Quick and dirty method of estimating torque of motors

[Arlo1]

Ok So I think with lam stacks in a inner runner and outer runner motor we should be able to have a rule of thumb for torque vs the side of the stator.
IE with good cooling 1ft/lb per sq inch.

So Lets use colossus for instance its a 2.5" long x 4" dia stator. So we could say that with 1ft?lb per sq" that would be 4*3.14*2.5= 31.4 ft/lbs avalible for continuous duty with good cooling.

Now If I use the same numbers for a couple induction motor stators I have 6" long* 5" dia *3.14 = 94 sq inches which could meen 94 ft/lbs.

I have a 256t frame induction motor and it could be 7 long * 6 dia * 3.14 =131 sq inches so 131ft lbs.

I use 31 ft lbs for colossus because I think I can make it run continuous at that level. Might be hi but its maybe safe to say a well designed motor would be >1ft/lb per sq inch. What do you guys think.

IM not talking about weight. I have 4 induction motors I thing the stators will be a good score for big power induction motors.

 

[flathill]

MOTOR DESIGN: RADIAL AND AXIAL TORQUE SCALING
There is another fundamental difference between radial and axial permanent magnet motor configurations. This notable and often misunderstood difference is the relationship of output torque with respect to motor diameter.
In all cases, torque is the result of force applied at a radius. In all motors, the radius increases directly with the increase in diameter of the motor, resulting in a linear increase in torque with diameter solely due to the increased radius. Also, in a permanent magnet motor, the available force is directly proportional to the flux-carrying capability of the stator. So, as diameter increases and the amount of flux that can be carried in the stator increases, the torque of a motor will also increase.
With a radial motor, as diameter increases, the circumference where the stator pole shoes are located also directly increases (Figure 5). This means that the flux-carrying area of the pole shoes linearly increases with diameter. Therefore, torque increases with the square of diameter, since both the radius and the total flux are linearly increasing with motor diameter.
Figure 5: Radial motor with diameter D and 2D
In addition, for a radial motor the flux-carrying area also directly increases with stator length. Since force is directly related to total flux, increasing the motor length also increases torque in a linear manner.
Therefore, in a radial motor, torque increases with the square of diameter and also increases with additional length of the motor, resulting in an overall cubic function for torque. This results in a radial motor increasing torque proportional to the volume of the motor, since the motor volume also increases with the square of diameter and linearly with the length of the motor.
For an axial motor, the torque increases with the cube of diameter and with

motor length. The reason for this is, again, that torque is force times radius. As the axial motor diameter increases, the radius increases, so the torque increases linearly with diameter due to this radius increase. The force in an axial motor is also directly proportional to the amount of flux that the stator of the motor can handle without saturating. The amount of flux is dependent on the area of the stator, and this area increases with the square of the motor diameter. This is shown in Figure 6 below as area A and 4A. Four times the area can carry four times the flux and this is the result of the stator poles increasing in two dimensions as diameter increases.
Figure 6: Axial motor with diameter D and 2D
Because the radius of the motor increases directly with motor diameter, and the force increases with the square of motor diameter, torque of an axial motor increases with the cube of diameter. Notice that since only the diameter is changing, the axial motor volume only increases by the square of the diameter, even though the torque increases by the cube of diameter. This leads
to the fact that for larger motors, an axial motor can use less material to achieve the same torque, compared to a radial motor. Please note that the above is a theoretical approach, and in actual practice the length of an axial motor does need to increase somewhat with increasing diameters, since the air gap dimension, strength of materials and realistic sized bearings must be considered. This still leaves the axial motor design at an advantage over radial motor designs when compared on the basis of torque generated per unit volume of active material required.
Increasing the length of an axial motor does not provide any additional torque capability for the motor, since the flux in the stator is along the axis of the motor. Adding length to an axial motor does provide more room for additional conductor volume and more surface area for heat dissipation. Adding conductor volume can decrease the total motor resistance and thereby decrease the total conduction (I2R) losses of the motor. This is an excellent approach for achieving high motor efficiencies at low motor speeds and leads to a broad, flat motor efficiency curve. Of course, the extension of the length of an axial motor adds both additional stator steel and conductor material, increasing both the weight and cost of the motor. In addition, the extra stator steel adds additional iron losses in the motor, which hurts high speed efficiency performance. However, up to a limit, by using low loss electrical lamination steel in the stator, overall efficiency can be improved by adding length to an axial motor configuration.
THERMAL CONSIDERATIONS

 

[Lebowski]

funny, for an axial motor he does not consider increasing the length by going to multi-stator (which DOES increase torque such that it increases linearly with motor length)

This would also be a more fair comparison, as the increase in length in a radial flux motor automatically implies more/bigger magnets in the rotor, while
the axial flux length increase he envisions does not increase the amount of magnets. Going multi-stator in an axial flux motor increases the amount of magnets,
making for a more fair comparison.

AF rules ! 8)

 

[Arlo1]

Yes lebowski.... I will play with an AF motor one day.

Thanks flat hill. So I might be able to look at it another way for some simple estimates. Maybe stator surface area x radius? because I forgot the further out from the center of the radius the air gap is the more torque you get because of the leverage.

I am crunching numbers on these motors first to see if they are worth spending ~$500 for magnets on and ~2000 for a controller build to make 100hp plus.
I have 2 cars I'm wanting to convert and another motorcycle but I need to have a guess as the torque and power I will have first.

 

[Kingfish]

Slight corrections to the AF panacea:

We can apply multiple stators (& magnets) to both RF and AF. I can't speak for RF, though with AF, adding more is not quite a linear improvement as FEMM will point out: There's a slight efficiency hit, say 10% per additional stator, so adding more is adding less, and only makes sense where pairing smaller-counts is not possible, e.g. one 10-stator AF machine over two 5-stator machines.

In addition, parasitic drag on AF will be profound unless it is ironless which is clearly an advantage over RF. However, going ironless requires more REE material which is a co$tly proposition unto itself. Ironless also poses more difficult challenges for heat removal.

The one salient point missing from this thread though is left unmentioned: If we're increasing diameter to increase torque - well that's less current through the coils, and less heat too. However we also increase mass and momentum. For wind turbines - this is good. For locomotives - this is good. For racing bikes, this is bad: Smaller fast spinning spindles geared way down are better suited for race day. The challenge then is to come up with a design that minimizes momentum, yet imbues high torque qualities without excessive heating.

The most efficient RF or AF system then is one of a single stator, followed closely by two, and less so with each thereafter. The value of having more stators though is that we again have opportunity to split the load and remove heat.

This is a longwinded way of saying that there's no easy way to figure torque accurately. For AF, the best approach is to take the area of the magnet facing the coil and apply it at the midpoint of the magnet length (and thickness), using that radius from the axle as one factor. Then use the number of turns divided the count of turns facing the magnet face as another factor - with the AWG size of conductor and the current going through it. Then the spacing between the magnet and stator-coils.

Playing with these values is a laborious challenge. I've often thought it might be worthwhile venture to create a parametric design program, but then surely others have walked here before me. This is where I tend to look at industry examples for clues on how to achieve my goals more cost-effectively. Think where you want to be, and look at what you can borrow from off-the-shelf to get you there, perhaps giving up the precise goal for one that's close-enough :wink:

Suddenly I'm thinking of what my Pa used to say: "Close only counts with horseshoes and hand grenades, and not too bad with depth charges". I always though a-bombs would fit right in there too. :lol:

But if you look at the sizes of RF motor diameters, particularly hub motors, you'll find a strange correlation between the back iron diameters and steel pipe.

Observing the obvious sometimes simplifies our calculus for us.
~KF

 

[Arlo1]

The one salient point missing from this thread though is left unmentioned: If we're increasing diameter to increase torque - well that's less current through the coils, and less heat too. However we also increase mass and momentum. For wind turbines - this is good. For locomotives - this is good. For racing bikes, this is bad: Smaller fast spinning spindles geared way down are better suited for race day. The challenge then is to come up with a design that minimizes momentum, yet imbues high torque qualities without excessive heating.

lol:

KF I chose this one thing to focus on so I can get an idea of what I want then I will wind for the RPM I need... Its not for a EBIKE its for a motorcycle or car.
Sure light weight is better but laminations are hard as hell to get cheep. So I found a very abundant source of them. Big ugly induction motors. I can make a new housing for them and cool them better and make a magnet rotor and BAM I have a very powerful motor.

I want 500ft lbs or more from a motor if I'm going to conceder a electric conversion to the Road Runner. I do not want to use brushed motors.
If I can figure out a rough rule of thumb for a continuous and max torque I can make I can see what will be worth the 100s od $ worth of magnets and all my time! I have found no mater what you are using there is a max rpm... You can spin some a little faster but remember eddy currents rise fast and to counter that you need thinner laminations which I've already explored.

So again the porous of this thread is to see if we can re use old motors for something cool

 

[flathill]

I have two Honda civic hybrid motors I was going to build into a motorcycle. One has a busted rotor so you get an extra stator to burn up with no worries (also to mockup case). Making a case is the hard part.

With liquid cooling one of these should be good for at least 60 kw. The efficiency map is out of this world (efficient at low rpm as well as high) with super high grade (temp) mags

They can be stacked easily.

Let me know if you are interested.

In any case I would recommend finding a motor from a hybrid or leaf

 

[jonescg]

I was just about to make a post asking the same kinds of questions, Arlo.

Basically, the AFM140 is going to stay in the bike, so I am looking for an alternative motor for the CRX. I still want direct drive, so it needs to be pretty torquey. It needs to be a short motor, as it will fit between an IRS differential and the front of the car. I can't make an AF motor with ease, so I have considered making a RF motor, but with a sizeable diameter.

And then I found this thread So what do you think? Is it possible to make a liquid cooled radial flux motor capable of spinning up to 5000 rpm while putting out >250 Nm of torque all day long? And is it possible to make it weigh less than 50 kg? The Evo motor would be perfect, but I can't buy another one unless I buy hundreds of them...

 

[Arlo1]

I was just about to make a post asking the same kinds of questions, Arlo.

Basically, the AFM140 is going to stay in the bike, so I am looking for an alternative motor for the CRX. I still want direct drive, so it needs to be pretty torquey. It needs to be a short motor, as it will fit between an IRS differential and the front of the car. I can't make an AF motor with ease, so I have considered making a RF motor, but with a sizeable diameter.

And then I found this thread So what do you think? Is it possible to make a liquid cooled radial flux motor capable of spinning up to 5000 rpm while putting out >250 Nm of torque all day long? And is it possible to make it weigh less than 50 kg? The Evo motor would be perfect, but I can't buy another one unless I buy hundreds of them...

Yes its totally possible. read this You could find a 3 phase and rewind it and oil cool it for some pretty good numbers as a diy project. So starting with a 10 hp and get 100 continuous and 225 peak or maybe a 20 hp and 200 continuous and 450 peak?

 

[speedmd]

Smaller fast spinning spindles geared way down are better suited for race day.

Not sure that is the case in the wet / muddy conditions on several race days I was party to. Give me smooth tractable torque, and lots of it in the wet.

Would the Max rpm's be similar on RF and AF motors and depend on the same type of criteria AF is a most Interesting design. Would love to try one on a bike.

As far as torque curves go on electric motors, it has been stated many times over that a electric motor has maximum torque right off the bottom, in the first revolution. I am not 100% sure this is possible given the controller / current limiting issues. There must be something I am missing in this. Spending most of my life riding with IC motors is most likely confusing the issue a bit for me.

 

[Arlo1]

Smaller fast spinning spindles geared way down are better suited for race day.

Not sure that is the case in the wet / muddy conditions on several race days I was party to. Give me smooth tractable torque, and lots of it in the wet.

Would the Max rpm's be similar on RF and AF motors and depend on the same type of criteria AF is a most Interesting design. Would love to try one on a bike.

As far as torque curves go on electric motors, it has been stated many times over that a electric motor has maximum torque right off the bottom, in the first revolution. I am not 100% sure this is possible given the controller / current limiting issues. There must be something I am missing in this. Spending most of my life riding with IC motors is most likely confusing the issue a bit for me.

When you gear down a high reving motor you have lots of torque!

 

[speedmd]

Hi Arlo 1

Love the burn outs!

Yes, Much More torque when geared down. What I am saying is that with equal torque in two motor/gearbox setups, the one that revs up quicker (steeper power / torque curve with a more sudden spike) vs another that has a smoother (less sudden) torque curve and revs up and power comes on smoother. The one with the smoother curve will be much easier to go fast on in the wet. At least that is my experience.

Very different with electric motors figuring what the true usable power band is. Controller, voltage, FET, rpm issues: eddy current losses, etc.. Lots Lots more to learn.

 

[Arlo1]

Hi Arlo 1

Love the burn outs!

Yes, Much More torque when geared down. What I am saying is that with equal torque in two motor/gearbox setups, the one that revs up quicker (steeper power / torque curve with a more sudden spike) vs another that has a smoother (less sudden) torque curve and revs up and power comes on smoother. The one with the smoother curve will be much easier to go fast on in the wet. At least that is my experience.

Very different with electric motors figuring what the true usable power band is. Controller, voltage, FET, rpm issues: eddy current losses, etc.. Lots Lots more to learn.

Ahhh but.... with electric you have full torque at 0 rpm.

So if you have a hi revving 10hp motor at say 6000rpm (which is 8.75 ft/lbs ) geared down at 4:1 to give 35 ft/lbs at the wheel, and a low revving 10hp motor at say 3000 rpm( which is 17.5 ft/lbs) at 2:1 gives 35 ft/lbs at the wheel Both will rev the wheel at the same speed and both will have the same torque at 0 rpm. Its not the same as a ICE.

 

[speedmd]

Hi Arlo 1

Good topic, and much to learn on the ele side. Forget for a moment gearing. In a ICE, if you have two near identical motors, one with a lightened crank/fly wheel. The one with the lighter crank will rev up much quicker. Expect the same with a electric with less mass. In this case the lighter one will also typically show a bit less low speed torque. Possibly for a host of reasons. Other ways, typically used to increase torque have been to change bore/stroke ratios, valve and ignition timing, compression ratios and even to extend the connecting rods. All proven out over many years of development and I understand this is only part of the power/speed equation.

With a electric motor, does the torque change with the change in rotating mass the same way? I have seen very little real test data on them. Simulations leave out much too much.

 

[major]

With a electric motor, does the torque change with the change in rotating mass the same way?

Yes and no. Basically the electric motor torque is related to the air gap area. For radial flux motors, this air gap is a cylinder between the rotor and stator. You need a certain thickness of iron, copper and magnet on either side of this air gap to make a motor, but this air gap area is the driving factor for torque production.

So, you want more torque? Make the cylinder area larger. That increases mass. However it is the moment of inertia of the rotor which enters into the angular acceleration equation. And that is different from simple mass. It depends on the shape. The shape of the cylinder in particular. So you can minimize the increase in moment of inertia by increasing the length of the cylinder rather than the diameter.

 

[speedmd]

Hi Major

So, you want more torque? Make the cylinder area larger. That increases mass. However it is the moment of inertia of the rotor which enters into the angular acceleration equation. And that is different from simple mass. It depends on the shape. The shape of the cylinder in particular. So you can minimize the increase in moment of inertia by increasing the length of the cylinder rather than the diameter.

Understood. Now with a rotor the same air gap diameter, length, area, magnetic strength, and one with a heavier more massive (Weight more concentrated at the perimeter) rotor, it effects torque how? Certainly will speed up a bit slower due to the moment increase.

 

[major]

Hi Major

So, you want more torque? Make the cylinder area larger. That increases mass. However it is the moment of inertia of the rotor which enters into the angular acceleration equation. And that is different from simple mass. It depends on the shape. The shape of the cylinder in particular. So you can minimize the increase in moment of inertia by increasing the length of the cylinder rather than the diameter.

Understood. Now with a rotor the same air gap diameter, length, area, magnetic strength, and one with a heavier more massive (Weight more concentrated at the perimeter) rotor, it effects torque how? Certainly will speed up a bit slower due to the moment increase.

I'm not sure I follow your question exactly. But the density of material adjacent to the air gap (in both the rotor and stator) affects the flux and current. So I don't see how you can maintain your stated conditions and concentrate weight at the perimeter. But for argument purpose, say you had the choice of two magnets of identical magnetic properties and size but with different mass. The developed torque at the air gap is the same. The shaft torque during changes in speed would differ due to the difference in moment of inertia.

 

[speedmd]

Understood that added material would change magnetic properties somewhat. I see both solid and also hollowed out PM rotors (in the non magnetic structural material) in use, but my question is exactly as you propose the argument.

The shaft torque during changes in speed would differ due to the difference in moment of inertia.

Greater moment would give greater torque as in a ICE? Would rev up slower also?

 

[major]

Understood that added material would change magnetic properties somewhat. I see both solid and also hollowed out PM rotors (in the non magnetic structural material) in use, but my question is exactly as you propose the argument.

The shaft torque during changes in speed would differ due to the difference in moment of inertia.

Greater moment would give greater torque as in a ICE? Would rev up slower also?

The same effect as adding a flywheel, right? So why would you want to do that with an electric motor? It doesn't change the steady state torque output.

 

[speedmd]

I am not suggesting this be done, but have noted that it does add more control in wet /low traction conditions on a ICE motor bike. Important when going over bumps/ shifting gears and the like.

 

[John in CR]

Speedmd,

For 2 motors capable of the same torque I'm pretty sure the issue you're bringing up will be a lot more about controller settings than rotor mass.

 

[speedmd]

Hi John

I did not want to hijack the thread anymore than I already have, but, I could not agree with you more on the controller point. Everyone keeps claiming the electric motor has maximum torque at zero RPM. I am not 100% sure of this without some proof. No way you can supply the draw current without burning up a less than way over sized controller in some very SHORT order. Plus the dynamic torque has not had a chance to add in. I understand that there will be little / no back emf when still but all the motors I have played with come on after they start spinning just a bit. Definitely better than ICE off the bottom, no comparison for certain, but controller is key.

BTW, more rotor mass would smooth things out for certain (even if archaic) if there was no other way to do it. On a ev / ebike with all the fancy controllers, much easier / lighter to smooth using the throttle side.

cheers

 

[John in CR]

The proof is that the BEMF is 0 only at 0 rpm. I just take it for granted since it shows up clearly in the torque curve of our motors. Try playing around with the simulator at Ebikes.ca

 

[speedmd]

The proof is that the BEMF is 0 only at 0 rpm.

Yes, but the coils still have their inductive reactance at a stand still, and what about all the stuttering/ vibration and other crazy noises most generate until they un-stumble themselves. Some things are just missed or ignored in simulations. Some are major. Most dyno traces I have seen ignore the first few rpms.

I agree, it is mainly a controller issue and most likely best long term approach is to get it moving just a bit before slamming it.

cheers

 

[Arlo1]

The proof is that the BEMF is 0 only at 0 rpm.

Yes, but the coils still have their inductive reactance at a stand still, and what about all the stuttering/ vibration and other crazy noises most generate until they un-stumble themselves. Some things are just missed or ignored in simulations. Some are major. Most dyno traces I have seen ignore the first few rpms.

I agree, it is mainly a controller issue and most likely best long term approach is to get it moving just a bit before slamming it.

cheers

I am not sure how long it takes with lebowski's programing. But its <1 sec to full torque and with the right settings I can have the throttle pinned with the brake holding the bike still (at a lot power setting) and the then let off the brake and it will accelerate at full troque from 0-top speed. I can also just pin the throttle and get the same results.
Its dead smooth.