Re: [femm] Re: Incorrect force direction

["Steven Stretz" <stretzsj@xxxxxxxxxx>]

I agree, this was a great problem. I, hopefully others, learned something.
Thanks
----- Original Message -----
From: "Rob Tickle" <tickle@xxxxxxxxxxx>
To: <femm@xxxxxxxxxxx>
Sent: Wednesday, January 03, 2001 10:51 PM
Subject: [femm] Re: Incorrect force direction


> This was an interesting problem to me because it showed me how my
intuition
> could be misled by the geometry of the problem. One would think by looking
> at the problem (I did anyway) that there could be a centering force. It's
> obvious that the magnet would experience an attractive force before it
gets
> fully between the bars - calculation with femm agrees with this if you
move
> the magnet a bit to the left. I think the position it is in for this
problem
> is right where it hits the fuzzy zone where that force is disappearing. To
> convince myself of this, I grabbed the magnitude and normal components of
B
> along the right and left side contours where the force line integral is
done
> and put them in a spreadsheet. Then I calculated the x-component of the
> force from the stress tensor for right and left sides and added them up.
The
> graph shows the right contour contribution (red) and the left contour
> contribution (grey) with the total x component of force in black. It
> illustrates what David mentioned in the previous post about two large but
> opposite sign contributions summing to zero (or close to it). 2 notes:
this
> illustrates force up to a multiplicative constant since I didn't multiply
by
> path lengths. Also, I didn't include the x-component of force from the top
> and bottom contours, since they are of equal magnitude but opposite sign,
> and I assumed from the symmetry of the problem that they would cancel.
> Based on this result though, I could see how the answer might be down in
the
> noise so to speak.
>
> I tend to trust this result, but I would be interested in seeing some pics
> of Greg's actual setup and magnets. I know the NdFeB magnets are highly
> anisotropic and very strong; one possible source of stray force to look
out
> for would be if the magnetization direction of the magnet were not
> completely perpendicular to the bar surfaces. Does the actual setup show a
> true centering force in the center of the bar; i.e. does the force
direction
> switch signs as you roll past the middle?
>
> Rob
>
>
>