We are pretty good about modeling air gaps, and when we know the power draw is low (and hence low saturation) we make very sure the gaps are there. Actually, I tend to over estimate the gaps, typically I would model a press fit as at least a .002 - .003" radial gap, more if we actually expect the gap to cause an issue. As for driving the solenoids - we have a few ways. But the most common for our group at least is "straight" DC, and that is where we have seen the specific problem I'm talking about.
Also, in answer to Dave's comment, we definitely are aware of how treacherous a force measurement can be on a solenoid, especially a small one. But the two most common methods we use to measure force (deadweight lifting and strain gage) have both shown the same type of problems to one degree or another, and we can usually get them to agree as long as the side loads aren't too hideous.
I'm pretty impressed - this question got more of a response than I expected. Thank you! By the way, I had a brief correspondence today with a company that does B-H analysis on steel samples, and the person I spoke with said that machining stresses can increase the Hc of a soft steel by a factor of 2 to 5, with a similar change in the maximum permiability. Since we are often dealing with thin sections, I'm thinking it probably is going largely back to this. I did a quick test today where I multiplied the Hc of my 1215 curve by 5 and it made quite a difference in the model.
Andy
--- In femm@xxxxxxxxxxxxxxx, Keith Gregory <k.gregory@xxxx> wrote: > Another thought - how is the solenoid put together? Have you got any small > parasitic air-gaps that are not being modelled? > > Keith.
Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/