship anywhere but at Duna.  Thus, if you want to build discoidal reentry vehicles, you don't have much choice except reaction wheels and batteries.

[Gojira1000]

I played with intentionally unstable no-heatshield designs and tumbling re-entries with recoverable odd-shaped first stages from orbit - which does work and is interesting to watch.

 

 

[SinBad]

Google up ' biconic lifting body ' to find out more info on the apollo style re-entry vehicle's inherent stability.

 

besides the spacecraft application, i believe  its the best way to get asteroidal ore to the surface. Foam it up with nitrogen until its at 990kg/m^3, mould it into a 100 ton biconic body, spin it up for gyroscopic stability, then just nudge it into aerocapture for ocean recovery by tug.

[Snark]

Really looking forward to the deployable 10m heatshield thing that's coming in 1.1. Put one of those on the back end of a reentering vehicle and it's all the "tailfeather" you could possibly want.

(Or in the front, to actually shield the vehicle, but then you may need to worry about mass distribution, or how long your vehicle is. Hard to know for sure how it will play out until we get our hands on it.)