Sizing an electric motor 2

[JeffreyLan]

Hi,

I am a Junior in ME and currently working as a graduate assistant at a wind tunnel research facility. I have been given the task of designing a mechanism which would remove a long pitot probe from the flow stream in the wind tunnel. The mechanism will be able to move a 21 in. long 5/16 in. diameter hollow metal tube 90 degrees (if you would like a diagram to better understand this, I can do that; I have already made up a CAD file that shows the scale motion). The wind tunnel will run at a maximum velocity of 70 m/s and I have calculated that the maximum torque on the pitot will be 9.36 in*lb (1.05 N*m, 149.8 oz*in) and that whatever motor I use must be able to supply this holding torque (plus about a 30% safety factor). Ideally, the operation of the motor will be done through the program LabView, and the graduates informed me that I have a 10V supply to work from.

At this point, I have run through the entire beginning to mid level design process stages, and I am looking at different motors to suit the application. Right now I am leaning toward using a stepper motor since I believe that a DC motor with encoder and feedback loop would be too complex. It is also necessary that for good pitot measurements, the pitot tube should be able to maintain its angular position +/- 1-2 degrees...the stepper motor, I think, should have no problem with this.

I am really interested in hearing what motors other people think might work better and especially WHERE reliable motor suppliers can be found. This is only my second exposure to electric motors (I took a Mechatronics class this past semester where we learned C and how to use the Dragon 12-Plus to run dc motors, steppers, and ultimately make a robot).

Thanks for the input in advance!

Jeff

 

[MikeHalloran]

Swinging the probe into and out of a stream doesn't require feedback, or at least not much.

An air cylinder is the first thing that comes to mind.

A stepper could do it, if its holding torque is high enough. Note that steppers typically take a half-step when you turn the power off completely, so you might run it at reduced current on two coils to hold it, and/or arrange some cooling, or deal with trying to make it lay against the stop magnetically when you turn it off, which may be too interesting.

For feedback you need two limit switches, to say {deployed | in transit | stowed}. Think of them as a two-bit encoder.

A DC motor could do the same thing, arranging the 'stowed' limit switch to shut off the motor,
and the 'deployed' switch to reduce the motor current to just enough to hold the probe in position,
against a mechanical stop that you would shim to set the deployed position to any precision you like.
Maybe you can find enough leakage flow to cool the motor, or let it hang into the tunnel a little.

If the DC motor heating bothers you, arrange it to run a quarter of a four-position Geneva wheel, so it locks mechanically against the mechanical stop and you don't need holding current to keep it there.




Mike Halloran
Pembroke Pines, FL, USA

 

[ScottyUK]

A DC motor with a leadscrew and limit switches would achieve sufficient accuracy I think.

Keep the leadscrew pitch fairly fine, so it resists any tendency to reverse drive. Short the motor terminals using a relay to achieve a rapid stop once you reach the desired position and have removed the power. Depending on how rapidly you need to deploy and retract the pitot tube, you could use a reduction gearbox prior to the leadscrew which would increase the resistance to reverse driving, reduce the motor power required, and allow greater positional accuracy.

 

[JeffreyLan]

Mike, thanks for the feedback! Actually, your first guess is pretty darn good. The original setup for this device at the first wind tunnel we used was indeed a pneumatically driven device (I've asked the grads and they really don't know how it worked though). The problem is, our pressure tab is on the INSIDE of the chamber and, well, I don't think any of us would like to stand in there with the hyper-hurricane force winds. I am going to recommend to them that we could possible re-route the line out of the chamber, where we are, and attach a valve there. The thing is though, I think they want a new system which would be operable through the computer program LabView. Would the stepper motor really get that hot during holding? I would estimate that a single test could last 5-10 minutes, and, as you can imagine, there is a large pressure gradient across the chamber which has a small airflow across it for cooling. Unfortunately, the testing we do does not permit the motor to be placed in the stream for maximum cooling, nor can we divert some of the incoming air. I do like the idea of the electromagnetic latch and the limit switches (but I was thinking I could get around that by just using the stepper motor and commanding an exact 90 deg angle change,then leave the voltage supplied to hold the position). In my opinion, the problem with the DC motor is that, when it is shut off after reaching one of the limits, it can still move...unless you create some sort of latch, which means more moving parts, 1 or 2 small servo motors, and, in general a more complex program. My effort is to make this as simple as possible.

Thanks for the insight!

Jeff

 

[JeffreyLan]

Mike,

I forgot to mention that I really like the Geneva wheel idea. I'm going to look into that some more!

Thanks!

Jeff

 

[JeffreyLan]

Scotty, thanks for the ideas. One of my first ideas was like this. I was thinking of using a driven worm gear to turn a granny which the pitot is mounted on. This would indeed prevent the back-driving. My estimate is that a normal opening would take around 3-7 seconds which is about 0.2 to 0.5 rad/s (2-5 rpm), but there is no limit on how much time it must take, just that it is reasonable. I am going to revisit your idea since it may be quite the bit simpler and cheaper than a stepper motor now that I think of it. The worm/leadscrew would also output a large amount of torque, and I would still need limit switches to indicate deployed|in transit|stowed like Mike said but I think this setup could be a winner.

Thanks Scotty!

Jeff

 

[MikeHalloran]

On reflection, I like Scotty's idea better. Steppers can get real complicated real fast unless you've dealt with them before and have the appropriate crap and expertise lying around. They're fun to learn about, if you have a spare month.

You can use the 'deployed' limit switch to reduce the drive current by switching in a resistor, and keep the probe jammed against a mechanical stop, which through even a low ratio worm gear, should work okay. You reverse the polarity and remove the resistor from the circuit to get it unjammed and going the other way. You can do the whole thing with relays, which are easier to troubleshoot when things don't quite work as you planned.



Mike Halloran
Pembroke Pines, FL, USA

 

[racookpe1978]

Your 10 volt supply - controlled however you wish from the LabView program - could actuate a solenoid on an air supply valve to the air drive motor.

The air drive supply - if you have none available - need only be a $200.00 off-the-shelf 120 volt compressor and hose from any hardware store.

Also. Don't be afraid to just drive it manually from outside the tunnel.

 

[ScottyUK]

Addressing a point you made earlier, if this is a typical low speed wind tunnel - i.e. not a supersonic / hypersonic type - then a standard brass cable gland should be more than adequate to route cables or impulse lines for an air actuator though the tunnel wall. If you are really worried about running airlines through the tunnel wall then Swagelok and their peers make fittings designed to provide a feedthough connection through a panel or bulkhead.

 

[JeffreyLan]

racookpe1978 and ScottyUK,

I was talking with the grad student the other day and he says that the main idea is to get away from the pneumatically operated system. Basically he wants a whole new electric motor system which will operate the pitot. Also, the problem with manually operating the mechanism by crude linkages from outside is that our available ports into the chamber are already stuffed with instrumentation lines, and our current setup would make it near impossible to have a rigid link attachment (its also about a 10 ft. run to the pitot from the ports with multiple obstacles in the way).

At this point, in my mind, it's being narrowed down to one of the following:

1. Stepper Motor
- Pros: With the correct holding torque, it will be able to maintain the angle exactly.
- Cons: Increased complexity and cost (need a motor driver and, I think, some sort of micro-controller), not to mention a programming code. Question: Would I need a motor driver and micro-controller for any mechanism I choose that involves a motor?

2. Geneva Wheel (

http://www.youtube.com/watch?v=t7ZnHDdIvn0)

-Pros: Exact 90 degree turn with guaranteed positioning of angle (cannot move once at 90 degree position). Also, the motor no longer is the device providing the reaction torque to keep the pitot in the same place. That job now rests on the physical contact between the two rotating members (which is guaranteed to work). Lower complexity (a dumb DC motor with a low rpm)and generally lower cost.
-Cons: Hard to find/buy a geneva wheel, but could be a good opportunity for me to head to the machine shop and learn to make the parts. Also, the DC motor must turn one way, then turn back the other way to make the Geneva wheel turn both directions. This should not be too difficult (either command it to turn for a certain time, or use limit switches like suggested by some of the responses).

Question: Is it a bad idea to run a brushed motor in one direction, stop it, and then run it in the opposite direction? I have heard that the brushes are angled in a certain direction and should only be rotated one way. I have also heard that they also make straight perpendicular brushes. Any help on this would be great!

3. Geared DC Motor or DC Motor turning a worm (leadscrew) which turns a larger gear - suggested previously by ScottyUK.
-Pros: For the worm mechanism, the motor does not have to supply a holding torque during the time in the wind tunnel because the worm prevents back-driving. The geared DC motor should not require any holding torque either since it will have a large gear ratio. Lower complexity and cost and readily available.
-Cons: Can't really think of any.

Thanks again guys for your help.

Jeff

 

[MikeHalloran]

A stepper will not hold an angle exactly against a variable load just from its detent torque. The detents are a little bit elastic.
Half a degree pretty much requires a mechanical stop of some kind no matter how you move the probe.







Mike Halloran
Pembroke Pines, FL, USA

 

[LionelHutz]

This sounds like a job for a 12V linear actuator. Basically a self holding screw drive with limit switches and a feedback pot built-in available with different travel lengths and cheaper than you could build. You can run it right off your 10VDC supply if it's capable of sourcing the current.

 

[itsmoked]

Heartily agreed with solution.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[roydm]

An Auto windscreen or window driver motor operates on 12 Volts so it will easily operate on 10
Just another cheap way to think about.

Roy

 

[JeffreyLan]

Thanks again everyone for the suggestions. I'm still working on getting this thing to fit in the structural constraints of the wind tunnel, but when I do get it finished, I'll be sure to post a picture or a video of it in operation and the details of the mechanism. Again thanks very much to everyone for the help!

LionelHutz and roydm: This question has been plaguing me since the beginning of this project, the answer to which was never taught in my Introduction to Electrical Engineering class. Is it possible/safe to run a 12V motor off of a 10V system? Does it not interfere with the operation of the motor? (not to mention that the torque rating will be lower and I would have to design around that too)

 

[LionelHutz]

Yes, you can easily run a 12VDC motor on 10VDC and yes will will lower the torque.