What is three dimensional magnetic flux?

[BigOutrunner]

Would anyone explain what 'three dimensional magnetic flux' is?

It seems that there have been lots of attempts by big manufacturers (like Honda)
to use compacted 'soft magnetic composites' as a replacement for laminations.

I've attached an image of the prototype motor Honda is testing.
It seems rather inefficient compared to a straight forward winding - that is, unless this 3D effect enhances it somehow.

I'm looking to make a core that's as light as possible, but cannot get my head around how this 3D effect can help.
Any insights to 3D magnetic flux, and comments on Honda's motor design - would be greatly appreciated.

Thanks!

 

[Miles]

When you have a laminated core, the flux is constrained within the laminations and so can only act in a planar direction (2D). For a SMC core there are no restrictions on the direction(3D)

 

[liveforphysics]

This paper goes into that motor a bit.

http://www.ipmd.net/news/002340.html


But it doesn't explain how transverse flux motors work, which this thread covers a bit:
http://endless-sphere.com/forums/viewtopic.php?f=30&t=23109

 

[Miles]

It seems rather inefficient compared to a straight forward winding - that is, unless this 3D effect enhances it somehow.

It's also put to better use at higher frequencies as the eddy current losses are lower than for a standard core (hysteresis losses are higher).

 

[BigOutrunner]

Thank you Gentlemen for the insights.

It seems to me that directed 2D flux is still better. 3D, will just disperse the limited amount of flux produced.
I'd prefer a sniper rifle over a shotgun approach if efficiency is the objective.

Has anybody ever used these SMC prototyping bricks from Hoganas (attached file) as a core?
I have yet to find out how much each disc costs.

The possibility of a non-lam motor is very intriguing.

Their largest disc size is 4.7 inches diameter by .787 inches thick. Machined and glued together, how much better would the motor be over standard lams by anyone's estimation?

Even these compacted soft magnetic composites (SMCs) don't seem to provide better weight against stacked lams. I see claimed weight savings up to 60%, but can't verify that since the weight figures were not posted.

However if an SMC core can be designed to be hollow for lightness and cooling - without seizing or melting down or dropping efficiency significantly, that would be fantastic!

CAN a 3D flux from this Hoganas Somaloy allow for a hollow core design (without major penalites for the finalized motor)?


Thank you all, in advance!

 

[Miles]

I looked into the protoyping blanks, they weren't that expensive but I forget the exact amount.

I'm not sure you're going to save weight, if you mean by that a better power to weight ratio.

Also, if you drop the motor, I should think it's likely to fracture the core....

More useful info in this paper (you probably have all the Somalloy literature):

 

[BigOutrunner]

Thank you Miles! This is document seems to have all of the information I need.
If this thing can support a hollow design, I'll try to order Somalloy blanks and will machine it.
You ROCK man!

 

[Miles]

I'm not sure what you mean by a "hollow design"? It's the cross-sectional area that counts. Air is a waste of space...

 

[BigOutrunner]

Hello Miles,

This is what I mean by hollow:

If a stator tooth is comprised of laminations the size of a box of matchsticks, since it is 2D the laminations are as usual, and heavy. Say we give it 10 windings of copper (this is certain to work), and

If you have another stator tooth comprised of Somaloy (being 3D isotropic), just the same matchbox size but hollow inside, with the same 10 windings of copper - will this 'work'?

That added one dimension - in theory (in my head at least) should allow the hollow Somaloy tooth to work just as well as the standard 2D laminations.

Do you think a hollow-toothed 3D stator comprised of Somaloy would work?

Thanks again!

BigO

 

[Miles]

If you have another stator tooth comprised of Somaloy (being 3D isotropic), just the same matchbox size but hollow inside, with the same 10 windings of copper - will this 'work'?

I don't think so. Because you need the cross-sectional area of the core to be the same (or greater for SMCs) to handle a similar level of flux.

 

[BigOutrunner]

LOL. Thanks Miles! So the 3D isotropy of SMCs doesn't seem to give much of an advantage then.

I think that the words "Rectangular Wave Shaped Coils" in the first image is a misnomer, since those are just Copper Rings sandwiched between separated stator phases - and not real 'coils'.

So the 'copper savings' and 'reduction in size' being touted by SMC suppliers and researchers is that there are none of the usual 'windings' around the stator teeth. Now I think I understand.

It really seems that the prototype 3D motor Honda is making would be far less powerful than a standard stator with laminations and real copper windings per tooth.
And as you pointed out, the usual lams would be far more robust than SMCs.

Your advice, comments and suggestions (everyone) is invaluable and much appreciated!

Love this forum! Thanks!

BigO

 

[Miles]

LOL. Thanks Miles! So the 3D isotropy of SMCs doesn't seem to give much of an advantage then.

None at all with a conventional motor.

You can extend the heads of the core teeth out over the endturns of the coils but the bottleneck is the section running through the centre of the coil.

 

[BigOutrunner]

Did anyone notice in the Honda prototype motor drawing that there are only two copper rings (not 'coils' in a real sense) a U phase and a W phase copper ring, but no V phase copper ring?

Can anyone take a stab at how the V phase 3 dimensional stator segment will be 'powered up' (if at all) without that missing third V phase 'coil'/copper ring?

Thanks everyone! Good night.

 

[major]

Did anyone notice in the Honda prototype motor drawing that there are only two copper rings (not 'coils' in a real sense) a U phase and a W phase copper ring, but no V phase copper ring?

Can anyone take a stab at how the V phase 3 dimensional stator segment will be 'powered up' (if at all) without that missing third V phase 'coil'/copper ring?

I noticed that also. I guess maybe they excite the 2 coils as 2 phase 120 degrees apart in the electrical timing. So U and W phases are excited and V phase is a consequence. Similar maybe to DC machines which use a consequence pole (like a 2 pole motor with only one field coil). I've never seen it done with armatures before, so just guessing here. But using the fact that balanced 3 phase has U + W + V = 0, you can derive V from the sum of - (U + W). So can the sum of the U and W mmf's produce the flux in the V stator ring? I think so.

 

[drebikes]

Did anyone notice in the Honda prototype motor drawing that there are only two copper rings (not 'coils' in a real sense) a U phase and a W phase copper ring, but no V phase copper ring?

Can anyone take a stab at how the V phase 3 dimensional stator segment will be 'powered up' (if at all) without that missing third V phase 'coil'/copper ring?

I noticed that also. I guess maybe they excite the 2 coils as 2 phase 120 degrees apart in the electrical timing. So U and W phases are excited and V phase is a consequence. Similar maybe to DC machines which use a consequence pole (like a 2 pole motor with only one field coil). I've never seen it done with armatures before, so just guessing here. But using the fact that balanced 3 phase has U + W + V = 0, you can derive V from the sum of - (U + W). So can the sum of the U and W mmf's produce the flux in the V stator ring? I think so.

I think so too, as long as the command is aware

Off-topic maybe, but in regards to various materials that would diminish weight of the motor, why not iron-cobalt or iron-nickel alloys? FeCo has a saturation point far higher than silicon steel, while FeNi has the same saturation point while having lower losses - one would decrease weight for the same losses, one would simply reduce losses.

 

[athlon]

Off-topic maybe, but in regards to various materials that would diminish weight of the motor, why not iron-cobalt or iron-nickel alloys? FeCo has a saturation point far higher than silicon steel, while FeNi has the same saturation point while having lower losses - one would decrease weight for the same losses, one would simply reduce losses.

a good lamination material not only have an high saturation but also need a high resistivity to reduce eddy current , silicon steel have this very high resistivity so is a good material for iron lamination.


Goin back on topic time ago I was dreaming about a finely powdered mu-metal , oxidized on suface for insultion , all keept together by resin ... mu-metal have a very high permeability , pulverization and thin ( few atoms ) insulation layer avoid the eddy current

 

[BigOutrunner]

Did anyone notice in the Honda prototype motor drawing that there are only two copper rings (not 'coils' in a real sense) a U phase and a W phase copper ring, but no V phase copper ring?

Can anyone take a stab at how the V phase 3 dimensional stator segment will be 'powered up' (if at all) without that missing third V phase 'coil'/copper ring?

I noticed that also. I guess maybe they excite the 2 coils as 2 phase 120 degrees apart in the electrical timing. So U and W phases are excited and V phase is a consequence. Similar maybe to DC machines which use a consequence pole (like a 2 pole motor with only one field coil). I've never seen it done with armatures before, so just guessing here. But using the fact that balanced 3 phase has U + W + V = 0, you can derive V from the sum of - (U + W). So can the sum of the U and W mmf's produce the flux in the V stator ring? I think so.

Honda is admitting that the max efficiency they are deriving from the prototype motor is only 71%.
I'm not surprised, since the flux is partially being wasted in 1 extra direction - since it is 3D.
So the flux isn't just moving towards the magnets below the segmented stator in the drawing - it is also flowing sideways, supposedly.

I'm not sure if exciting the copper '120 degrees apart' is possible since these are copper rings and not the usual windings around each teeth.
Which should make the back emf sinusoidal - but I'm guessing here.

Since both U and W phase copper rings are in contact with the V phase stator (which is 3D), the V phase stator is always half active if either copper ring is on.
Thus you are right, if both U and W copper rings the V phase stator would then be fully active.

But then, wouldn't all of the stator phases be active at the same time as well?

Perhaps the sneakier way that the V phase stator is activated is by exciting half of each of the copper rings that are facing or directly touching the V phase stator instead? So the U copper ring is actually half U and half V, powered at the same time, and same goes for the W copper ring.

If you look at the drawing, only 75% of the U phase stator tooth is in direct contact with the U phase 'coil'. The copper isn't wound completely around the tooth.
Because the tooth isn't wrapped 100% like a coil, 25% of the potential to max flux is lost.

Then add say 4% of flux lost to its 3 dimensional nature, so the 71% efficiency Honda is seeing is just about right.

All of this is just wild guess of course.

Our resident motor guru friends would have a better picture, for sure.

As Miles said, as a direct replacement for laminations in motors we normally work on, using 3D somaloy core is very inefficient.

But is there any practical, basic configuration where 3D somaloy would be useful, or better than laminations in electric motors? Even a single simple example may broaden our understanding of this new 3D motor technology.

Honda's prototype does not inspire confidence that this 3D thing is actually better than standard laminations.

Shall we label 3D just marketing hype and call it a day?

Thanks All!

 

[Miles]

But is there any practical, basic configuration where 3D somaloy would be useful, or better than laminations in electric motors? Even a single simple example may broaden our understanding of this new 3D motor technology.

The YASA motor is an example of a high performance motor that uses SMC cores. Mostly for practical reasons. It doesn't really take advantage of the isotropic properties. http://www.yasamotors.com/
http://www.soue.org.uk/souenews/issue9/yasa.html

Hobby King were trumpeting how they were going to make a super efficient motor using SMC.
After testing, the project was quietly dropped..... http://www.hobbyking.com/hobbycity/store/twit.asp?id=120 LOL!

 

[BigOutrunner]

But is there any practical, basic configuration where 3D somaloy would be useful, or better than laminations in electric motors? Even a single simple example may broaden our understanding of this new 3D motor technology.

The YASA motor is an example of a high performance motor that uses SMC cores. Mostly for practical reasons. It doesn't really take advantage of the isotropic properties. http://www.yasamotors.com/
http://www.soue.org.uk/souenews/issue9/yasa.html

Hobby King were trumpeting how they were going to make a super efficient motor using SMC.
After testing, the project was quietly dropped..... http://www.hobbyking.com/hobbycity/store/twit.asp?id=120 LOL!

Thanks again Miles.
It may be of interest to eBikers, that Hoganas has produced one specifically with their Somaloy, saying its way more efficient than other motors.
Here it is.