The limits of torque production for a given weight of motor

Electric Motors and Controllers
User avatar
Miles   100 GW

100 GW
Posts: 11032
Joined: Mar 16 2007 4:15pm
Location: London UK

Re: The limits of torque production for a given weight of mo

Post by Miles » Mar 20 2016 6:47pm

We're not really tackling the variable gearbox argument, here. That's been done to death in the other thread.

User avatar
speedmd   1 GW

1 GW
Posts: 3210
Joined: Nov 14 2012 12:16pm
Location: new england

Re: The limits of torque production for a given weight of mo

Post by speedmd » Mar 20 2016 6:50pm

John in CR wrote:What leads you guys to believe that the smaller diameter motor won't always have an advantage? I ask because it's the non-productive weight in the stator and rotor supports that will kill the the large diameter motors in a direct comparison based on weight.
This is exactly why I think we need to establish what is needed for bottom end torque as well as top rpm suitability. A small motor will be fine in most situations if you can wind it up and gear it low. We need to weight the bottom end requirements if we truly want to find a more universal configuration - solution for a direct drive configuration.

John in CR   100 GW

100 GW
Posts: 14955
Joined: May 20 2008 12:58am
Location: Paradise

Re: The limits of torque production for a given weight of mo

Post by John in CR » Mar 20 2016 6:50pm

Miles wrote:I'm not believing anything. I'm an agnostic.
I love hubmotors, but know better than to get into that comparison. The magnetic working area still has to be supported just as rigidly, so doing it at a larger diameter has to weigh more assuming equal materials...or are the real limitations the same as those encountered when trying to come up with a practical hubless wheel?

User avatar
speedmd   1 GW

1 GW
Posts: 3210
Joined: Nov 14 2012 12:16pm
Location: new england

Re: The limits of torque production for a given weight of mo

Post by speedmd » Mar 20 2016 6:53pm

Miles wrote:We're not really tackling the variable gearbox argument, here. That's been done to death in the other thread.
Understood. Better to keep this thread to finding some golden rule of length to diameter if possible.

John in CR   100 GW

100 GW
Posts: 14955
Joined: May 20 2008 12:58am
Location: Paradise

Re: The limits of torque production for a given weight of mo

Post by John in CR » Mar 20 2016 7:01pm

speedmd wrote:
Miles wrote:We're not really tackling the variable gearbox argument, here. That's been done to death in the other thread.
....Better to keep this thread to finding some golden rule of length to diameter if possible.
LOL, that's not going to happen.

Hillhater   100 GW

100 GW
Posts: 12450
Joined: Aug 03 2010 10:33pm
Location: Sydney ..(Hilly part !) .. Australia/ Down under !

Re: The limits of torque production for a given weight of mo

Post by Hillhater » Mar 20 2016 7:33pm

Miles wrote:We can run the Astro 3220 at 10,000 rpm.
It can make about 5Nm of torque continuously.
If we assume a reference wheel speed of 500rpm that means a reduction of 20:1

So, we're looking at a hub motor of 2 kg, plus whatever the weight of reduction gear for the Astro is, capable of putting out about 100 Nm continuously.

Is that fair?
Gearbox weight ? Why
I thought you were "simply" going to see if (for example) that Astro could be reconfigured to produce 100NM at 500rpm for the same 2kg (?) weight of "active" materials ? At similar efficiency levels ?
And if not, what are the issues preventing it
Then maybe assesing the practical issues of such a configuration ?
This forum owes its existence to Justin of ebikes.ca

User avatar
Miles   100 GW

100 GW
Posts: 11032
Joined: Mar 16 2007 4:15pm
Location: London UK

Re: The limits of torque production for a given weight of mo

Post by Miles » Mar 21 2016 2:06am

Hillhater wrote:Gearbox weight ? Why
We don't have to. In which case, we don't have to do anything much. I can't envisage a 2kg hub motor, capable of 50Nm/kg, using present technology....

Hillhater   100 GW

100 GW
Posts: 12450
Joined: Aug 03 2010 10:33pm
Location: Sydney ..(Hilly part !) .. Australia/ Down under !

Re: The limits of torque production for a given weight of mo

Post by Hillhater » Mar 21 2016 4:24am

But do you know the max potential torque or the limiting reasons using available tech ?
This forum owes its existence to Justin of ebikes.ca

User avatar
Miles   100 GW

100 GW
Posts: 11032
Joined: Mar 16 2007 4:15pm
Location: London UK

Re: The limits of torque production for a given weight of mo

Post by Miles » Mar 21 2016 4:33am

We could explore that. It's kind of what we were doing here, too: viewtopic.php?f=30&t=57371

Hillhater   100 GW

100 GW
Posts: 12450
Joined: Aug 03 2010 10:33pm
Location: Sydney ..(Hilly part !) .. Australia/ Down under !

Re: The limits of torque production for a given weight of mo

Post by Hillhater » Mar 21 2016 6:23am

.? I thought that was the objective of this thread...without the constraints of being a hub motor ?

PS ..what happened to the hub motor project ?...no posts in 18 months ?
This forum owes its existence to Justin of ebikes.ca

User avatar
speedmd   1 GW

1 GW
Posts: 3210
Joined: Nov 14 2012 12:16pm
Location: new england

Re: The limits of torque production for a given weight of mo

Post by speedmd » Mar 21 2016 7:58am

Miles wrote:
Hillhater wrote:Gearbox weight ? Why
We don't have to. In which case, we don't have to do anything much. I can't envisage a 2kg hub motor, capable of 50Nm/kg, using present technology....
Your simulations on the ultimate hub were showing some impressive specific torque values. Active weight was in the 2.5kg range. No way a small astro style motor would come close to the calculated torque without a gear box and its associated weight.

Punx0r   100 GW

100 GW
Posts: 5492
Joined: May 03 2012 8:16am
Location: England

Re: The limits of torque production for a given weight of mo

Post by Punx0r » Mar 21 2016 12:16pm

Ignoring the weight and torque loss of the gear reduction does simplify the problem but also reveals an obvious logical limit to the idea of simply increasing the diameter to trade torque for speed using a given weight of active material & supporting structure (packaging): 2kg of motor clearly cannot produce infinite torque being infinitely great in diameter as the packaging would have to be infinitely strong.

Therefore, I think there must be an optimum diameter to optimise torque output per unit weight of motor, dictated by the strength of the available packaging material, after which a gear reduction possibly becomes the better solution. That point would be affected by the weight and efficiency of the reduction, hence those variables get thrown back into the mix.

Why do engineering problems always have to be a mess of compromises? ;)

User avatar
Miles   100 GW

100 GW
Posts: 11032
Joined: Mar 16 2007 4:15pm
Location: London UK

Re: The limits of torque production for a given weight of mo

Post by Miles » Mar 21 2016 1:06pm

OK. Let's try something a bit more extreme than on the ultimate hub motor thread.

I think it will be important to have a high number of symmetries in the layout for this............

Hillhater   100 GW

100 GW
Posts: 12450
Joined: Aug 03 2010 10:33pm
Location: Sydney ..(Hilly part !) .. Australia/ Down under !

Re: The limits of torque production for a given weight of mo

Post by Hillhater » Mar 21 2016 4:25pm

Punx0r wrote:Ignoring the weight and torque loss of the gear reduction does simplify the problem but also reveals an obvious logical limit to the idea of simply increasing the diameter to trade torque for speed using a given weight of active material & supporting structure (packaging): 2kg of motor clearly cannot produce infinite torque being infinitely great in diameter as the packaging would have to be infinitely strong.

Therefore, I think there must be an optimum diameter to optimise torque output per unit weight of motor, dictated by the strength of the available packaging material, after which a gear reduction possibly becomes the better solution. That point would be affected by the weight and efficiency of the reduction, hence those variables get thrown back into the mix.

Why do engineering problems always have to be a mess of compromises? ;)
I think you are jumping a few steps there..
Why not First see what can be done with "2 kg of ACTIVE Materials"...find the limiting factors
. Then review what that means in practice when you consider the "packaging ?
IE..throw some random test numbers into the mix... 0.5m, 1.0m dia air gap ? ...what does that give us with 2kg AM's ?
"Shake the tree" until something falls out ?
This forum owes its existence to Justin of ebikes.ca

User avatar
Miles   100 GW

100 GW
Posts: 11032
Joined: Mar 16 2007 4:15pm
Location: London UK

Re: The limits of torque production for a given weight of mo

Post by Miles » Mar 21 2016 5:19pm

Hillhater wrote:Why not First see what can be done with "2 kg of ACTIVE Materials"...find the limiting factors
Made a start.

This is peak eta.
Attachments
Limits 1.001-102t-108p-20a Pk eta.pdf
(513.38 KiB) Downloaded 134 times

Hillhater   100 GW

100 GW
Posts: 12450
Joined: Aug 03 2010 10:33pm
Location: Sydney ..(Hilly part !) .. Australia/ Down under !

Re: The limits of torque production for a given weight of mo

Post by Hillhater » Mar 21 2016 6:12pm

Interesting Miles.
Does that imply that is viable ?...or is there any issues you see in those results ?
Is the 20A an input parameter to maintain a 1kW power rating ?
What would be the amp limiting factor, and what would be the result ?
This forum owes its existence to Justin of ebikes.ca

User avatar
Miles   100 GW

100 GW
Posts: 11032
Joined: Mar 16 2007 4:15pm
Location: London UK

Re: The limits of torque production for a given weight of mo

Post by Miles » Mar 21 2016 6:19pm

We could certainly double the amps. How much further beyond that depends on the ability to shed heat.

20 amps was a guess to achieve peak eta. I was lucky :)

All the parameters were guesstimates. Still a lot of tweaking to be done.

Looks like we can get 50Nm continuous from 2kg of active materials, though...

This slot pole combination gives us 3 layout symmetries which will help distribute radial forces.

Winding distribution factor is 0.953

LCM is 1836 so, not a lot of cogging.....

Punx0r   100 GW

100 GW
Posts: 5492
Joined: May 03 2012 8:16am
Location: England

Re: The limits of torque production for a given weight of mo

Post by Punx0r » Mar 21 2016 6:23pm

Interesting simulation :)

Comparing to the Astro:
Miles wrote:We can run the Astro 3220 at 10,000 rpm.

It can make about 5Nm of torque continuously.
The simulation has 23Nm at 400rpm for 1000W input (I assume that is continuous, as based on 80°C winding temp?). The Astro would be 5.2Kw output.

Does the ~2T flux density of the simulation indicate its already close to its peak torque?

User avatar
Miles   100 GW

100 GW
Posts: 11032
Joined: Mar 16 2007 4:15pm
Location: London UK

Re: The limits of torque production for a given weight of mo

Post by Miles » Mar 21 2016 6:28pm

Punx0r wrote:Does the ~2T flux density of the simulation indicate its already close to its peak torque?
No. You have to look at the distribution in the graphics.

Punx0r   100 GW

100 GW
Posts: 5492
Joined: May 03 2012 8:16am
Location: England

Re: The limits of torque production for a given weight of mo

Post by Punx0r » Mar 21 2016 6:30pm

Oops, posted at the same time. Since it's clearly not near saturation, how would it perform at 4 or 5kW input? If the torque can be quadrupled it would hit the Astro plus gear reduction target. Well, minus the packaging...

Astro state the 3220 is at 90% efficiency for 4.2kW continuous output.

I'd hope that equal or better heat shedding for a physically larger motor wouldn't be too hard to achieve :)

Punx0r   100 GW

100 GW
Posts: 5492
Joined: May 03 2012 8:16am
Location: England

Re: The limits of torque production for a given weight of mo

Post by Punx0r » Mar 21 2016 6:31pm

I plead ignorance ;)

Teh Stork   1 kW

1 kW
Posts: 461
Joined: May 25 2011 8:02am

Re: The limits of torque production for a given weight of mo

Post by Teh Stork » Mar 21 2016 8:36pm

Miles wrote:I thought it would be good to analyse a reasonably simple case - that we can swap speed for torque and keep motor weight constant. Is it not relevant as a first step? Any other discussion can carry on on the original thread......
Ok.

My understanding of motor design is that power production is a function of active area and magnet speed. Active area gives torque. Magnet speed supplies the necessary angular speed to produce power.

Bigger diameter: more magnet speed - more power.
Bigger active area: More torque

As in: more RPM -> more magnet speed -> more power. As in: Wider stator and magnets -> more torque -> more power.

Proposed case: Take one motor. Change one parameter, change everything else - does the results match up?

I'd like to entertain this, but I am not able to do the FEMM analysis - so I'll start with a proposed method.

1. Find a suitable hub motor.
2. Design another hub motor with different parameters, but same performance.
3. What material is needed to achieve this?
4. Does material/performance match?

A test:

1. Selecting Crystalyte HS3540 ish.
Active area: ~220mm diameter, Stator height: 40mm. Total active area ~27 646 square mm.
Max RPM: 350 RPM

Keep RPM constant.

2. New hub motor. Slimline rimalichious hubmotor. Diameter 550 mm. The question now is the new active area needed to match performance?
taking regard to T=k*D^2*L. T=Torque, k = constant, D=Diameter, L= Length; I'm going to take a (wild) guess at what new active area is needed.
550/220=2,5.
2,5^2=6,25.
27646/6,25 = 4400.
New active area: 4400 square mm.

Motor magnets and stator width: 4400/(550*pi) = 2,5 mm

Remember RPM is constant. As the magnets in the new design has a 2,5 better lever arm than the HS3540 - only 0,4x the force must be generated by the magnets.

3 and 4. "What material is needed to achieve this?" and "Does material/performance match?"
How thick must the magnets be?
How thick must the stator be?
How much copper is needed?
Will the simulation even match the original motor performance?

My intuition tells me the calculations in nr 2 is wrong - but from my moderate experience with BL motors i think that my intuition is wrong. (My intuition tells me new motor active area should be divided by 2,5). Would be cool if someone would entertain the groundwork in this post and finding the fault(s).
2013 VW E-UP!
Kona mtb: Cromotor, Kelly KEB, ~6kW peak, 1,4kWh 90V battery 29E cells, 38 kg

Hillhater   100 GW

100 GW
Posts: 12450
Joined: Aug 03 2010 10:33pm
Location: Sydney ..(Hilly part !) .. Australia/ Down under !

Re: The limits of torque production for a given weight of mo

Post by Hillhater » Mar 21 2016 10:39pm

TS,....did you read the thread title ?
The objective is to see how much torque can be produced from a given weight of active materials .
This forum owes its existence to Justin of ebikes.ca

User avatar
speedmd   1 GW

1 GW
Posts: 3210
Joined: Nov 14 2012 12:16pm
Location: new england

Re: The limits of torque production for a given weight of mo

Post by speedmd » Mar 22 2016 12:07pm

Are we just looking for max torque? It is the power band that is key IMO. If you have enough torque, you can effectively create acceleration. Being able to do this efficiently throughout the design speed range is what determines a motors worth. Thinking that a Broad/ large Area under the torque- rpm curve is key. Not sure what we should do on limiting to one design frequency/ design rpm/ turn count etc., or if that matters at all in just rough size comparisons.

Hillhater   100 GW

100 GW
Posts: 12450
Joined: Aug 03 2010 10:33pm
Location: Sydney ..(Hilly part !) .. Australia/ Down under !

Re: The limits of torque production for a given weight of mo

Post by Hillhater » Mar 22 2016 5:48pm

One parameter at a time.
So , aim for Max torque,...then look to see what the implications are in terms of efficiency, power, rpm range, etc etc. ?
This forum owes its existence to Justin of ebikes.ca

Post Reply