attitude control rockets for model spacecraft 6

[KarlChwe]

This is a little outside my expertise, so I am asking you, the excellecnt spacecraft engineering community (or one of them.) Sorry if I use incorrect terminology and such, or if this is the wrong kind of post for this forum.

I am a volunteer at the Denver Museum of Nature and Science, in the astronomy and space exploration section called "Space Odyssey". We have an exhibit that is meant to allow visitors to experience maneuvering a spacecraft in zero-g (in 2 dimentions.) A 1/60 or so model of the space shuttle moves frictionlessly on a flat surface within a plexiglas enclosure. Visitors control the yaw and forward thrust of the model with a joystick. They have to "dock" the model with painted targets on the sides of the en

The problem is "fuel" for attitude control and forward thrust rockets. Previously we had used compressed air piped in by an umbilical suspended vertically overhead. The joystick controlled valves that would allow jets of air to shoot out of tiny nozzles mounted in the sides and back of the model. That approach caused lots of problems. The umbilical resisted twisting, and the weight of it created a constant pull toward the center, where the umbilical could become vertical (like a hanging rope.)

So we are trying to come up with a way to create self-contained reaction jets on the model.

Here are the design contstraints. The materials should be lightweight (say, less than five pounds for fuel, valves, nozzles, etc.) The motors should be able to operate for about 12 hours, on and off, on the onboard fuel. Everything should be cheap and readily available, or improvisable. The materials and their end products should be safe to breathe, touch and preferrably, to eat. The exhaust stream should be relatively cool. The amount of thrust required is small (I don't know how small), since it is moving an object on a relatively frictionless surface. In fact, we don't want too high thrust, because we don't want the craft to end up bouncing rapidly around the exhibit area. We can't really use sophisticated computer logic.

I thought about using cans of compressed CO2, but I decided that they would be depleted too quickly. I wonder if there is some way to continuously create CO2, perhaps chemically inside a pressurized vessel. Or perhaps a small compressor could pressurize air stored in a vessel. We could have a pressure regulator maintain steady pressure to the nozzles.

So 1. How do I create pressurized reaction gas? 2. How do I even figure out how much pressure is required, and how much gas? 3. What kind of nozzles, valves, tubing, etc. do I use?

Any tips or hints of where to go is much appreciated.

Thanks!

Karl

 

[IRstuff]

I think your original implementation was probably in the right direction. If you can't handle CO2, there probably isn't any other self-contained approach that would work any better.

What's missing from the original is a swivel joint to allow the vertical portion of the hose to swivel without twisting. A leaky swivel would seem to be a solution. The air leaking around the joint would act as an air bearing, thus minimizing the stiction.

The second part of that solution would be to have the horizontal portion of the supply and the fixed end of the swivel follow the craft. If the swivel was more like a ball joint, then you could perhaps place a ring of microswitches to sense the direction that the craft wants to go.

TTFN

 

[scooter911]

Are you doing the thruster valving on or off the shuttle model. I'm thinking you stay with what you have and look at a better way to get the air to the model. Take a look at

http://www.beswick.com,

they specialize in miniature tubing and fittings.

Scooter

 

[KarlChwe]

Thank you, Scooter and IRStuff.

IRSTuff, the swivel joint is a good idea. I could imagine a ball joint, at the lower end of the air hose where it attaches to the model, to both relieve both the twisting and the bending forces exerted by the hose.

There is still the problem of the weight of the hose, which wants to hang vertically, pulling the model toward the center. I didn't quite follow the second part of your solution; perhaps that is what you are talking about. I could see having the top end of the hose mounted to a free-moving arm that can swing in a horizontal plane, like you find in those sprayer wands in those self-service car washes. You could have a joint in the middle of the arm to allow the hose to hang vertically no matter where the model is.

Those joints would have to be very low friction. Scooter mentioned Beswick, which carries valves with swiveling inlet ports. I wonder how much friction they have.

Scooter, I was imagining the valving would be on the shuttle model, so that we need only one hose to the model. To control the valves, we could have a direct wire connection. That was what we had, but again, I would prefer an untethered approach. I was thinking about some kind of RC setup? I looked at Beswick's products, but I didn't see any electrically actuated valves (though they do have rotary valves that could be actuated with a servo motor.)

I may have to switch to using electrical motors and ducted fans used by RC model aircraft. It would have the wrong look, but it would be easy to set up.

Thanks again!

 

[GregLocock]

So why not combine the approaches. Hide a ducted fan somewhere on board and then use electrically operated valves to open exhaust ports to thrust the thing?

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[scooter911]

Check the model airplane industry. They use allot of R/C servos and valves. Problem with everything on board, is you have to power it with batteries, they are heavy and have to be charged. I was thinking a real small coiled hose from an air supply and small multi-ribbon wire for controls supported by a spring hanging as high as possible. We have an air bearing here at The Johnson Spce Center and they have a person walk behing the chair carring the hose and cables.

Scooter

 

[btrueblood]

I like the idea of a small on-board tank, operating at a fairly low pressure (15-30 psig), and serviced by a small diaphragm air compressor. It's a lot easier to create flexible electrical connections (power and/or signal wiring) than pneumatic tubing.

 

[KarlChwe]

Thank you, Scooter and btrueblood. (GregLocock, you suggested I look at the FAQs. Am I not doing this right? Please let me know if I am. karl_chwe_@_comcast_._net, remove the underscores.)

Scooter,
I have stumbled upon the rc model field, and I think we will end up using some servos and rc control units. They also have those tiny ducted fans. They even have rockets, but the rockets they use put out lots of thrust for about 2 seconds. What we need is a rocket that has tiny thrust, but can be stopped and restarted, like a tiny liquid fuel rocket. The rc field doesn't seem to have a need for such things. (So I thought I'd ask you guys!)

The problem with any kind of hose is that if it is fixed at any point above the model, it will want to hang straight down, like a pendulum. So whenever the model moves to one side, model will pull the hose out of vertical, and the weight of the hose will try to drag it back. Since the thrusters on the model are weak, and because the model floats frictionlessly, the effect will not be negligible.

Oh, and I called Beswick. They don't have electrically controlled valves, but they may have some others I can use. Thanks for the tip, Scooter.

btrueblood, the small compressor sounds good, if it isn't too heavy. Any idea where I can find such? Ideally the compressor would run on 12 VDC or so, since the lifter fan will be running on batteries.

The tank could be PET drink bottles or aerosol cans, if I can figure out how to fit valves. Or a balloon, to provide constant pressure automatically!

I am beginning to think that both lift and forward thrust could be accomplished with ducted fans. Apparently when they power on there is a little lag time before they develop full thrust, but the lift motor will operate continously, so no problem there. For thrust, a little lag may even feel realistic. For the yaw control rockets, that compressed air idea might be best. I still don't have electrically operable valves that could shut the jets on and off, though.

Karl

 

[RobWard]

Personally I'd look at a perforated floor with pressurised air pumped through it (think air hockey table) which would remove the largest source of power requirement from the model spacecraft. I would have thought that some 12V servo valves swithched via remote control and a bottle of compressed air would give a long duration of use and no umbilical cord to pull the model around.

Small compressor to recharge the air supply and produce the lift through the table and a rechargeable battery and you have very few consumables.

"I love deadlines. I love the whooshing noise they make as they go past." Douglas Adams

 

[btrueblood]

Oy. Yes, there are quite a few manufacturer's of small diaphragm compressors, operating at a wide range of voltages from 3 to 220, ac or dc. Try the usual culprits (

http://www.mcmaster.com,

http://www.coleparmer.com),

and also try googling or a Thomas Register search for small diaphragm compressors. I received a sample of one of these from a very small manufacturer several years ago; their product would be perfect for you...but there's no way this tired old brain will remember the name, sorry!

I like RobWard's suggestion too (make the table perforated, reducing the work needing to be done by the model). You could also off-load the compressor by making the act of docking actually make a connection of an external air supply to refill the on-board tank. You would need to make an alternate refill connection (maybe on top?) for when people fail to complete the docking...

A reason I'm not too thrilled about ducted fans (other than the reasons you mentioned above) is that you may end up with un-compensated torques due to the rotation of the fans/motors. You'll need to be careful about thrust axis alignment too, irregardless of whether the thruster is a fan or a rocket. A final reason for my bias is that, having been a rocket engineer in a former life, the thought of a model shuttle with working "bang-bang" thrusters is very appealing!

 

[GregLocock]

Well, use two contrarotating fans if torque is a big issue.





Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[btrueblood]

Greg,

Two contra-rot. fans, if not coaxial, and you have a thrust vector alignment issue (the fans would need to be both synch'd to reduce torque about thrust axes, as well as thrust-matched to prevent torques about the "vertical" axis). If coaxial, the thrust vector alignment problem goes away, but... where to source a 3" coax. contra-rotating fan box?

 

[KarlChwe]

Greg, btrueblood and Rob,

Thanks for all your comments and suggestions.

The perforated floor was an idea we played with for a while, but I guess it doesn't work with our existing location restraints. We would have to build a new raised floor inside the viewing chamber. There was also the noise from the blower to think about. I like the idea, though. If I had my way, that is what we would do (stupid administrators ;-)

Rob, the compressed bottle of air with a tiny compressor might be the way to go for the attitude control jets, now that I have a source for servo valves. Thanks, btrueblood.

Recharging a compressed air bottle at a docking port is conceptually elegant, since it could mean offloading all sources of power (if we use the perforated floor.) But it seems difficult. You would need some kind of clamp or attachment mechanism to hold the model in place while the air is refilled (magnets, maybe?) It needs to press the model to itself hard enough to open a check valve.

We did consider recharging a battery at the docking port, since electrical contacts don't need much pressure or precise alignment.

The issue of uncompensated torque from the ducted fans is a good one, especially since they would start and stop often. It would cause one side of the model to dip, and perhaps drag on the floor.

There are also procession effects, now that I think about it. And gyroscopic effects, though that would be smaller. You are right, I don't like ducted fans either.

Okay, compressed air it is. Any idea what kind of pressure or volume I need? The model might weigh 10 pounds or so.

Maybe I need to go read up on rocket thrust...

Thanks again, gentlemen (and gentlewomen?)

Karl

 

[hydrae]

Karl
To toss the ideas on the wall a see if they stick.

I was recalling high school physics class
For a source of compressed air, how about dry ice inside a pressure vessel, the pressure is not very high but pressure over surface area will give the lift needed. The dry ice will also give the cool smoky effect.

As a source of power for the unit, there is an amusement park ride similar to bumper cars where the power was completely in the floor with a series of alternated painted conductor strips and multiple pickups on the cars. I assume there was a complex set of diodes on the electrical pickups in the cars to get a reliable dc current without shorting the grid. but then you will get some friction...

But if you were to supply ac to the grid add a thin coating for electrical resistance to the surface and have coils onboard the craft set in multiple directions that would act like a transformer again with the rectifying diodes. you need enough coils set in apposing directions so the attraction will be canceled out no matter what orientation. [I might be missing something here] You might have a power source without batteries...

As for ducted fans, have two fans feeding a common header the fans would be facing each with the header in between. Then tubes going to your control points with the servo valves. The fans would be rated half the supply voltage and wired in series, the torque is proportional to amps which would be the same for each fan thus cancellation of torque...

Hydrae

 

[KarlChwe]

Thanks, Hydrae.

Here's what we have settled on. We are going to have a blower floor, like in air hockey. (They are cool with the raised floor.) That handles flotation.

For attitude control and thrust, our best idea appears to be battery-powered ducted fans. Your idea of fans blowing into a plenum that are valved is good, except it is a lot easier to turn fans on and off than it is to open and close the largish valves required for the high volume, low pressure flow created by the fans.

Maybe we could use a gyroscope principle to rotate the craft, like in those satellites.

Or simple rotational friction. With a wheel mounted on a vertical axis, accelerating it would cause the craft to rotate in the opposite direction...

Karl

 

[btrueblood]

Well, it might not be the coolest way, but I can still respect your engineering choice. I think using an air hockey table probably negates one of my objections to fans (the off-axis torque problem) -- you may be able to kick up the air table pressure enough to prevent the orbiter from tilting and dragging, also use contra-rot. fans in pairs as Greg suggested.

One way to balance the torques would be deliberately mount the fans so they have fairly big "moment arms" about the vertical axis. Each fan is paired with a thruster (pointing the same direction) mounted on the opposite side of the vehicle so that their torques more or less cancel out. Then, always "fire" the fans in pairs, and pulse-modulate (and use gyro compensation) to cancel the remaining (residual) torque. You end up using 8 fans instead of four, but at least have a way to compensate for torques. Also, you can use differential fans to "dump" stored gyro torque (i.e. when the gyro is reaching, say 70% of its rpm limit, fire fans to generate a torque in opposite direction. The gyro must then brake or decelerate to compensate for the fan torque, "dumping" its excess speed).

The above torque balancing scheme is pretty much "out of the book" for 3-axis stabilized spacecraft. However, it's pretty unlikely that your human visitors can do it very well themselves. The torque/rotation limiting should be built in to an on-board micro controller (which implies an orientation sensor too, maybe a magnetic compass?). You might want to build in a speed limiting routine as well (my sons would have great fun banging such an exhibit from one wall to the other all day long...)

 

[rburton]

I have been following the discussion and wish to put in a vote against the ducted fans, for the reason that they are hard to turn on and off quickly. With the air table, the vehicle will have very low friction. Docking requires a delicate combination of thrust times time applied in a certain direction. For a 10 lb vehicle, say 5 kg, a change in velocity of 1 cm/s requires a thrust times time (called an impulse) of 0.05 Newton-sec. This is equivalent to the mass of a nickel applied for only 1 second. This is hard to do with a ducted fan.

A CO2 thruster would need about 0.1 grams of CO2 to create the change of 1 cm/s. So 1 kg of CO2 would last for 10,000 of these impulses.

Esthetically, you would have a hovercraft with ducted fans, not a rocket.

 

[KarlChwe]

Thank you blueblood and rburton.

It appears fans of any type have lots of issues: torques about their axis of rotation caused by accelerating or slowing the fans, as well as that caused by simply biting into the air, gyroscopic forces, aligning thrust axes (axises?) with the various axes of rotation in the craft...

It occured to me that the simplest solution might simply be a flywheel mounted on a vertical axis, with an electric motor. To turn the craft one way, turn the flywheel the other way. Conservation of angular momentum. Stop the wheel, and the craft stops turning.

If we want the coasting action you get with reaction rockets in space, once the wheel is spinning we just let the flywheel spin freely on its bearings. There would be some friction in the bearings, but that could serve to slow down the craft so it doesn't spin forever. We could use fake jets (like light bulbs, say) for visual effect. To speed up the spinning, just accelerate the flywheel.

The problem is in slowing the craft's spin. We'd have to somehow get the motor to apply a (variable) opposing torque to the flywheel without necessarily stopping or reversing it (although that would be possible), so that the illusion of reaction rockets is preserved.

What kind of motor can apply a certain amount of torque without fixing the speed or direction of the rotation?

Would it have to be some kind of air jet and turbine, so if the turbine is spinning, it will continue spinning against the jet for a short time before it reverses direction? Or some a mechanical connection with some kind of very controlled slipping, as in a clutch?

Could the motor act as a generator (or dynamo or whatever the right word is) drawing energy from the flywheel? One that can create a variable load?

Or a second flywheel that acts in the opposite direction? Twice the weight and cost. Also what would keep the two flywheels from spinning faster and faster forever? In satellites I think that is called "momentum saturation" or something.

Would it require electronic feedback? A custom wound motor/generator? Would it require (shudder) software?

Karl

 

[rburton]

Karl--

You have two problems: rotation and translation. Your suggestion to use a flywheel, called a momentum wheel for spacecraft (see

http://www.answers.com/topic/momentum-wheel)

is a good one. Spacecraft use very high RPM to save mass. You could use a heavy one with a reversible motor. Instead of batteries use a helical electric power cord of litz wire which is ultra-flexible.

You still have the problem of translation, but with the flywheel you can rotate first and then fire a thruster, so you need only two thrusters for docking. I am still in favor of the CO2.

 

[scooter911]

Karl,

To see a good example of ducted fans I suggest you go to ebay and type in Hovercraft, it'll give you a good idea of how long batteries will last and only cost a few bucks, be fun playing with the toy too. To simplify this, how about a few dimensions on the playing field and anticipated size for the shuttle model. As for flywheels, assuming a frictionless surface the shuttle will turn one way when spinning up the disk, will it not turn the other way when stopping the disk or at least have a tendency to turn the other way due to bearing friction?

Scooter