Steel laminate core radial PMA build

[CraigM]

Hello to all,

I recently came across a source for silicon steel sheet (8" x 36" x .013") at a local metal salvage yard. After posting in Newbies about options for using this material and some helpful advice from Flux I thought I'd try my hand at building a radial PMA. This is a first build for me so hopefully I'll make a bunch of mistakes and in the process maybe learn a thing or two.


I did some preliminary drawings and in doing so wondered how I would lower the stator core into the ring of magnets (drum) without it crashing or losing fingers. I studied the jacking screw method outlined in Hugh Piggotts book and thought it may work with the radial design as well.

Some of the obstacles I read about concerning the radial PMA design is finding suitable core material and then if a core is found how to fabricate a matching drum. I believe I've found a good source for core material so now I can fabricate a core to match an existing drum inner dimension. Another obstacle I've often read about is a radial design is not adjustable for air gap (flux weakening). This is where the jacking screw idea comes in. If I can control lowering the stator into the drum then I can also adjust how much of the stator is contained in the drum. If the stator is fully positioned perpendicular to the magnets then I have 100 percent flux capacity. If I back the stator out from the drum half way then I have 50 percent flux capacity and have effectively reduced the alternators strength. By doing so I can now adjust the power curve of the alternator to the power curve of the blades. So tell me, am I all wet with this theory?


This drawing shows a cross section of the radial PMA. The drum is made from a 6 inch section of 8 inch nominal schedule 40 pipe. Hub and spindle are dexter trailer 4 on 4" center. Magnets shown are 2" x 1" x .5". Stator is 2 inches wide with diameter to fit opening. Round or square axle pipe is used to carry the stator which can slide along the pipe with an adjustment screw.

My concern is; Will this bind up when the stator is lowered using only one 1/2"-20 bolt? Will this method allow adjustment similar to gap adjustment in an axial machine?



Here's where I'm at so far. Thought if I could make a good perpendicular cut to the length of the pipe I may end up with a fairly concentric drum. Didn't quite work that way but in a pinch instead of turning the ID on a lathe the magnets could have been shimmed to fit. Turning on a lathe was just easier at this point.

The plate with the bolt pattern was cut with a laser and is just for reference at this time. Now that we know it fits it will be cut from 1/4" steel plate and welded in place.

Still have a bunch of questions concerning coil fabrication and how to attach them to the core but trying to not get too far ahead of myself.

Thanks everyone for getting me this far.

CM

[SparWeb]

Holy smoke, that is a neat idea.

I've got a few questions.

I'm going to assume that the stator is the central part of the design and you will have magnets on the ID of the 8" pipe, which sounds reasonable.  However, the stator undergoes torque when the generator generates power.  How would the stator be fixed to the mounting shaft and be movable at the same time?  Maybe a spline...

You may be "out there" on the cutting edge when it comes to de-cogging your generator.  The models for the typical motor-conversion may work, but maybe not...

Have you selected a number of poles/windings yet?

There is never "too much flux".  I hope you bench test the generator before trying to make the rotor blades.  But take steps beforehand to select wire that will get you in the ballpark.  Later tests will give you the cut-in speed and wire connection types that you can suitably use.  Or show you need to re-wind and try again.

Something very fundamental:  which way are you going to stack the laminations?  Hopefully you aren't going to "wrap them" around the spindle axis.  The photo of your scrap iron doesn't give a scale to judge how big the circles can be cut from the strips.

Good luck, and I'll be watching with interest!

[Flux]

I thought I replied to this earlier but it seems to have got lost.

Yes there are some interesting ideas here. Although I thought the core material would better suit a toroidal core, it would be possible to make a small radial machine with the strip being 8" wide.

I would have found it far easier to adopt a standard motor core and make the magnet drum to fit, but it can be done the other way, it will take time and patience to do one disc at a time without suitable tooling to punch a disc at a time but can be done even by laser cutting if necessary.

I don't think you will manage to slide a core on the shaft but you can try. Moving the stator axially does give a degree of control over the characteristic, I don't really like it but that is partly as a result of never running machines off magnetic centre, it doesn't really matter but there is a slight issue that you are bringing the end connections under the influence of the magnets ( probably won't matter)  I would aim for the correct winding ( test coil or similar) but you could do some fine adjustment with axial displacement. I would shift the magnet drum rather than the stator bu spacers between the drum and the hub.

I assume these will be smooth punchings with the coils in the air gap so you have no cogging problems.

With such a small machine you can place the magnet drum over the stator by hand if the stator is securely held and you keep fingers well clear of the danger area. Alternatively you can clamp the magnet drum and lower the stator with a hoist. You could also use long jacking screws through the drum and on to the core but for such a small machine I wouldn't bother, its a slow process and the other methods work ok.

I built something that size with voice coil motor magnets and it took a 6ft prop quite nicely, with your better magnets you may get up to perhaps 8ft.

[CraigM]

I'm going to assume that the stator is the central part of the design and you will have magnets on the ID of the 8" pipe, which sounds reasonable.  However, the stator undergoes torque when the generator generates power.  How would the stator be fixed to the mounting shaft and be movable at the same time?  Maybe a spline...

My thoughts (and realize this is a newbie speaking) are to use 2” square tubing as used in trailer axles for the mounting shaft. The square tubing will not allow the stator to rotate on the shaft as long as the thru hole in the stator is also square. It should however give the stator enough bearing surface area to allow it to slide fore and aft even with a close tolerance fit. The square bearing surface of the stator will most likely be made of plywood for ease of construction with epoxy holding the laminates to the bearing surface.

You may be "out there" on the cutting edge when it comes to de-cogging your generator.  The models for the typical motor-conversion may work, but maybe not...

The mounting surface for the coils will be a smooth cylinder with the coils mounted in the air gap. From my research this will allow zero cogging.

Have you selected a number of poles/windings yet?

For the inner diameter available, and using 1” wide magnets, 14 poles seems to be the best fit. Wound with overlapping coils will allow 3 phase, 7 coils per phase, 21 coils total.

There is never "too much flux".  I hope you bench test the generator before trying to make the rotor blades.  But take steps beforehand to select wire that will get you in the ballpark.  Later tests will give you the cut-in speed and wire connection types that you can suitably use.  Or show you need to re-wind and try again.

Single coil bench testing will be required and I'm sure I'll need help when I get to this point. Star, Delta, Jerry will also need consideration and no doubt I'll be leaning on you guys for advice.

Something very fundamental:  which way are you going to stack the laminations?  Hopefully you aren't going to "wrap them" around the spindle axis.  The photo of your scrap iron doesn't give a scale to judge how big the circles can be cut from the strips.

The material I found is 8” wide and I'll need to cut approximately 7” O.D. rings and stack them into a laminated core. At .013” thickness I need to cut over 150 rings to make a 2” thick core. I haven't settled on an exact method to manufacture the core but I'm leaning towards building the bearing surface first and then stacking the rings over the bearing surface while using thinned epoxy to glue and clamp the whole mess together.

I would have found it far easier to adopt a standard motor core and make the magnet drum to fit, but it can be done the other way, it will take time and patience to do one disc at a time without suitable tooling to punch a disc at a time but can be done even by laser cutting if necessary.

The upside is I'm fortunate to work at a machine shop so laser cutting and other machine tools are available to me. The downside is I'm not a machinist. But I'm also trying to apply a simple approach that would be successful for anyone with a drill press and hand tools. To cut the sheet metal rings for the laminations I'll start my cutting the 8” wide strips into approximate 8” x 8” squares. A pilot hole will then be drilled in the approximate center of each square, doesn't have to be exact and probably stacking a few together at a time. A jig will then be made by inserting a pin slightly smaller than the pilot hole into a hole drill in a piece of plywood. The pre-drilled squares are then stacked on the jig (pin) a few at a time. Clamp your plywood jig on a drill press table and use an end mill cutting tool in place of a drill bit. Adjust your jig to give you the radius needed and clamp the jig to the drill press. Now all you have to do is plunge into the stack with the end mill on your drill press and rotate the stack to cut an accurate circle. Do the same for the inner circle and you end up with ring. Hundreds if needed and all exactly the same.

I don't think you will manage to slide a core on the shaft but you can try. Moving the stator axially does give a degree of control over the characteristic, I don't really like it but that is partly as a result of never running machines off magnetic centre, it doesn't really matter but there is a slight issue that you are bringing the end connections under the influence of the magnets ( probably won't matter)  I would aim for the correct winding ( test coil or similar) but you could do some fine adjustment with axial displacement. I would shift the magnet drum rather than the stator bu spacers between the drum and the hub.

I have to agree, it would be best to match the alternator to the blades right from the start by running a test coil. With my inexperience and lack of knowledge on the subject I wanted to be able to compensate if I missed my target by using something similar as the air gap adjustment on an axial machine. I need to do further reading and study how alternator/blade matching is accomplished before proceeding with coil winding.

I assume these will be smooth punchings with the coils in the air gap so you have no cogging problems.

That is correct, I'm also aiming for .25 inch coil thickness and an air gap no wider than .3125 inch.

With such a small machine you can place the magnet drum over the stator by hand if the stator is securely held and you keep fingers well clear of the danger area. Alternatively you can clamp the magnet drum and lower the stator with a hoist. You could also use long jacking screws through the drum and on to the core but for such a small machine I wouldn't bother, its a slow process and the other methods work ok.

Thanks, knowing this helps a lot. I want to have a healthy respect for the power of neo magnets and not put myself or others in danger by being uneducated on the subject.

Thanks again,
CM

[JW]

There's alot more stuff in the database of old topics than one would expect. I was just messing around with the enhanced search function, thought this was relevent-



Here's the link to actual thread.-
http://www.fieldlines.com/index.php/topic,139891.msg934249.html#msg934249

 1OO% of the old topics are accessable, but about 7.5% of images are not. Yes some have pointed out, there are some missing images, but all the old posts are there...

[oztules]

Not sure the salient pole unit pictured is helpful in this case, but these are.
From two different people:




Give you an idea of what the finished and part built thing will look like.



................oztules

[Flux]

I see you have given this a lot, of thought and you are on the right lines.

Although I admire your idea of producing a core in a way that others can replicate, it probably fails on the basis that they won't have access to your pallet of scrap corer material, so I would just choose the easiest way for your attempt and not deliberately do it the hard way for others to follow. That sheet is fairly tricky to deal with.

Your 14 poles looks fine but it is worth thinking how you are going to fit the coils. Oztules shows two stators both with the same construction with half the coils with bent ends and the other half without bends that fit in the gaps. This is easier to do than a 2 layer diamond coil lap winding. It is in fact derived from the concentric coil designs for large alternators and motors but with only one coil per pole per phase.

To do this arrangement with 14 poles you need a dog leg coil with one bent side and one straight one. My first small machine was 10 pole and the problem arises when the number of poles is not divisible by 4. You can do it ok but the odd coil will make you scratch your head.

Flux

[tecker]

You may want to consider a stator wound with no core segments . Just wind it flat and cast it .ther then would be no cog .

[CraigM]

Thanks everyone,

I took my stator building idea to the guys in the sheet metal shop and they rolled their eyes and had a good laugh muttering something about slicing my fingers off if I attempted such a thing. They pointed to the laser and said if you want to do it right, use the right tool. I was coerced out of another 6 pack of beer and was told they would take care of me. Great bunch of guys but I think they know they have a fish on the line and I'm the one that's hooked. In return I won't schedule to have them work this weekend… a win-win solution for all.

I believe I'll throw the screw adjustable stator positioning idea out the window… at least for this build. A simple bracket welded to the axle tube will give me something to bolt the stator to and it can still be shimmed for fine adjustment. I'm still intrigued by this however. What prompted the idea was a youtube video I saw of a rotating can DC motor (hobby airplane motor) with a similar screw adjustment. A constant voltage was applied to the motor and it ran at a fixed RPM. The stator was then backed out of the rotating can and as it was RPM's increased significantly without a change in voltage to the motor, although I'd assume much less torque. The gears started turning in my head and I envisioned a microprocessor controlled stepper motor that would position the stator fore and aft in relationship to the ideal TSR of a blade set. As wind speed changed the alternator would adjust itself to suit best conditions.

Coil layout, winding fixture and how to adhere the coils to the stator will all require some additional thought. I realize for a newbie as myself this type of build isn't the easiest path to take but with the kind of help I'm finding here on this forum my confidence level is pretty high.

Thank you to all,
CM

[Flux]

Yes you can shim the stator as you suggest for fine adjustment that should be ok. Trying to have it fully adjustable would be a much greater challenge.

The idea of matching the blades this way is a sensible one, but it fails in the same way as all the fixed winding schemes with flux weakening. It keeps the prop loaded correctly but you are stuck with a fixed winding resistance and if you wind for good cut in, you have too much resistance to get reasonable efficiency at full load unless you make it 10 times oversize.

You would certainly gain over the normal scheme in that you would avoid stall and keep the blade power up. By doing it this way you may get up to near 70% electrical efficiency if you don't go for too big a prop but the mechanical complexity of the sliding stator and servo would be a major piece of engineering. Nice to see some new ides though.

Flux

[CraigM]

To do this arrangement with 14 poles you need a dog leg coil with one bent side and one straight one. My first small machine was 10 pole and the problem arises when the number of poles is not divisible by 4. You can do it ok but the odd coil will make you scratch your head.

It took me awhile to understand "the problem arises when the number of poles in not divisible by 4" but when I did the cad work it made sense. In the 14 pole drawing you can see two dog legs next to each other at the 12 oclock position instead of every other coil bent over as in the photo inserted by oztules.

14 Poles

In the 12 pole drawing (divisible by 4) you can see every other coil is bent over.

12 Poles

Stator building is next in line and with 150 or so laminations to cut and assemble it may take awhile.

More questions and learning are surely on the horizon.

Thanks to all,
CM