DIY Toroidal Axial Flux PM

Electric Motors and Controllers
User avatar
HalbachHero   100 W

100 W
Posts: 209
Joined: Apr 05 2021 8:35pm

Re: DIY Toroidal Axial Flux PM

Post by HalbachHero » Dec 18 2021 7:36pm

Jrbe wrote:
Dec 18 2021 6:56am
Great job optimizing all of the parts, these look excellent!

Do you have metal collars for the bolts between the 2 back irons? Just having plastic there is likely adding to the flexibility. Ideally the bolt hole collars would hold the shape of the 2 back irons with just the irons, bolts, and spacers bolted together. You can experiment without the plastic parts in the metal stack then mod the plastic parts to fit or reprint if necessary.

You could likely print an impeller that bolts onto the side.
Thanks!
No metal collars, but it sounds like it could be helpful. They would have to be pretty small diameter and tall. Not sure the best way to make something like that.... I've thought about breaking the spacer down into a bunch of layers that could be bonded together, they could provide airflow but making it possible to laser cut. I might be able to achieve tighter tolerances when it comes to height, and the laser cut stuff is way cheaper then CNC stuff.

I have been thinking of a way to incorporate an impeller still, but it really adds to design complexity, so I'm not sure how I will proceed with that.

I have made a silicone mold, but the printed shell is still in the works. I should be able to pot some nylon test stators soon, and I have a few other things I want to try now that I have the silicone. But right now the goal is to make a stator with epoxy that is stiff, but also has very little variability in thickness throughout.

I have assembled the new rotors too. Those are curing are the moment.

Another thought I had, was that with all these turns of copper, its a significant total length. I wonder if it would make sense to being the magnets as close together as possible instead of following the "rule of thumb" for the air gap. I imagine if the rotors can handle the stresses, it might be worth using less copper and maximizing the magnetic field. I would think this would only apply for air cores too... Thoughts anyone?

Jrbe   100 W

100 W
Posts: 113
Joined: Mar 23 2020 6:28pm
Location: CT USA

Re: DIY Toroidal Axial Flux PM

Post by Jrbe » Dec 19 2021 7:14am

Shop for the collars at McMaster, MSC, or similar. They typically have cad for their stuff and it's usually accurate for simple things like this. Not cost effective or worth the headache to try to make them unless you buy tube cut to length. I put in some filters I thought would help narrow down the selection. They're on the left side. Just check to make sure they make sense for you. I'd guess you'd want the split collars (but not the alignment collars with thinned out ends.)
https://www.mcmaster.com/spacing-collar ... l~ceramic/

The impeller could be open https://www.idolz.com/wp-content/uploads/2021/06/2.png and use the side of the motor to be the other side of the impeller. Should make it easier to print.
You could extend a second set of stand offs and use the same outer bolts to hold the impeller in place and make it optional / replaceable.

There is a process called "forged carbon fiber". It's a ridiculous name but mixes up resin and short carbon fibers in a slurry then is put in a pressed mold to cure. I thought about your project when I saw it. There are lots of bad youtube videos on it, some showing their failures. Most of the failures were from flexible molds, not having enough reinforcement / resin slurry and left voids, or flexible molds that distorted and left thin and thick spots. You could likely use fiberglass, Kevlar, carbon, a mix, or maybe even sheets along with the slurry. You can buy pulp too (really short pieces of reinforcement.) Just follow the don't breathe the dust recommendations if you sand carbon.
https://youtu.be/25PmqM24HEk

The aluminum plate he used on the back of his molds is why I don't think silicone will work well for molding this. The mold resin could make the press easy to make or it could be 3d printed. I think you will need some little bumps or some geometric circular pattern to hold the stator centered and allow for a void for the slurry to flow evenly and allow for an end result flat face. I don't think the dimples / pattern left in the stator faces from them would have any negative effect as long as the edges are radiused. And thinking about the copper probably touching I'm guessing you'll need at least 1 layer top and bottom of fiberglass to make sure there's an insulating material protecting the stator.

(Edit - new)
You could add a viewing window to see the stator gap and very slowly spin the motor to watch what the stator does. If your coils and magnets are even and the stator and hub are rigid and the rotor balanced well you could probably minimize the gap.
You could also add thin paint lines on the magnets and or the stator to see where they rub.

User avatar
HalbachHero   100 W

100 W
Posts: 209
Joined: Apr 05 2021 8:35pm

Re: DIY Toroidal Axial Flux PM

Post by HalbachHero » Dec 19 2021 2:22pm

Jrbe wrote:
Dec 19 2021 7:14am
Shop for the collars at McMaster, MSC, or similar. They typically have cad for their stuff and it's usually accurate for simple things like this. Not cost effective or worth the headache to try to make them unless you buy tube cut to length. I put in some filters I thought would help narrow down the selection. They're on the left side. Just check to make sure they make sense for you. I'd guess you'd want the split collars (but not the alignment collars with thinned out ends.)
https://www.mcmaster.com/spacing-collar ... l~ceramic/

The impeller could be open https://www.idolz.com/wp-content/uploads/2021/06/2.png and use the side of the motor to be the other side of the impeller. Should make it easier to print.
You could extend a second set of stand offs and use the same outer bolts to hold the impeller in place and make it optional / replaceable.

There is a process called "forged carbon fiber". It's a ridiculous name but mixes up resin and short carbon fibers in a slurry then is put in a pressed mold to cure. I thought about your project when I saw it. There are lots of bad youtube videos on it, some showing their failures. Most of the failures were from flexible molds, not having enough reinforcement / resin slurry and left voids, or flexible molds that distorted and left thin and thick spots. You could likely use fiberglass, Kevlar, carbon, a mix, or maybe even sheets along with the slurry. You can buy pulp too (really short pieces of reinforcement.) Just follow the don't breathe the dust recommendations if you sand carbon.
https://youtu.be/25PmqM24HEk

The aluminum plate he used on the back of his molds is why I don't think silicone will work well for molding this. The mold resin could make the press easy to make or it could be 3d printed. I think you will need some little bumps or some geometric circular pattern to hold the stator centered and allow for a void for the slurry to flow evenly and allow for an end result flat face. I don't think the dimples / pattern left in the stator faces from them would have any negative effect as long as the edges are radiused. And thinking about the copper probably touching I'm guessing you'll need at least 1 layer top and bottom of fiberglass to make sure there's an insulating material protecting the stator.

(Edit - new)
You could add a viewing window to see the stator gap and very slowly spin the motor to watch what the stator does. If your coils and magnets are even and the stator and hub are rigid and the rotor balanced well you could probably minimize the gap.
You could also add thin paint lines on the magnets and or the stator to see where they rub.
All good ideas. Thanks. I have actually already purchased some chopped fiberglass to achieve a similar effect to what you are talking about with the CF. Others have mentioned that its possible that CF in the stator could effect performance in a negative way. Not really sure. I am hoping I wont need a textured surface, as the wires themselves are strands, so I'm hoping the resin will still find its way through things, but again, I'm unsure.

I did intend to use fiberglass sheets on top bottom, and even the OD in a rig, since that seems to be the most exposed on the past stators. I also intend to incorporate fiberglass tow into the litz to help with things.

The viewing window might be built in with the holes I have made on the spacer for the rotors, not sure yet, but that's a good idea.



So the shell of the press is printed, and it feels solid, but I may have underdone it again, I might have to make it thicker with more infill (I used 50%). There is a silicone insert in there so it releases after pressing, but there are a few things I still don't quite know, and will take a bit of trial and error to nail down.
I made 3mm inner and outer diameter walls with the expectation that it will compress a bit when I squish everything in there. So there's a fudge factor in there to make sure the expanded stator is not too big. I'm sure this will take a few tries. I also added some countersunk holes for M3 screws to mount the stator to the hub. This will allow for swapping out hubs to get tolerances nailed down, or to be able to swap stators. (Trying to make it more modular). I have concerns I made the spots for the screws too close to the ID of where the copper will be. So I may need to adjust things there too.

I will be testing this first with the 6 turn nylon stator first. While the diameter of the nylon thread is similar in total to the litz wire, it acts differently. With the litz wire the strands being untwisted, they are able to mush together a bit more, so the 6 turn nylon stator is a bit taller than the equivalent copper stator of the same number of turns. I might make a 8 turn nylon stator to play with, but that might be as far as I need to go with that. Ill probably have to make several of those. At least I'm getting quicker at it.
resized-image-Promo - 2021-12-19T135647.460.jpeg
resized-image-Promo - 2021-12-19T135647.460.jpeg (105.98 KiB) Viewed 1063 times

On the rotor font. I got them all glued up. These ones look pretty cool, cuz its basically all magnets and steel now. Together they weigh ~275g. I bought some brass machine screws for the inner diameter since the steel ones are obviously ferrous and I'm not sure how that will affect the flux lines.

The nice thing with these rotors is that I have no plastic backing, so I can see if there is flux leakage directly on the steel. I cant get anything to stick. If I put the rotors back to back perfectly, there is nearly nothing, but slightly offset and there is, so I'm not sure if that's flux getting around the metal, and actually pulling from the front or what, but its very very weak. So I think I'm good with these ones. Also, with the new designs they are much harder to get apart since I cant pull from the side of the rotor to separate them. So ill have to find an easy way to work with that, especially if I'm trying to get shrink the air gap.
resized-image-Promo - 2021-12-19T135626.178.jpeg
resized-image-Promo - 2021-12-19T135626.178.jpeg (113.66 KiB) Viewed 1063 times
resized-image-Promo - 2021-12-19T135608.110.jpeg
resized-image-Promo - 2021-12-19T135608.110.jpeg (67.16 KiB) Viewed 1063 times

User avatar
HalbachHero   100 W

100 W
Posts: 209
Joined: Apr 05 2021 8:35pm

Re: DIY Toroidal Axial Flux PM

Post by HalbachHero » Dec 21 2021 9:43pm

Okay, first test. I used the 6 turn nylon stator, which stood slightly above the wall of the jig, so I had to squish it down a little bit. The flat surface for the lid was a bad idea. which I'm realizing now I never took a photo of. But basically there is no compression happening. Also I did not use enough epoxy. So nailing down that amount will be good. I have to assume though that the nylon will soak it up a bit better than enameled wire.

The resulting "stator" was about 8.5mm thick, which interestingly enough was the 6.5mm height of the wall +the 2mm thickness of the silicone. So that completely bottomed out?? I have modified the mold to have an extension to compress the stator and compensated for the silicone completely giving out its thickness on the bottom.

anyway, here's your pic
resized-image-Promo - 2021-12-21T213145.833.jpeg
resized-image-Promo - 2021-12-21T213145.833.jpeg (111.46 KiB) Viewed 1008 times

Oh also,
Jrbe wrote:
Dec 19 2021 7:14am
There is a process called "forged carbon fiber". It's a ridiculous name but mixes up resin and short carbon fibers in a slurry then is put in a pressed mold to cure. I thought about your project when I saw it. There are lots of bad youtube videos on it, some showing their failures. Most of the failures were from flexible molds, not having enough reinforcement / resin slurry and left voids, or flexible molds that distorted and left thin and thick spots. You could likely use fiberglass, Kevlar, carbon, a mix, or maybe even sheets along with the slurry. You can buy pulp too (really short pieces of reinforcement.) Just follow the don't breathe the dust recommendations if you sand carbon.
https://youtu.be/25PmqM24HEk
I have re-watched this video a couple times. which is funny because I remember seeing it pop up on my phone and disregarding it. I really want to try again without the silicone, and actually using a mold release. I have tried cooking spray and silicone lubricant to no avail. I'm really tempted to buy their mold release and do all the things he mentions. PETG, sanding it, chamfers to insert something to pop it loose, make it a 3 part mold. I think that would help with getting tolerances right.
However, I plan on trying some weird things with the silicone which I do not think would be possible with just the 3d printed stuff. but I guess we will see.

Jrbe   100 W

100 W
Posts: 113
Joined: Mar 23 2020 6:28pm
Location: CT USA

Re: DIY Toroidal Axial Flux PM

Post by Jrbe » Dec 22 2021 8:33am

Some ideas for you.

There are vapor smoothing and other techniques you could try to smooth out your mold.

6+ layers of pva mold release should release any smooth mold. You can apply it with a fine mist spray bottle, something like a "preval sprayer", a paint gun, etc. It's like spraying clear coat. Some have a color added to help you see where you sprayed it. Your mold color ideally would contrast the pva so you can easily see where it is. Likely not worth reprinting just for contrast though.

Do you have any draft in the mold? There's a chance the epoxy shrinkage is enough to not get stuck.
I think a cylinder would do better for compressing this. Could add a lip or some feet on the id and od to bottom out on your ideal height.

There are air bubbles in the epoxy. Do you have a way to vacuum pot the resin before using it? Voids could do unexpected things. I'd guess they might cause hot spots.

Do you have any risers to allow air and extra resin to escape the mold? You could drill a bunch of small holes in the outer edge. These might get hardened resin stuck in them though. A small tapered drill bit could work well. I can't remember if I shared this link, https://m.youtube.com/watch?v=PyoBOBDkouE he's making copies of parts in urethane with silicone molds. He explains the process well. A hybrid 3D printed mold with silicone inserts in the non critical / riser / sprue areas seems like it could work well too.

User avatar
HalbachHero   100 W

100 W
Posts: 209
Joined: Apr 05 2021 8:35pm

Re: DIY Toroidal Axial Flux PM

Post by HalbachHero » Dec 22 2021 12:42pm

Jrbe wrote:
Dec 22 2021 8:33am
Some ideas for you.

There are vapor smoothing and other techniques you could try to smooth out your mold.

6+ layers of pva mold release should release any smooth mold. You can apply it with a fine mist spray bottle, something like a "preval sprayer", a paint gun, etc. It's like spraying clear coat. Some have a color added to help you see where you sprayed it. Your mold color ideally would contrast the pva so you can easily see where it is. Likely not worth reprinting just for contrast though.

Do you have any draft in the mold? There's a chance the epoxy shrinkage is enough to not get stuck.
I think a cylinder would do better for compressing this. Could add a lip or some feet on the id and od to bottom out on your ideal height.

There are air bubbles in the epoxy. Do you have a way to vacuum pot the resin before using it? Voids could do unexpected things. I'd guess they might cause hot spots.

Do you have any risers to allow air and extra resin to escape the mold? You could drill a bunch of small holes in the outer edge. These might get hardened resin stuck in them though. A small tapered drill bit could work well. I can't remember if I shared this link, https://m.youtube.com/watch?v=PyoBOBDkouE he's making copies of parts in urethane with silicone molds. He explains the process well. A hybrid 3D printed mold with silicone inserts in the non critical / riser / sprue areas seems like it could work well too.
Thanks again, I really appreciate the advice and search terms. Its really been helpful to find what I may need. I'm still not familiar with some of the jargon though. Such as a draft. Do you mean an angle build into the wall of the mold for easy release? Also I did add some things into the mold and modified it to not use silicone at all, and we will just try with the mold release and dowels for pushing it out.

I'm not totally familiar with the vacuum stuff yet, but I'm interested. That's probably something I will try if I cannot get rid of air bubbles on my own.

I didn't want to drill into the print because of the voids in the infill, so I added them into the model. I have not seen that link before either, so thanks again for sharing that too. Very good stuff.
resized-image-Promo - 2021-12-22T123242.744.jpeg
resized-image-Promo - 2021-12-22T123242.744.jpeg (114.63 KiB) Viewed 980 times
resized-image-Promo - 2021-12-22T123321.596.jpeg
resized-image-Promo - 2021-12-22T123321.596.jpeg (137.92 KiB) Viewed 980 times
I printed these parts, very happy with the rigidity of things, and the quality, but then I changed the design again. So Ill have to reprint, but I will try pressing a nylon stator with the one I just made and see if that extension helps, and how that plays with the silicone around the outer wall.

Here's the new model, haven't printed yet. With the infill at 75% they each take quite a bit of time and filament.
Screenshot 2021-12-22 123025.png
Screenshot 2021-12-22 123025.png (365.23 KiB) Viewed 980 times
Screenshot 2021-12-22 123014.png
Screenshot 2021-12-22 123014.png (310.81 KiB) Viewed 980 times

Jrbe   100 W

100 W
Posts: 113
Joined: Mar 23 2020 6:28pm
Location: CT USA

Re: DIY Toroidal Axial Flux PM

Post by Jrbe » Dec 22 2021 1:57pm

No problem. I really like the work you've done so far and I think this has a lot of potential. Also please feel free to ignore anything I suggest.

Yeah, draft is an angle on the part so it can pop out of the mold. 1 to 5 degrees per inch of length is typical depending on materials & method. Your mold might be flexible enough and stator short enough to not have to worry about draft.

Risers are typical in casting / molding to allow the air out of the edges. If your holes around the outside edge don't have draft the resin that goes up the riser will likely get stuck.
Or if they are for dowels or pins to help pop the stator out these are normally called ejector pins in injection molding. Those would help get it apart.
You could consider carefully using very low air pressure to pop it apart too. Could push wax, silicone, or some kind of plug in the air holes before assembling the mold so they don't fill up.

If you do decide to get a vacuum pot look for one that can be pressurized too. Vacuum helps get the air out, pressure helps force the resin in and minimize any leftover air pockets / voids. I think a vacuum / pressure pot would be very helpful for what you're trying to do and a big step forward in quality once you work the bugs out.

User avatar
HalbachHero   100 W

100 W
Posts: 209
Joined: Apr 05 2021 8:35pm

Re: DIY Toroidal Axial Flux PM

Post by HalbachHero » Dec 22 2021 9:45pm

Jrbe wrote:
Dec 22 2021 1:57pm
No problem. I really like the work you've done so far and I think this has a lot of potential. Also please feel free to ignore anything I suggest.

Yeah, draft is an angle on the part so it can pop out of the mold. 1 to 5 degrees per inch of length is typical depending on materials & method. Your mold might be flexible enough and stator short enough to not have to worry about draft.

Risers are typical in casting / molding to allow the air out of the edges. If your holes around the outside edge don't have draft the resin that goes up the riser will likely get stuck.
Or if they are for dowels or pins to help pop the stator out these are normally called ejector pins in injection molding. Those would help get it apart.
You could consider carefully using very low air pressure to pop it apart too. Could push wax, silicone, or some kind of plug in the air holes before assembling the mold so they don't fill up.

If you do decide to get a vacuum pot look for one that can be pressurized too. Vacuum helps get the air out, pressure helps force the resin in and minimize any leftover air pockets / voids. I think a vacuum / pressure pot would be very helpful for what you're trying to do and a big step forward in quality once you work the bugs out.
Thanks. I appreciate your, and others interest. I'm definitely enjoying the journey

I added in the chamfered edges to insert a wedge so I'm hoping I can avoid using drafts, but I can certainly modify it if I need to. I made the risers on the lid have a draft now. Does it make sense for the small opening to be closest to the part being pressed? I was going to use wax or clay ( I have clay readily available) to plug the bottom holes, and use a dowel to knock out the part after. I will be getting some mold release and wax, and will spend the time to prep the surface better. Id love to be able to reuse a printed mold.

I will do my "research" on the vacuum/pressure pots. But would I essentially just press the part in the mold, then put it in the vacuum/pressure pot?

Jrbe   100 W

100 W
Posts: 113
Joined: Mar 23 2020 6:28pm
Location: CT USA

Re: DIY Toroidal Axial Flux PM

Post by Jrbe » Dec 23 2021 10:05am

The small end of the draft of the riser tubes should ideally be at the top / away from the part. This should allow the pieces to slide apart. You can flip it so the small end is closest to the part if you think you can snap them all off when separating the mold by twisting or separating. Still need to figure out how to get the hardened resin out of the risers if you go this route.

You could also swap over to just straight riser holes and push / force some silicone tubes into them. This way the worst that happens should be a tube pops out but doesn't get stuck. Extra resin can flow up the inside of the tube. It will be important to cut the inside end of the silicone hose straight where it touches the stator, the end shape will end up in the stator. Silicone tubes might get pushed out by the clamping pressure though. Some ribs or teeth inside the risers could help hold the silicone tubes in place. Could do the same for the fill tube in a larger diameter. Having this fill tube could allow a funnel to be swapped for an air syringe to push resin in. The ends of the silicone tubes could distort and leave odd bumps in the stator if they move.
Looking at your mold I'm not sure how you're filling it.

Not sure what your press process looks like now. There are a few ways to do this. If you could build the press into the mold that should keep it squeezed while it cures in the pressure pot. You could also go the vacuum bag route instead and likely use your same press method.

Vacuum helps expand the bubbles in the resin so they pop / go away. Likely won't be able to get all the bubbles out. Vacuum helps remove the bubbles and voids when in the mold.

Pressure forces any left over air or voids to shrink drastically.

- Mix up your resin. You can try to calculate how much but you will want some extra.
- Vacuum air bubbles out of the resin in its mixing cup. Make sure your container is only 1/3 to 1/2 full so the resin won't "boil over." A large cup inside your vacuum chamber will catch spill over. The resin might not be able to go to full vacuum. If it turns into a gas you could ruin your vacuum pump. You can try to mix without adding air bubbles but it typically doesn't work.
- Load up the stator, fiberglass, etc into the mold and set it so it's sealed (partially clamped?) Pour the resin into the mold like you pour a beer, constant stream of resin on a mostly vertical wall. This will help not add extra air bubbles. As the voids are filled more resin will flow into the mold. There will be a learning curve here, how much to fill and how to fit extra resin capacity to be able to flow in.
- clamp it and watch for overflow. Top it off if necessary / wipe overflow.
- Load it into the pressure pot, vacuum bag, or go gravity fed. Splits now.

Gravity fill - keep it mostly full and watch the riser tubes. Don't overfill. Hard to avoid voids. Can somehow force resin in to try to make sure air is out. Will likely have to wipe the riser tubes as you do this.

Vacuum bag - vacuum it and put it in the press. Probably worth taking it out after getting to half vacuum level to check the resin level. Make sure there's enough resin in there. You can likely open it and add resin if needed. Keep it under vacuum until it's cured. The bag will suck into the fill spout as the air goes out, a small chamfer or radius could help it not pop the bag here.
There are options down the vacuum infusion route you could consider. Using vacuum to suck the air out then at full vacuum opening a valve to suck the resin in. This uses atmospheric pressure (14.7 psi) to fill the stator / mold. You can then put vacuum back on after filling but watch that the vacuum doesn't suck up resin (happens quick!) This would help clamp the mold with atmospheric pressure. This can be a pain to setup but can work really well. Has a steep learning curve and a bunch of variables to watch.
A resin trap is a good idea any time you vacuum resin. Can be a mason jar with 2 tubes in the lid.

Pressure pot - is a lot like the gravity fill but the pressure will shrink any bubbles / voids. Keep an eye on the resin level. Keep the pressure on until it's cured.
The potential downside here is tiny bubbles at 45- 60?psi. Probably won't do much when they get hot until it gets to the resin heat deflection temp but it's hard to know for sure.

For clamping the mold you could:
- add in bolt holes inside and out of the mold
- use a v band clamp outside and bolts inside
- go the vacuum bag or gravity fill route and clamp it your normal way.

Consider waxing the outside of the mold too. Should help it stay clean and maybe not stuck together when it gets messy.

User avatar
APL   100 kW

100 kW
Posts: 1062
Joined: Aug 06 2018 8:44pm

Re: DIY Toroidal Axial Flux PM

Post by APL » Dec 24 2021 3:18pm

Great progress! Working with resin has it's learning curve for sure, I'm no expert, but have done a lot of it. Keep it simple, shouldn't be very hard for a small part.

I've used some cling wrap for some things, resin doesn't stick to it, but it might wrinkle slightly. Clear packing tape doesn't stick either. Fiberglass polyester resin is cheap for simple test's if your epoxy is costly.

Came across this photo of the Marand motor the other day,.. looks like they use a thin fiberglass board in the center, and just catch the edge of the end windings. Might be something to try, add strength, and could drill the holes after the cure? Outer edge looks like pure resin.
Screenshot 2021-12-24 140000.png
Screenshot 2021-12-24 140000.png (681.87 KiB) Viewed 913 times
(They relieved the back iron in the magnet centers to save weight.)

Anyway, keep up the good work, looking forward to more progress!

User avatar
HalbachHero   100 W

100 W
Posts: 209
Joined: Apr 05 2021 8:35pm

Re: DIY Toroidal Axial Flux PM

Post by HalbachHero » Dec 31 2021 3:36pm

Hope everyone is enjoying the holidays. I've been busy with other things lately. So not much progress, but wanted to catch up.
Jrbe wrote:
Dec 23 2021 10:05am
The small end of the draft of the riser tubes should ideally be at the top / away from the part. This should allow the pieces to slide apart. You can flip it so the small end is closest to the part if you think you can snap them all off when separating the mold by twisting or separating. Still need to figure out how to get the hardened resin out of the risers if you go this route.
For this reason I moved the small end of the risers to the outside edge of the mold.
Jrbe wrote:
Dec 23 2021 10:05am
You could also swap over to just straight riser holes and push / force some silicone tubes into them. This way the worst that happens should be a tube pops out but doesn't get stuck. Extra resin can flow up the inside of the tube. It will be important to cut the inside end of the silicone hose straight where it touches the stator, the end shape will end up in the stator. Silicone tubes might get pushed out by the clamping pressure though. Some ribs or teeth inside the risers could help hold the silicone tubes in place. Could do the same for the fill tube in a larger diameter. Having this fill tube could allow a funnel to be swapped for an air syringe to push resin in. The ends of the silicone tubes could distort and leave odd bumps in the stator if they move.
Looking at your mold I'm not sure how you're filling it.
Ill see if I have issues with the drafted risers. If I do, I will try this method instead. Currently I am filling the mold by pouring the epoxy directly into it, the placing the top on an compressing it.
Jrbe wrote:
Dec 23 2021 10:05am
Consider waxing the outside of the mold too. Should help it stay clean and maybe not stuck together when it gets messy.
I did buy some wax too I will definitely use this in conjunction with the mold release.

Thanks for all the information on the resin stuff. It sounds like this could be an adventure. I am going to attempt making some nylon stators using pour and press method that I have done so far. This should allow me to at least test the mold dimensions and the mold release stuff, along with other things, like the using clay to plug holes so I can pop it out with a dowel after.
I will in the mean time convince my wife to let me spend even more money on this project, and Ill see if I can get a vacuum chamber and make a resin trap.
I will start with the gravity method, and if this causes issues, I will attempt to make the vacuum infusion work, but I have concerns with putting it in the bag then putting that in the press. I worry that the shape will be somewhat compromised, but who knows, this might work well.

I'm trying to iterate on the complexity as the original idea here was simplicity. And its getting away from that a bit by involving a vacuum chamber/ vacuum infusion stuff, but that's not going to stop the project.



APL wrote:
Dec 24 2021 3:18pm
I've used some cling wrap for some things, resin doesn't stick to it, but it might wrinkle slightly. Clear packing tape doesn't stick either. Fiberglass polyester resin is cheap for simple test's if your epoxy is costly.
Thanks for the suggestions, that could be an easy way to make the mold non stick, as there are no really complex geometries in the mold itself. I bought some cheap table top epoxy on amazon. Its got a really long cure time, and is not something I will ultimately use on the real stator, but it was cheap for a good amount. Ill look at the fiberglass polyester stuff next though.
APL wrote:
Dec 24 2021 3:18pm
Came across this photo of the Marand motor the other day,.. looks like they use a thin fiberglass board in the center, and just catch the edge of the end windings. Might be something to try, add strength, and could drill the holes after the cure? Outer edge looks like pure resin.
Thanks for sharing that one. I'm not sure I saw that from Marand before, but since I moved away from the 3d printed core, my motor is very similar to theirs.
Unfortunately, since mine is an out runner, I cant/don't want to do anything on the outer diameter, but that same concept could be applied to the inner, currently I am planning on making a puck basically, and use some machine screws to fasten the stator to the hub. I'm sure there is room for optimization here, that might come with future generations, where I could make a disk with mounting points on the inner diameter that are smaller, and would result in needing less epoxy.
Also I redid the calculations on the weight savings of the back iron reliefs. I just don't see the point. At least not for my case. Theirs is a much larger diameter, so it might be a more significant weight savings, but it would only save ~20g in my case, which is less than 5% of the total weight. Obviously Marand's goals outweigh the complexity of machining/design, but mine does not unfortunately.



going back to the second silicone mold I had made. I tried to make a stator with it. and it sort of worked. but I dont like how inconsistently the silicone would compress. Its adding too much variation to the back face. I might revisit using this in the inner/outer diameter, but for now I am going for a purely 3D printed mold.

Here's some pics
resized-image-Promo - 2021-12-25T093701.468.jpeg
resized-image-Promo - 2021-12-25T093701.468.jpeg (117.33 KiB) Viewed 804 times
resized-image-Promo - 2021-12-25T093721.070.jpeg
resized-image-Promo - 2021-12-25T093721.070.jpeg (88.07 KiB) Viewed 804 times


and here's the new mold
resized-image-Promo - 2021-12-25T093824.490.jpeg
resized-image-Promo - 2021-12-25T093824.490.jpeg (74.58 KiB) Viewed 804 times

User avatar
HalbachHero   100 W

100 W
Posts: 209
Joined: Apr 05 2021 8:35pm

Re: DIY Toroidal Axial Flux PM

Post by HalbachHero » Jan 04 2022 1:17am

More progress with the stator tests.

I sanded down the mold with two different grits of sandpaper, and tried to be thorough. Then plugged the holes up with clay and buffed it with parting wax.
I poured the mold release into a spray bottle. Not a paint sprayer, but just a little squirt bottle that window cleaner comes in. I realize this is not ideal but thought it might suffice. I gave it 3 or so "coats", but the spray was not even, and the box I had it in was not level. This caused the liquid to pool. I used a paint brush to smooth out the liquid, but might not have been the best idea either that seemed to take a lot off in the process. So I have my doubts there is enough there to actually prevent it from sticking, but we will see...
resized-image-Promo - 2022-01-04T010723.223.jpeg
resized-image-Promo - 2022-01-04T010723.223.jpeg (140.81 KiB) Viewed 753 times
resized-image-Promo - 2022-01-04T010748.242.jpeg
resized-image-Promo - 2022-01-04T010748.242.jpeg (115.99 KiB) Viewed 753 times
I cut a piece of plywood and bored a hole in it that was slightly larger diameter than the inner set of risers. The outer ones were covered. Again, we will see how this goes.
resized-image-Promo - 2022-01-04T010806.634.jpeg
resized-image-Promo - 2022-01-04T010806.634.jpeg (92.98 KiB) Viewed 753 times
resized-image-Promo - 2022-01-04T010825.525.jpeg
resized-image-Promo - 2022-01-04T010825.525.jpeg (91.64 KiB) Viewed 753 times
This epoxy has a 72 hour cure time, so mark your calendars

Hummina Shadeeba   10 MW

10 MW
Posts: 2397
Joined: Apr 25 2014 10:26pm
Location: Sausalito

Re: DIY Toroidal Axial Flux PM

Post by Hummina Shadeeba » Jan 04 2022 11:32am

It’s a long thread and haven’t read it all but why not add a bit of iron for torque and as it is now I imagine it won’t be able to put out much. Maybe iron powder? Will put more mechanical stress on everything but when that’s solved you’d have a motor that’s more practical. Then again it’s still unlikely to put out much torque and maybe not the goal.

User avatar
HalbachHero   100 W

100 W
Posts: 209
Joined: Apr 05 2021 8:35pm

Re: DIY Toroidal Axial Flux PM

Post by HalbachHero » Jan 04 2022 2:58pm

Hummina Shadeeba wrote:
Jan 04 2022 11:32am
It’s a long thread and haven’t read it all but why not add a bit of iron for torque and as it is now I imagine it won’t be able to put out much. Maybe iron powder? Will put more mechanical stress on everything but when that’s solved you’d have a motor that’s more practical. Then again it’s still unlikely to put out much torque and maybe not the goal.
The original idea was to keep it simple. Make it easy to DIY, but to still have some purpose. By making it an air core, a lot of complexity was avoided. The original purpose of this was to be a lightweight generator, but I had fun making motors, and I can still make a generator down the road.

I have been able to consistently make these stators with the jig and the winding pattern I've come up with. I think I have some simple ways of adding iron to the equation, but yeah mechanical stresses could be an issue.

I think there are still a few iterations that this could go through, but ultimately, It was and still is a platform to experiment with, and a project for me to learn some DIY crafting skills or whatever.

There are still things I would like to try.
Adding iron to the core, seeing how close I can get things together, seeing how fast I can spin things. I have thoughts of scaling this up in size, or using PCB/ribbon cables instead of hand winding, making a rotor with both magnet faces usable to potentially stack these to use them in parallel, and gear reductions

Then there's all sorts of things I could do with them. RC cars/drones/e-foil. Maybe a scooter/e-bike. Lawn equipment?? or back to the generator idea.

If nothing else it's been fun, and who knows maybe with more improvement it could have some torque.

Hummina Shadeeba   10 MW

10 MW
Posts: 2397
Joined: Apr 25 2014 10:26pm
Location: Sausalito

Re: DIY Toroidal Axial Flux PM

Post by Hummina Shadeeba » Jan 04 2022 4:17pm

A super wide thrust bearing and iron powder sounds good to me.

https://www.amazon.com/Othmro-Needle-Be ... SVDL&psc=1

User avatar
HalbachHero   100 W

100 W
Posts: 209
Joined: Apr 05 2021 8:35pm

Re: DIY Toroidal Axial Flux PM

Post by HalbachHero » Jan 05 2022 9:59am

Yeah I have considered the thrust bearing if I went to larger diameter. I was thinking two of them on the top and bottom of the stator on the inner diameter. But I think at this small of a size it does not provide much benefit. I think the biggest gains that I can still get from it, are going to be adding iron or iron powder to the stator, or by shrinking the air gap, while keeping a reasonable number of turns (I think there's some optimization given the resistance of a 8-10 turn stator at the current diameter)





I got impatient and popped the stator out of the mold at about the 30 hour mark. This was fine, as the epoxy was set enough, but it was flexible. I will be patient with the real ones. But they came out great! There is one major bubble, and there are a bunch of smaller ones but its still the best one yet, and I expected at least some anyway. Obviously the vacuum chamber could help improve this further, but I'm really happy with how the mold performed. The clay plugged the holes nicely, and the parting wax and mold release did exactly what they were supposed to.

The nylon stator was 12mm before I pressed it, and how its about 6.5 all the way around. There are small tufts of nylon sticking through, but I think that will be fixed by using fiberglass on the top and bottom face. I will do another test next including the fiberglass, and try to do a strength comparison.

I also might make some changes to the press to make it less cumbersome to press, like the suggestion of using bolts to press it down, or a V clamp. that would be a step toward being able to fit it in a vacuum chamber.
resized-image-Promo - 2022-01-05T091956.650.jpeg
resized-image-Promo - 2022-01-05T091956.650.jpeg (83.92 KiB) Viewed 696 times
resized-image-Promo - 2022-01-05T092102.334.jpeg
resized-image-Promo - 2022-01-05T092102.334.jpeg (97.71 KiB) Viewed 696 times
resized-image-Promo - 2022-01-05T091937.428.jpeg
resized-image-Promo - 2022-01-05T091937.428.jpeg (90.04 KiB) Viewed 696 times
Did a dry fit to check the tolerances. Everything looks good so far
resized-image-Promo - 2022-01-05T092123.203.jpeg
resized-image-Promo - 2022-01-05T092123.203.jpeg (97.4 KiB) Viewed 696 times

User avatar
fechter   100 GW

100 GW
Posts: 15763
Joined: Dec 31 2006 3:23pm
Location: California Bay Area, USA

Re: DIY Toroidal Axial Flux PM

Post by fechter » Jan 06 2022 1:04am

Won't the string have sort of high resistance losses?

Just kidding, it looks great. It will be interesting to get some performance data. I still think adding iron will make a big improvement.
"One test is worth a thousand opinions"

Hummina Shadeeba   10 MW

10 MW
Posts: 2397
Joined: Apr 25 2014 10:26pm
Location: Sausalito

Re: DIY Toroidal Axial Flux PM

Post by Hummina Shadeeba » Jan 06 2022 3:11am

motor needs thicker string

User avatar
HalbachHero   100 W

100 W
Posts: 209
Joined: Apr 05 2021 8:35pm

Re: DIY Toroidal Axial Flux PM

Post by HalbachHero » Jan 06 2022 3:31pm

Ha! yeah, I wonder if you could generate any amount of voltage. The conductivity of Nylon is quite poor...


I also look forward to getting some data. I figure if I can make really consistent stators. I want to start playing with fewer turns and smaller air gap. I made a spreadsheet in which I tried to calculate the specific Km of the motor given a range of potential Kv values, and the calculated resistance of various number of turns. It seems that there is an optimization to be had here. too many turns results in higher resistive losses. It seems the sweet spot for my motor is something like 4-8 turns.

All of this is based on the current data I have for the 6 turn stator. I scaled Kv linearly by the number of turns (I realize that it would not be the case, but until I run better simulations, that's what I'm going with). You can see there is sort of a point of diminishing returns from a Km standpoint.

I scaled the weight based on the base weight of the rotors and other parts, and then made an estimate on the amount of extra weight per turn. I know for a fact, 1 turn is equal to 10g. So I added an extra 10g for the extra epoxy and thickness of other parts. Changing this weight increase per turn really affects how quickly the extra turn become relevant or not. This might allow for the argument or removing that extra weight from the iron back....

Please let me know if you think my math is wrong.
Screenshot 2022-01-06 151715.png
Screenshot 2022-01-06 151715.png (159.17 KiB) Viewed 645 times
Think this also validates my thought of getting things closer together. Less turns, but more flux, which means Kv would not change linearly given a change in total air gap, but a static gap between the rotor face and stator. I think I am more likely to achieve the -10% Kv part of the chart if I have a smaller airgap.

User avatar
HalbachHero   100 W

100 W
Posts: 209
Joined: Apr 05 2021 8:35pm

Re: DIY Toroidal Axial Flux PM

Post by HalbachHero » Jan 06 2022 5:07pm

Just played with that chart a bit more. After adding exponential increases and decreases to Kv the results are even more interesting. I still think simulations would make these more real, since playing with the rate of change of Kv between turns. the green "parabola" moves up and down a bit
Screenshot 2022-01-06 170410.png
Screenshot 2022-01-06 170410.png (114.64 KiB) Viewed 640 times

User avatar
HalbachHero   100 W

100 W
Posts: 209
Joined: Apr 05 2021 8:35pm

Re: DIY Toroidal Axial Flux PM

Post by HalbachHero » Jan 09 2022 7:07pm

I made a few small modifications to the top part of the press, to help deal with the wire, and I added numbers to the top to know what thickness it will produce when used with the bottom I used last time.

I also made a part that I am hoping I can just press right into the center of the stator. I'm hoping this will help maintain a consistent inner diameter, and will give more consistent mounting points. I added fins just to increase the surface area just to make it stronger. I also have to imagine this will slightly decrease weight while its 3D printed plastic. But it would be an expensive part to machine and would weigh more than just keeping it as epoxy. We will see if this is a bad design decision.

I made a new stencil for the fiberglass rings. I will be using these on the next test and soon I'll start playing with chopped fiberglass tow for added strength.

I need to add bolt holes to the middle of the press, but my concern is the amount of resin that escapes from the center. I think I will end up locking the bolts in place. I guess its worth a test. I also will look into getting a v clamp for the outer part. or maybe just bolt the outside to. I need to eliminate the need for the clamps if I am going to fit it in a vacuum chamber.

Oh yeah I bought a vacuum chamber. Also I bought a pressure pot and a vacuum pump. the pump is only 72 microns. If that doesn't work, I can try something better. We're in it now!
resized-image-Promo - 2022-01-09T184850.033.jpeg
resized-image-Promo - 2022-01-09T184850.033.jpeg (103.06 KiB) Viewed 597 times
resized-image-Promo - 2022-01-09T184910.963.jpeg
resized-image-Promo - 2022-01-09T184910.963.jpeg (109.11 KiB) Viewed 597 times

User avatar
APL   100 kW

100 kW
Posts: 1062
Joined: Aug 06 2018 8:44pm

Re: DIY Toroidal Axial Flux PM

Post by APL » Jan 10 2022 9:16pm

Congrat's on your last stator, that's looking much better,.. I like your newest design too, your on a roll! This is going to be the best one yet! :thumb:

User avatar
HalbachHero   100 W

100 W
Posts: 209
Joined: Apr 05 2021 8:35pm

Re: DIY Toroidal Axial Flux PM

Post by HalbachHero » Jan 16 2022 9:50am

APL wrote:
Jan 10 2022 9:16pm
Congrat's on your last stator, that's looking much better,.. I like your newest design too, your on a roll! This is going to be the best one yet! :thumb:
Thanks! yeah this project has certainly been a learning process every step of the way, but my workshop is now armed with quite a few tools for crafting now. Looking forward to putting them to use over the years.

I just made a 5 turn nylon stator, and used the new press to squish it to 4.5mm. I used the fiberglass tow to push into the slots between each leg of the coils, but I will try a different method for the next attempt. I tried to put the fiberglass down, then the resin then push it in. That didn't work out so well. Next time I will try to push it in a bit fist add resin, and repeat until its full.

It was definitely harder to get it totally flat on this go. I'm not sure if that's due to the use of the fiberglass, or the 3d printed part in the center, but we will see how the end result turns out.

I had done a test with the pressure pot prior to this, and its incredible to see how clear the epoxy turns out. I almost did another nylon stator without the tow, and just using the pressure to get a better look that how things sat when pressed. But I'm getting tired of making nylon stators. I want to get back to copper.

If this one cures nicely with few bubbles, I think its back to copper. I pretty excited for this next version

User avatar
APL   100 kW

100 kW
Posts: 1062
Joined: Aug 06 2018 8:44pm

Re: DIY Toroidal Axial Flux PM

Post by APL » Jan 16 2022 1:55pm

Awesome! Good to hear about the pressure pot results. 4.5mm is a new record, I can see where it's tempting to get back to copper,
but since every nylon stator you've made is like twice as good as the last one, it would seem logical to stay the course.

Getting very close though, and you can only get a rough estimate with nylon, so your probably right. Either way, best of luck! :thumb:

User avatar
HalbachHero   100 W

100 W
Posts: 209
Joined: Apr 05 2021 8:35pm

Re: DIY Toroidal Axial Flux PM

Post by HalbachHero » Jan 17 2022 11:12am

So the potting went pretty well. There were a few air bubbles around the printed stator hub. I think I can avoid these in the future by brushing some epoxy onto the center part first, also the fiberglass sheets will help.

This did not compress to the desired 4.5mm even though the mold appeared to press down completely. Either way. Its 5mm, only off by 0.5mm so I'll accept that margin of error.

Also I added holes on the inner diameter to help push out the stator. This worked, but I punched a hole in the plastic in the process. So I might try to avoid using those or at least be gentler.

I think if I do this one or two more times I will get the hang of it. The mold needs a better pressing mechanism for the outer diameter. I used clamps to get it pressed then screwed a block of wood on either side to each other to hold it there. It seems once its pressed it doesn't back out on its own, but I also didn't need to squish the 5 turn nylon stator down much. The nylon doesn't compress as much as all the strands of litz wire, which means that prior to pressing things, the copper stators will stand a bit taller than the hub. I'm hoping this doesn't become an issue.

I think I am going to wind up a copper one. I'm going to plan for 6 turns @ 4.5mm and I'll use the fiberglass sheets as well as the tow. I'll have to get some better epoxy too, so it might be a bit before its done.

Nothing super special to look at, but it's the next step in the evolution
resized-image-Promo - 2022-01-17T111138.759.jpeg
resized-image-Promo - 2022-01-17T111138.759.jpeg (103.83 KiB) Viewed 467 times
resized-image-Promo - 2022-01-17T111113.825.jpeg
resized-image-Promo - 2022-01-17T111113.825.jpeg (87.77 KiB) Viewed 467 times

Post Reply