Optimal Hub Motor design? Break down of BionX D-Series

[madin88]

https://wk.baidu.com/view/a3647f25aaea998fcc220eba?pcf=2#2

That steel would help its design work substantially better, but still a crutch.

Luke you have had right in terms of that.
It seems Bionx made some compromises on the D-motor. As they were using 0.35mm lams, they had to reduce the air gap flux density on purpose to keep iron losses on a normal level.
I did some research about magnetic designes and noticed follwing issues:

First:
The thickness of the back iron is insufficient.
If we assume the grade of the magnets is N35 (about 1,2T), the thickness should be 1/3 of the width of the magnets for optimal (close to 100%) steel return.
The D-motor magnets have a width of 7,6mm and the back iron is 1,6mm. If we do the math now (7,6mm / 3) it should be at least 2,5mm thick.

-> therefore the initial field strength of the magnets is already reduced!

Next thing is the effect of the airgap width in combination with the thickness of the magnets:

The flux density which reaches the stator teeth can be calculated like this:

thickness of magents / (thickness of magnets + air gap width)
so: 2mm / (2mm + 1mm) = 0,66%

-> this means the remaining field strenght, which is already lower due to the too thin back iron, will be reduced further to only 66%


If Bionx would have used thinner laminations of better or mentiond steel ($$$), they could have kept the air gap flux density on a MUCH higher level making it alot more torquey (probably up to twice as much?), and so much more interesting for hot-rodding.
Unfortunately, and as real world experince shows, it seems like it was just designed to be as cheap as possible for the torque target of 50Nm without any headroom.

 

[neptronix]

It's quite a shame isn't it?
MAC has 0.27mm laminations in their motors, whereas other geared motor makers are still using 0.35mm.
DD hub makers don't seem to have wised up to using 0.27mm or less yet.

The jump from 0.5 to 0.35mm in common DD hubs was so huge that it increased continuous power by ~33% and efficiency jumped up from low 80's to upper 80's, even into the low 90's.

The first company to sell a DD with 0.27mm lams will knock the leafmotor off my recommendations list 8)

Bionx with their proprietaryness, high price, and lack of exceptional efficiency has gotta be low on my list..

 

[justin_le]

It's quite a shame isn't it?
MAC has 0.27mm laminations in their motors, whereas other geared motor makers are still using 0.35mm.
DD hub makers don't seem to have wised up to using 0.27mm or less yet.

True but the benefit of thinner laminations is really only present when you are running the motor at increasingly high eRPMs when eddie currents in the stator make up a good chunk of the stator losses. With skateboard DD motors we see 0.27 and even thinner lams all the time since the eRPM's are crazy. With a lot of ebike motors (both geared and DD) that I've dyno characterized the core losses at normal running speeds are still mostly made up of hysteresis losses and not eddie losses, so there's diminishing returns to thinner and thinner laminations especially since this reduces the iron fill factor of the stator.

On the subject at hand of the BionX D motor, the core drag at 250 rpm (typical ebike speed) is 0.68 Nm, of which 0.45 Nm comes from hystersis of the iron (not affected by lamination thickness) and ~0.23 Nm comes from eddie currents. If they went with 0.27 instead of 0.35mm, the expected reduction in drag would be 40% of the 0.23Nm, or a total of 0.09 Nm less drag. That's a 13% total reduction in drag, not nothing, but it's not something anyone would readily notice either.

They could then yes increase the air gap field strength so that the drag was back up to the 0.68 Nm and had a slightly more powerful motor, but not a vastly more powerful motor!

If Bionx would have used thinner laminations of better or mentiond steel ($$$), they could have kept the air gap flux density on a MUCH higher level making it alot more torquey (probably up to twice as much?), and so much more interesting for hot-rodding.

Those claims are certainly a little optimistic No doubt there are design tradeoffs where marginal increases in performance aren't worth significant increases in cost. Anyways I'm no BionX fanboy at all but you can't dispute the fact that even with their design tradeoffs, the BionX 'D' motor had the best torque to weight ratio of ANY direct drive ebike hub motor that has made it to market.

One way is to look at the ratio of Ki^2 / R which is a normalized metric for motor power capability. The BionX 'D' motor is 8.02, while the common 45mm stator DD hubs like the MXUS 3KW motors or Leaf motors etc. are about 10.3- 10.7 from my tests. But the BionX 'D' motor is 4.2kg, the MXUS and kin are 9kg.

If we normalize by mass: Ki^2/R*m, then we have a true scale of relative motor performance. The BionX D is at 1.91, all other DD motors I look at tend to be between 0.8 - 1.2 on this scale.

It should still be any hot rodder's dream, If only they didn't use a permanent glue to seal the damn things shut!

 

[madin88]

On the subject at hand of the BionX D motor, the core drag at 250 rpm (typical ebike speed) is 0.68 Nm, of which 0.45 Nm comes from hystersis of the iron (not affected by lamination thickness) and ~0.23 Nm comes from eddie currents. If they went with 0.27 instead of 0.35mm, the expected reduction in drag would be 40% of the 0.23Nm, or a total of 0.09 Nm less drag. That's a 13% total reduction in drag, not nothing, but it's not something anyone would readily notice either.

They could then yes increase the air gap field strength so that the drag was back up to the 0.68 Nm and had a slightly more powerful motor, but not a vastly more powerful motor!

Thank you for the detailed explanation.
I wonder what would be possible in terms of reduction of hysteresis losses when using newest electrical steel grades and no spare of expenses.

Those claims are certainly a little optimistic No doubt there are design tradeoffs where marginal increases in performance aren't worth significant increases in cost. Anyways I'm no BionX fanboy at all but you can't dispute the fact that even with their design tradeoffs, the BionX 'D' motor had the best torque to weight ratio of ANY direct drive ebike hub motor that has made it to market.

Yes, if operated within specs, it is the best DD ebike hub motor in the world ever made
But if we talk about taking 70Nm or 80Nm from it which isn't that much for hot-rodding, i am quite sure that things will look different and it rather would be only on par with other Hubmotors in this power class (in the weight range of 6-7kg).

As a guess, the D-motor probably already operates close at the "knee" of the the BH-curve at it's specs:

The reason from where my claim about twice the torque came from is that most BLDC motors have a flux density of around 0.9-1T, but on this D-motor it is more like 0,5-0,6T due to the combination of insufficienct back iron thickness, thin magnets and large airgap.
The other issue is why the hell they were using 11,5mm wide magnets on a stator that is 12mm wide and wasting the potential for 4% more torque.
I have done the calculations so those numbers are quite true, but you have absolutely right that this claim is way to optimistic and not practical as the drag in this case would be awefully high even with better and thinner steel lams.

It should still be any hot rodder's dream, If only they didn't use a permanent glue to seal the damn things shut!

Yes too bad that almost averything that came from BionX was built so that it has to be replaced instead of the possibility for a repair.
Speaking about theire batteries mostly, but the D-motor is a good example as well. In case of a controller defect the entire motor needs to be replaced which is a waste of material and should have been forbidden long time ago.

Did you ever had the chance to measure saturation (and compare it with the all-axle hub)?
It would be really useful to know how much amps it can take before kT starts to drop off to much, but i know that such test would be very time consuming and probably not worth the hassle for a motor like this which isn't produced anymore and rarely can be found as a used part.

 

[amberwolf]

It should still be any hot rodder's dream, If only they didn't use a permanent glue to seal the damn things shut!

Is it glue, or ultrasonic welding?

(probably doesn't matter)

 

[justin_le]

It should still be any hot rodder's dream, If only they didn't use a permanent glue to seal the damn things shut!

Is it glue, or ultrasonic welding?

(probably doesn't matter)

Well whatever it is it's a real PITA to separate the two halves apart in a way that it can be neatly put back together again. It could be that I just didn't have the technique down but the combination of a heat gun and xacto blade still resulted in a lot of cracked plastic and such.

If the process of opening and resealing these motors was more straightforward then I really do think that they would be high on the list of hotrodders ideal hub motor to hack. It's got a 50% higher torque to weight ratio of most other performance DD hub motors, and even with the composite shell it has a much better thermal performance too in account of the large diameter and surface area for shedding the heat.

 

[madin88]

The glue they used is very hard and it doesn't soften much when heating it up, but i had the feeling that it was at least a bit helpful and that we did it really slowly. Surely it needs some luck that nothing cracks.

It's got a 50% higher torque to weight ratio of most other performance DD hub motors, and even with the composite shell it has a much better thermal performance too in account of the large diameter and surface area for shedding the heat.

Yep! I could ride past 70kmh at 2000W+ forever on the flat while the temperature did stay below 70-80°C without any cooling mods. That was with a total weight of 125-130kg.
Thats impressive and shows how good the thermal performance is, but on the next slight hill the winding temperature jumped almost instantly up to 110-120°where i have set the cut off, and i had to go off the throttle and decrease the speed alot.
Maybe i am too much used to the better "heat sponge" of heavier hub motors, but i still think that this motor goes sooner into saturation as other hubmotors do. The other issue which was killing most of the fun was the terrible noise it made in the Vector frame.

Now it is lying unused in the basement and i have absolutely no idea for which project i could use it as i would like to keep it laced into the 20" wheel (which wasn't cheap).

Any ideas?
With the fat tire there still would be enough place for a 5-6 speed cassette. For a true high power e-bike or light moped conversion it definitely has to low torque reserves (at least without any cooling mods).
I also would be willing to sell it or trade in exchange for something else. If someone is interested -> PM

 

[Mboyle1210]

Greetings. Much of this thread (the EE part) is beyond me, but I have a DV500 here in Thailand that, after about 1700 km and then, over the course of 50 or 60 km, began to make so much noise as to become unrideable. The unit otherwise seems to operate normally and I don’t mind tearing it apart if there is some light at the end of the tunnel. It cost me $1100 (!!) to get the battery over here so any options are better than none.

Initially I would have described the sound as an amplified cogging and perhaps it still is, though now so loud that the growling approaches a shrieking at anything other over 10 kmh (26” on an HPVelo trike). On a stand and freewheeling it is similar to the vid posted by Madin88 last Oct. Unmounted I can hold the wheel out and turn it back and forth and get the same rumble as the stator sloshes back and forth. At that speed it does sound like a resonance in the plastic housing, but it seems it must be mechanical to occur at such a slow speed. Is there anything in there that could have shifted and is now rubbing on something?

One other question regarding the well documented difficulty of opening the motor up. I certainly don’t have the tools and cannot even figure out the style of the case fixing bolts. Not hex and too big for a T10 and too small for a T15. Mine are a bit chewed up as well. Guidance?

I bought two of these last summer, one before BionX went tits up and one after — because I liked the first so much and there is nothing similar available here in Thailand. I’m going back for a couple months soon and will bring the US unit back here when I return. If there is any reason to bring the noisey unit back to the US before I tear it down, I.e. certified service (very doubtful) I can do so.

Thanks for any help, opinions, etc.

 

[flathill]

Ultrasonic xacto knife will open it right up. Honda makes good ones

 

[madin88]

Greetings. Much of this thread (the EE part) is beyond me, but I have a DV500 here in Thailand that, after about 1700 km and then, over the course of 50 or 60 km, began to make so much noise as to become unrideable. The unit otherwise seems to operate normally and I don’t mind tearing it apart if there is some light at the end of the tunnel. It cost me $1100 (!!) to get the battery over here so any options are better than none.

I don't think that there is anything in there that could have shifted making the niose, but what i noticed on my motor was a brown-bluish color on the bearings which is normally a sign that it did overheat (from to much friction).
That makes me think that it might be possible that worn out bearings could be the reason for the noise.
We should go into the matter!

 

[Mboyle1210]

I have never heard a bad bearing generate a noise like this. It is absolutely regular, and is a “thrrum” rather than a “crunch”. As I mentioned, it’ll generate the sound when the wheel is hand held and rocked back and forth— same with or without load — and really seems to be coming from inside the case rather than a resonance without. But hey. Never say never!

What tool did you use to get the case fixing bolts out?

Thanks

 

[cferron]

Ok guys you are way too techhy for me! I'm impressed by your know how.
I have this issue with my D500 motor, it won't power up when the data cable is connected (power prot message was showing prior on the console) - see this video: (https://www.dropbox.com/s/3o9rj0843e9qbu4/VID_20190321_180332.mp4?dl=0).

since I have nothing to loose, I have decided to open and see what could be wrong inside. Man this thing is hard to open...
Can you give me more cue? I'm trying to access the inside...



Do you have to remove something here?


Do you have to remove something here?


Once you put all this back together, what have you used to reglue or reseal?

Thanks

 

[cferron]

Update

I was able to remove the side where the freehub is attached. By hammering on the shaft I was able to remove this side.
The other side is more complicated. Any cues are appreciated.


View attachment 1

 

[markz]

Once you put all this back together, what have you used to reglue or reseal?

High Temp RTV sealant from an automotive store.

 

[Mboyle1210]

Mark, I’ve got a couple of dv500s, one of which I’ve managed (now on my 3rd rc3 controller) to keep going. One motor is a goner as per my postings a while back, the other, aside from the controller issues, was fine until yesterday, when I experienced some rattling in the free hub. I just thought the cassette lock ring had loosened. VERY disappointed$ to discover the bearings in the free hub are crunchy and on the way out! Less than 3000 km on this motor and never driven in bad weather. Junk.... hard for me to figure out where BionX came by their decent reputation and easier to imagine why they folded up in such a hurry and headed for the hills.

Anyway, I don’t know if you had any success with your controller issues. Both of the failures I’ve had have been identical to each other and suggest nothing more than a loose wire i.e. with a little strategic tugging and pushing Of the wire I could get it to light up. Even though I have a new rc3 in hand, I’m still getting on that way with the second one, but it’ll go eventually as the first one did. As I said, this is the newer controller and they are backward compatible but, unless you’ve gotten the thing back together, it’s too late anyway.

My question is, now that you’ve got the free hub half of the motor separated can, you see anyway the free hub might be serviced without cracking the motor open. I have the old motor and the free hub is smooth so I could just swap them. I cannot get what appears to be a lock ring at the base of the free hub to budge and it seems to be more attached to something on the other side anyway. Of course I can just bomb the old motor to get at the parts, but if I have to do a pristine dismantle and reassembly of the other motor.... well, I needn’t tell you!

Thanks
Michael

 

[bikemikem]

I figured out a way to take my free hub off a couple months ago to fix my strain gauge sensor:

https://endless-sphere.com/forums/viewtopic.php?f=30&t=99379#p1457561

 

[Mboyle1210]

Thank you! I missed that thread.
No piece of cake by any means, particularly as I haven’t the tools to make the splined wrench here in Thailand, but that’ll be what needs, obviously, to be done. I’ll find some shade tree machinist to do it for me. As I have the old motor unlaced and in hand, that should be doable.
Thanks again.

 

[Krishnakirtan]

How fast does the D500 go hacked?

 

[madin88]

How fast does the D500 go hacked?

Do you mean without speed limit on a stock setup (with original Bionx battery and controller), or as in my case with external controller?

 

[rowbiker]

How fast does the D500 go hacked?

Since you say "hacked" I'm assuming you are not interested in the proprietary limits.

The Grin Motor Simulator (https://www.ebikes.ca/tools/simulator2.html) includes the BionX D500 (and the PL350), so I suggest you use that wonderful tool to obtain your answer. Select the "All Motors" option in the dropdown selector to reveal the BionX motor options.

 

[rowbiker]

Whoops! ... 404! my bad. The trouble with storing links in a changing world.

Grin has moved simulator to: https://www.ebikes.ca/tools/simulator.html

 

[Krishnakirtan]

How fast does the D500 go hacked?

Do you mean without speed limit on a stock setup (with original Bionx battery and controller), or as in my case with external controller?

Yes you got it. Whats the maximum MPH you can go?

My d500 can do about 28-30 miles per hour with stock bionx controller and unlocked via software.

 

[madin88]

Yes you got it. Whats the maximum MPH you can go?

My d500 can do about 28-30 miles per hour with stock bionx controller and unlocked via software.

It has been a while since i was riding with the D motor on that heavy bike ( ~50kg), but i remember that it was somewhere between 75 to 80kmh on the flat which it could do for quite a long time. The small wheel makes it possible
That was with 22s 80V battery without any field weakening.

 

[devilbass]

Hi! I have the exact same problem with this motor, now the sound are very loud, i suspect bearing problem, is not the best for this kind of motor our is just pretty cheap, i have access bearing very easy with one side only (freewheel) but for they other, i need to open motor, that suck really.