Slashdot Mirror
CEV Revolutionary Gimballed Thrusters
simonbp writes "A Tennessee Tech Professor has proposed an innovative gimbal mount for 'inclusion to the design of [NASA's] CEV (Crew Exploration Vehicle), revolutionizing the vehicle's RCS (Reaction Control System) and solar panel orientation capabilities.' This will allow for nimble maneuvering and for the solar cells to maximize power production."
If you're like me and are wondering what the heck a gimbal is, wikipedia has an article. Not being an engineer, I still only have but the fuzziest idea of what's going on here; blame a liberal arts background.
Gimball looks to me like a perfectly cromulent word!
Hosting 20G hd, 1Tb bw! ssh $7.95
And I'm pretty sure the orginal CSM did not have a gimbled engine. The Saturn did, but with no where near the same range of motion as being discussed here.
Any sect, cult, or religion will legislate its creed into law if it acquires the political power to do so.
It's standard practice to gimbal the main engines in a launch vehicle, or an upper stage. That gives you pitch and yaw (and sometimes, if you have 2 or more engines, roll) orientation control while you're under thrust from those main engines, without having to use the smaller RCS engines as well.
Other ways of doing it include using a RCS anyways, with fixed main engines; putting vanes or paddles in the main engine exhaust stream but keeping it fixed; using aerodynamic control fins (only works in an atmosphere during certain speed ranges, useless at liftoff or in space); injecting a liquid or gas into the main rocket engine nozzles on one side but not the other, to give side thrust (LITVC or Liquid Injection Thrust Vector Control, though it can technically use gas as well).
It is not standard practice to gimbal the reaction control system used in space. The assumption to date has been that the four fixed quads approach gives the best reliability under circumstances where part of the system suffers a failure. If you lose one of these oriented thrusters (stuck actuator or thruster fails) then it's like losing a whole standard quad, in terms of the vehicle's remaining dynamics. Lose two, and your maneuverability is severely impaired.
As a rocket engine gimbal, this doesn't look promising. It's a rather bulky mechanism; the linkage is much larger than the engine bell. It requires fifteen bearings, not including the three motors. The standard solution, a gimbal ring arrangement, only requires four. The bearings also have to handle off-center loads, never a good thing. Bearings in space are headaches; lubrication is tough and temperature changes can jam them.
The motors are in a weak position from a leverage standpoint; the engine thrust is applied directly to the motor shafts, so they (and their gear trains) must be strong enough to overpower the thruster. In a gimbal ring arrangement, the bearings are usually placed so that the center of thrust is at the center of the gimbal, so that the bearings, not the actuators, take almost all the thrust. Very large engines, like the Space Shuttle and Saturn V main engines, have been successfully gimballed that way.
The three motors don't seem to add redundancy; it looks like they all have to be working.
For comparison, here's a simple gimbal from Amadillo Aerospace, Carmack's rocket program.
In reality, having many fixed reaction thrusters is probably more reliable than have a few steerable ones. Fewer moving parts.
Gimbals have been around since...well, if you believe Needham, at least the 10th Century AD in China. A gimbal is just a way of mounting something so it can rotate relative to something else while still being attached to it and moving linearly with it, and the main application has been on boats where equipment like lamps and compasses is suspended in mounts so it can swing. http://www.sailgb.com/p/captains_cabin_lamp/ is a picture of a small gimballed lamp. So long as the centre of gravity of the equipment is below the plane of the mount, the boat can rock underneath and the lamp, compass, cooker or whatever will stay more or less upright.
By using an outer pair of pivots to hold a ring which then has another pair of pivots at 90 degress to which the equipment is attached, you get two axis gimbals which allow for rocking and for pitch, which is important on small boats. It isn't practical to suspend (say) a marine stove from a chain because it would swing all over the place, whereas suspending it from pivots near the top means that the base can swing a bit while the pans stay more or less in the same place.
So all the stuff in Wikipedia about Euler angles is all very well, but a gimbal is just a way of allowing one thing to be attached to another while being able to rotate in one, two or three dimensions relative to it. There are various designs and obviously the Canfield one is a clever one, but there is nothing mysterious about gimbals themselves.
Pining for the fjords
I am a Mechanical Engineering Undergraduate who attends TTU and does research for the department in which this design originated. I work in the lab where the device was first prototyped. Just as an FYI, the device is revolutionary because of the elimination of repetitive structures. Granted the bearing are an issue but the gimble can achieve a full 360 degree spherical change in attitude with only the use of 3 stepper motors. Nothing else does exactly that at this time. That's why the device is interesting to NASA. Think of replacing the current arrangement of 5 motors with just one. Can you say cost savings? Just thought I would post my 2 cents since I have had to demo the device on several occassions and have first hand experience with the mechanism. OUT
Also the control valving is highly decoupled from the combustion chamber which means high dribble volume and terrible min Ibit. Those simple stepper motors also have to operate at 165R for prolonged periods- this denies you most lubricants and requires special resolvers and the like. There is also no way that such a mechanism can deliver the frequency response of multiple small thrusters pointed in multiple directions. There is also the need to interface either a fiber optic or high voltage spark igniter lead to the thruster across large motions- could be a problem for the non-optical approach.
The issue is : just what problem are we trying to solve? is it cost of the combustion chamber? Number of valves? Weight? Overall complexity? Or is this just an interesting exercise for a kinematics class? The vehicle attitude control function can be performed two active and two standby modules- not four fully active as was used on Apollo. This is highly optimal for cryogenic thrusters since it minimizes the number of lines which must be chilled and pure 6 DOF operations are rare as opposed to simple maneuvers with coupled rotations and translations. This solution was proposed to NASA and rejected as being "just too different from what Apollo did". I cannot imagine them actually flying this contraption.