You are here:

Physics/Rotation using magnetic attraction


Magnetic rotation
Magnetic rotation  
QUESTION: Hello Steve,

Thank you for all your help with my previous questions. I have been doing some experiments with magnetic fields, and the movement of metal objects near those fields, and I am wondering if I could ask you a question on rotating an object using magnetic attraction.
I have a design for a plastic cube-like object that is suspended on a plastic axle. As shown in the attached image, inside the curved underside of the cube there is a ferrous metal section, and in the upper part of the plastic axle, is a permanent magnet which attracts the ferrous metal section of the cube. (The cube can be held in place mechanically, or rotate freely.)
The cube is positioned as shown in the image and held in place mechanically, with the ferrous metal section of the cube directly underneath the permanent magnet in the axle. Two inactive electromagnets are positioned below the cube, equidistant from the ferrous metal section.
I want to rotate the cube 45 degrees. To do this, I want to release the cube so that it may rotate freely, and at the same time I want to power up one of the electromagnets in order to attract the ferrous metal in the cube, rotating the cube 10 degrees. If I turn off the electromagnet after rotating the the cube 10 degrees, will the attraction of the ferrous metal section of the cube to the permanent magnet in the axle cause the cube to continue to rotate a further 35 degrees?

(apologies I have not explained clearly!)
Thanks again and best regards,

ANSWER: Hello Eddie,

It would tend to do that yes. After rotating 10 degrees, the ferrous metal section would be closer to the permanent magnet in the axle and therefore there would be a torque to cause the cube to continue rotating. Whether or not that torque would be enough to overcome the torque of the friction in the bearing is something I don't have the information to determine.

Looking at your drawing, it seems that there might be adjustments that I could suggest. There may be reasons these adjustments are not possible. I don't have sufficient overall picture of your project to know your existing restraints. But consider these thoughts:
> Could you revise the permanent magnet in the axle such that instead of north-south being up-down, north-south is left-right?
> Could the permanent magnet in the axle be in the lower part of the cylinder?

My crude sketch should give you an idea of what I mean. Then, when the cube has rotated 10 degrees, it would significantly decrease the distance between the ferrous metal section and one of the poles of the permanent magnet and should increase the torque to achieve the remaining 35 degree rotation.

I hope this helps,

---------- FOLLOW-UP ----------

QUESTION: Hello Steve,

Thanks very much for your answer. I apologize, I should have noted that I actually need to be able to rotate the cube 45 degrees in both a clockwise, and also a counter-clockwise direction from that original position. That is why I had the permanent magnet positioned in the upper part of the cylinder with the north-south/up-down orientation. Do you think this is the best way to position the magnet with this goal in mind? What would you say is the ideal point on the cylinder to put the magnet? Upper, lower, middle?

Regarding the overcoming of the torque of the friction that needs to be achieved when the permanent magnet's attraction "takes over" from the electromagnet and further rotates the cube, I am assuming that one of the conditions governing the torque would be the strength of the permanent magnet's attraction - would that be correct? Would increasing the permanent magnet's strength make the operation easier? Should I aim to make the permanent magnet as strong as possible, or would there be any adverse effects of the permanent magnet being too strong that I should try to avoid?

Thanks again. Your answers are so thorough, and I really appreciate the time you take.

Best regards,

ANSWER: Hello Eddie,

Yes, I expected that you would want to be able to rotate it either direction. I suspect you made a point of explaining that because we have confusion someplace. I didn't think the ferrous metal section was magnetized. Is it? If not, perhaps you didn't realize either the north or south pole could attract unmagnetized ferrous metal. You have been playing with magnets. Try picking up a nail with either end of a bar magnet. It'll work.

And yes, one of the conditions governing the torque would be the strength of the permanent magnet's attraction. The only possible disadvantage I can imagine due to excess magnet strength is 2 of your units interacting if they got too close together.

I hope this helps,

---------- FOLLOW-UP ----------

Magnetic Rotation
Magnetic Rotation  
QUESTION: Hello Steve,

Thanks again for your answer. Sorry, my laborious description was just me trying to get it all clear in my own mind :-) I was not actually planning on having the metal in the cube magnetized, so it is just plain old ferrous.
I am now trying to work out how much attractive power is needed by the electromagnet to rotate the cube the initial 10 degrees, and also how much time the entire 45 degree rotation will take. Last year, you were kind enough to explain how I could calculate the angular velocity when rotating the same object 90 degrees.
At that time you gave me the following formula
alpha = tau/I
You also gave me some help calculating the torque and the rotational inertia for the particular cube-like object I want to rotate -
alpha = (F*d/2) / ((1/12)*m*d^2)

I am wondering if I can use this equation for calculating the angular velocity of this new rotation. I am assuming that the rotational inertia of the cube in this new position would be determined by a new calculation, is that correct? Would the torque be the same, or would that also require a different calculation?

Actually I was also wondering whether it might be feasible to rotate the cube through the initial electromagnetic attraction to a degree less than 10 degrees. I would like to reduce the power used by the electromagnet as much as possible. In your last answer you confirmed that the strength of the permanent magnet in the axle would influence the torque. In theory, if the permanent magnet was very strong, could the cube be rotated 1 or 2 degrees before the permanent magnet's attraction caused it to rotate the full 45 degrees? What do you think would be a safe working model to calculate how much the initial rotation would need to be in order to pull the ferrous section into the permanent magnet's field and cause the cube to rotate further?

Sorry to pile so many questions into one! One answer spawns more questions.
Thanks again for your kind help. I appreciate it.


Hello Eddie,

Before I get into your new questions, I want to revisit the answer I gave you 2 replies ago in which I proposed a revision to the permanent magnet embedded into the axle and included a sketch. I think that idea has potential and I want to be sure you understood the revision I proposed. In your first followup after that proposal, it seemed that you thought my proposal would not allow clockwise rotation. But it would!

Before I go on, I need to confirm something. The image that came with this current question doesn't show the permanent magnet in the axle. Can I assume it is still in your current plan and is oriented as shown in your figure that did show it? Assuming yes, I'll continue.

If the permanent magnet is oriented with N-S axis horizontal, the distance between the magnet and the ferrous section decreases more significantly after the initial rotation of 10 degrees. And again, it could be rotated either direction after giving it a start in the desired direction with an electromagnet. The ferrous section will be attracted by the end of the magnet that it is close to -- either the N or S end. I believe the attraction would be significantly stronger with N-S axis horizontal than if it were vertical (as shown in your first drawing showing the permanent magnet in the axis).

Now about the current questions you have asked. The rotational inertia does not vary with orientation. It is a measure of how hard it is to give an object a spin. (Imagine that you had a bicycle inverted, and used your hand to give the rear tire a spin. Then assume you added weights to the wheel around the circumference -- at regular intervals so it would still be balanced, and then again used your hand to give the wheel a spin. It would be harder to get it up to the speed it went without the added weights. The weights increased the rotational inertia. Just like with linear motion, increasing inertia makes it harder to change the speed of an object.)

The rotational position of the cube does affect the torque. It would depend on the geometry and the force of attraction.

I am not optimistic that 1 or 2 degrees of rotation due to the electromagnet would be enough to be confident that the permanent magnet could do the rest. But it seems that decreasing the task of the electromagnet phase of the rotation could be the subject of tests after you get this built. I am guessing that you could adjust the time duration for the pulse to the electromagnet.

I hope this helps,


All Answers

Answers by Expert:

Ask Experts


Steve Johnson


I would be delighted to help with questions up through the first year of college Physics. Particularly Electricity, Electronics and Newtonian Mechanics (motion, acceleration etc.). I decline questions on relativity and Atomic Physics. I also could discuss the Space Shuttle and space flight in general.


I have a BS in Physics and an MS in Electrical Engineering. I am retired now. My professional career was in Electrical Engineering with considerable time spent working with accelerometers, gyroscopes and flight dynamics (Physics related topics) while working on the Space Shuttle. I gave formal classroom lessons to technical co-workers periodically over a several year period.

BS Physics, North Dakota State University
MS Electrical Engineering, North Dakota State University

©2017 All rights reserved.