Science Buddies: "Ask an Expert"

Good day to you.

I'm in the process of designing a tubular motor that travels along a heavy gauge cable, with the cable running through the centre of the motor. It's a slight variation on the maglev train story.

The basic idea is to use 4 narrow AC linear induction motors, joined together at 90 degrees to each other, with each motor's pole faces pointing inwards towards the cable. The motor gets its power from a separate feeder cable (AC single phase supply with constant frequency).

I'm using a FE package (Quickfield 5.4 Lite) for magnetostatic analysis which is very useful, but I'm having some trouble with optimising the design. The best approach I can come up with is as follow:

1. Assume a cable diameter (along which the motor will move) which remains constant.
2. Using the frequency supplied to the motor and the required speed, choose the pole pitch. Once chosen, this will also remain constant (speed = 2 x freq x p.pitch).
3. Design a basic geometry around the chosen pole pitch, making initial assumptions where necessary.
4. Choose a wire gauge for coils and work out the maximum space available for the coils (packing factor = 0.6). After allowing enough space for insulation, work out maximum number of turns and resulting current density in each slot.
5. Calculate the mass of the entire device.
6. Calculate static friction and gravitational forces that need to be overcome before motor starts moving.
7. Work out how much magnetic flux is produced by the four sections of the motor, using software simulation in Quickfield. Calculate the induced current in the heavy gauge cable.
8. Work out if the interaction of this induced current and the magnetic field from the coils provides enough thrust to get the motor moving.
9. If not, return to step 3 and try again!

That's the theory, anyway. Is this the best approach?

The main problem that I'm facing is that I'm also not entirely sure how to calculate the thrust developed by the motor. This makes any attempt to optimise the design a hit or miss affair. Any ideas?

If anyone knows of useful online resources or books worth ordering that relate to this subject, please let me know.

Thanks in advance,

Jack

As you have already discovered, this is a complex electro-mechanical problem. I'm not sure we can help all that much. Have you searched for research papers? Can you point us to them? This is a case were unless somebody just happens to have done work in this field, you are ahead of us.

Thanks for the reply.

As you have already discovered, this is a complex electro-mechanical problem. I'm not sure we can help all that much. Have you searched for research papers? Can you point us to them?

As you rightly say, it seems rather tricky. However, once the initial shock factor has subsided, it's conceptually not too complicated and I've managed to make some progress. There is some literature on linear and tubular motors in research journals like IEEE Transactions but nothing directly related. I'm basically looking through every textbook on electrical machines I can find and trying to piece it all together that way. "Propulsion Without Wheels" is a classic book on the subject by Laithwaite (the inventor of linear induction motors?) but it may be a bit out of date (and out of print?). Anyway, if anyone's interested, I'll be more than happy to share whatever I come up with once I've managed to digest all the theory!

I had hoped to purchase a few books on the subject via amazon but there are loads of titles available on electric motor design and I've no idea which to go for. Perhaps someone can point me in the right direction?

Cheers4now

Jack

Ok, so you think you are dealing with something unique and you have to start from first priciples...

Do you understand Lorentz Forces? http://hyperphysics.phy-astr.gsu.edu/hb ... agfor.html If not, you need to understand them. That will get you an initial guess at what the static trust might be with respect to time before you start moving.

Unfortunately, because you are using alternating current, you have the four Maxwell equations to solve concurrently and they all interact and when you start moving, the equations change quickly.

I'm also not sure you can predict or control in which direction you will that will be without some "starting circuit".

I'm thinking the problem might be simpler either using DC stepper motor concepts where you can effectively vary the frequency or using a 3 phase design where the starting conditions are more easily controlled. You might also consider supplying a current through the cable to simplify the equations by biasing the operating point.

Well I'm proving I'm not an expert proof reader! I meant to say:

I'm also not sure you can predict or control in which direction that it will go without some "starting circuit".

I'll need a bit of time to check out the Lorentz Force link, but that's great; somewhere to start. For now, the basic idea is just to get the thing moving, albeit it jerky movement or whatever. Of course it would be good to have it gliding gracefully like a swan. A bit unrealistic, but who knows? It might just happen. I've got some idea about Maxwell's Equations but I can see the whole thing getting a bit scary if all variables are taken into account. As I say, getting it to move at all will be a good start.

I'm thinking the problem might be simpler either using DC stepper motor concepts where you can effectively vary the frequency or using a 3 phase design where the starting conditions are more easily controlled.

This might in fact be the way to go. I'll need to give the control aspect some thought, which will definitely present a few decent challenges as well!

You might also consider supplying a current through the cable to simplify the equations by biasing the operating point.

The first stage of the project is to get the motor working when fed with a pair of independent feeder cables. However, the next stage involves incorporating a current transformer into the motor (in the form of toroidal rings) to draw current from the inner cable itself (which will be carrying something in the order of 360A!). I don't imagine this is the type of biasing you are talking about. If anything, the situation then becomes even more complicated, but we shall have to cross that bridge when we come to it.

Anyway, it's all good fun and gets the grey matter working! Let's see how things develop. Thanks for your advice. I'll post details of any progress that is made.

At 360A you are out of the science fair league which is what this site is intended for.

Right. Thanks for letting me know.