: Interesting. The Lee Valley site mentions "magnet cups" as a method
: for "focusing" the magnetic field: "When a magnet is placed
: in a ferromagnetic cup (Figure 3), the cup further magnifies the effect by
: eliminating the air gap (air is a poor conductor of magnetic fields) and
: brings both poles of the magnet to grip on the same surface. This is
: similar in principle to a horseshoe magnet. A rare-earth magnet in a steel
: cup provides four times the strength of a bare magnet. A cup provides a
: disc magnet the optimal magnetic flux focus into the smallest gap
: area."

Thanks for the comments,people. I have spent a little time with the 1 inch dia. magnets in the appropriate cups and washers (Lee Valley-items-one of our few Canadian claims to fame). I agree that best attraction is obtained with surface contact. However, I also played with them separated by what I approximated to be the thickness of two layers of 6 oz fibreglass (one on each surface, and found the vertical retention quite good. Sliding retention not as good, but sliding would be restricted by the cockpit rim. I guessed the need for four 1 inch magnets with cups would suffice. I suspect that with careful insetting of the gaskets that an adequate amount of compression could be accomplished--at least as good as trying to compress with the catenary of straps.

Interesting thoughts brought up regarding the other effects. I will have to play with a compass to see whether the effect will affect my ability to navigate with the cup/washer system. My credit is such that the effect will probably not be noticed.

Jim Lindamood

: (http://www.leevalley.com/hardware/page.asp?page=40077&category=3,42363&ccurrency=2&SID=)

: Have your tried these steel cups? It seems to this nonphysicist that a
: magnet-to-magnet setup would double the attractive force, while the steel
: cup-to-magnet would quadruple it, if the Lee Valley claim is correct.
: Intuitively it would appear that either setup would follow the inverse
: square law of force vs. distance, but maybe you have some experimental
: results otherwise?