Mike: I am using ALNICO 8HC down to approximately zero flux and recoiling from there. This magnet us used in a latching electromechanical relay and I am attempting to model the magnetic circuit which makes it necessary to have a fairly good estimate of the recoil permeability. I have been using u=1.9 and am planning to make some measurments to confirm the value. The flux paths are well contained by the pole pieces, coil core and armature. I don't have the tools for a FEM analysis however this may be the next step. I have another question regarding the demagnetization characteristics of 8HC. Does this happen instantaneously or is there a time constant. In other words if the magnet is driven to a lower flux level by the coil for a short interval (~100 milliseconds)will the magnet permanently recoil from this new point? If you have any sugesstions or guidance I would really appreciate it.
It'll be difficult (but not impossible) to calculate that analytically (FEM would be better). When the magnet is pushed down to zero flux (i.e., the coercive point), it will not follow the 1.9 to 2.2 recoil permeability line. It will initially follow the permeability below the knee (roughly 5.0 to 7.0 IIRC) and then return to a permeability nearer to 2.0. This is because it was pushed beyond the knee of the curve.
You might want to approach your magnetic material supplier. If they are something other than a reseller, it is likely they'll have engineers on staff to help out their customers in these situations.
Regarding your 2nd question: There is a time constant for Alnico and I would suspect that 100 milliseconds is probably going be too fast of an interval for the Alnico to completely respond. I don't think it will respond as cleanly you like. This type of behavior is even more difficult to model/predict. Testing & careful measurements are by far the best way to determine what will happen.
Mike:
Thanks for the input. Since I am new to working with permanent magnets I had thought, based on what information I could find, that the recoil permeability was basically a straight line recoiling with a slope equal to the recoil permeability from the lowest point reached on the demagnetization curve. Based on the air gap and the affect of the coils the two magnets used in the relay are operated between their maximum energy points and zero flux. When the relay is latched in in either the set or reset position one of the magnets will be at (or near) its maximum energy point (minimum air gap) and the other magnet will be at a low flux level (~100 gauss)and the maximum air gap. To release the latched magnet the coil neutralizes the flux forcing the magnet to near zero flux until it transfers. Does this mean I should be using a recoil permeability of 5 to 7 since it is always below the curve or will it be a non-linear curve? Also what do you mean by IIRC?
The recoil permeability is constant (i.e., "straight line") only for SmCo and most high-coercivity NdFeB permanent magnets, because they usually do not have a knee in the BH curve.
It is that knee in the curve for Alnico magnets that prevents one from using straight line behavior. When an Alnico magnet is pushed to its coercive point, it'll follow the path of a minor BH loop that is non-linear. That minor loop will resemble the shape of the major loop, but it'll be at lower values.
It's difficult to say which permeability should be used for your application, it requires a lot of analysis to determine where the magnet is operating at every moment in the cycle you described. It can be done, I remember working through the example in Moskowitz's book, but I have to admit that I would approach it via FEA nowadays.
I've written an FAQ (faq340-1112) of various packages for magnetic modeling - both shareware & full-featured.
I'd recommend starting with one of the shareware programs first, but I can't recommend one over the others as that might violate EngTips rules against promoting/selling.
I use & like the software from IES, both Magneto (2D) and Lorentz (3D) practically every day. I've found that if the model is set up correctly, they give accurate numbers.
I've tried FEMM and liked it as well.
I've been told that Flux2D & Flux3D are rather versatile as well.
Krux: If the geometry of your latch is rotationally symmetric, then you'll be able to try one of the shareware 2D/RS programs listed in the FAQ. They will be a lot easier to learn too.
If your geometry requires 3D, be prepared to spend quite a lot of money (& time for training) since I believe all 3D solvers are quite expensive.
sreid:
Unfortunately the device (a mag latch relay) already has the ALNICO 8HC magnets in the basic design. It is an old design dating at least 50 years but still being manufactured by the buckets full.
High coercivity Al8 will have a fairly straight recoil. You can probably get away with this estimate, maybe add a offset to it to shift the B value up slightly.
The time contant will also be a function of the size of the part. Very small diameter parts will not have much, thick parts can have a lot. As I recall we tried to keep magnetizing pulses longer than 500ms to assure sturation.
There are still good reasons to use alnico, such as; mechanically robust, corrosion resistant, temperature stability.
= = = = = = = = = = = = = = = = = = = =
Corrosion, every where, all the time.
Manage it or it will manage you.
EdStainless and MagMike:
The ALNICO 8HC magnets I am concerned with have an area of 2 cm^2 and a thickness of 1/16 inch. The Hc value for a fully saturated magnet is 1800 Oersteds with a Br of 7.6 kG. The magnets operate from approximately zero flux to approximately the maximum energy point and are generally not fully saturated. At the maximum energy point the air gap is minimum and the permeability of the magnet will have the greatest effect on the magnetic circuit. As the air gap increases the reluctance of the air gap dominates and the permeability and the resulting reluctance of the magnet has only a small effect on the total circuit reluctance. Given the recoil permeability at the maximum energy point I should be able to construct a linearized model that is accurate at this point. I will be attempting to measure the inductance of the circuit on both sides of the maximum energy point and extract the recoil permeability from the calculated circuit relucance. Let me know what you think of this method.
I had always assumed that magnetically latched relays only used the permanent magnets to hold the relay armature in one position or the other. I thought that the set and reset coils were a separate magnetic circuits that moved the armature from one position to the other.
sreid:
This particular relay is of the full crystal can variety and has both set and reset magnets. The two coils are on a common core and are attached to the inner (south) pole pieces of each magnet. A moving iron armature connects the outer (north) pole piece of the "set" magnet to the inner pole piece of the "reset" magnet when in the "set" position and vice-versa when in the reset position thereby. In the set position magnetic circuit includes only the set magnet leaving the reset magnet with a relatively large air gap and vice-versa for the reset position. The electrical contacts are attached to the moving armature with a springy material such as beryllium copper. As the contacts close part of the magnetic force bends the spring thereby storing some energy. To transfer the relay from set to reset sufficient ampere-turns are applied to the reset coil to reduce the latching force generated by the set magnet until it equals the spring force whereupon the contact jump open transferring the spring force into rotational inertia. The combination of the magnetic force generated by the reset coil plus the rotational inertia imparted by the spring moves the armature toward the reset position where the reset coil now aids the attractive force generated by the reset magnet. These two forces combine to overcome the spring force and latch the relay in the reset position. I hope this description is not to confusing. I have some cartoons that illustrate this action; however I don't see a way to attach them.
