In-runner VS Out-runner .. inherent Advantages/Disadvantages

[LI-ghtcycle]

I am curious, if all things are equal, and there isn't a problem with how the power is transmitted to the wheel, is there any inherent efficiency differences between In-runner and Out-runner motors?

I'm guessing not, and just the way the power is transfered to the wheel determines which would be ideal (hub motors are a perfect example of a clever use of Out-runner motors, and of course, In-runners are great for using a multi stage drive.) if you could use either.

In my case, a In-runner makes a whole lot more sense given that I want a chain running from the motor to my Nuvinci hub, but I haven't heard of too many 10 kV DC in-runners to replace my hub motor as "mid-drive" but I would love to have a more compact In-runner for the job with similar properties to my current 9 x 7 motor.

 

[Alan B]

RC folks discuss this topic. There are a number of differences and tradeoffs. Outrunners tend to have lower KV and are lighter in weight IIRC. Inrunners have higher RPM capability and are easier to cool. Search the RC forums.

 

[spinningmagnets]

My opinion? outrunners have a smaller size for the same power.

I am no expert, but I recently became fascinated by the emerging application of LiPo and outrunners. I have pondered the benefits and drawbacks. My interest in outrunners is based on my belief that they are the most compact motor configuration capable of a given HP output (using the same voltages/amps).

The weight might be slightly better for an outrunner of equal power, but the volume difference is definitely noticeable. My 63mm RC motor is roughly the same size as my coffee cup, and using 24V it is peaking at 2000W (roughly 3-HP). Conventional permanent magnet motors place the coils on the non-rotating shell to allow some ambient heat-shedding to the motor case-fins. Pull one of these apart and you will see that in order to give the magnet-coil interaction any leverage, the diameter of the motor must be increased, leaving some airspace in the center.

Outrunners reverse this, to the detriment of their coil-heat-shedding. But, if properly sized, they run cool enough, and are still smaller per a given power output. The tiny 50mm RC motors have about the same output capability as a Kollmorgan motor (400W @ 24V) of conventional construction (in-runner, finned case), though you would need a reduction to achieve parity.

edit: StudEbiker is about to adapt a 600W BMC motor of conventional in-runner configuration onto a bike. It is a little larger than the 400w Kollmorgan, and it may suit your plans. I believe when using 36V instead of 24V, this may be a good "stoke-monkey" candidate:
http://endless-sphere.com/forums/viewtopic.php?uid=10448&f=28&t=25980

 

[Miles]

edit: StudEbiker is about to adapt a 600W BMC motor of conventional in-runner configuration onto a bike.

Spinning,

I think you'll find that BMC motor is an out-runner....

 

[Miles]

The most significant difference between an outrunner and an inrunner is that, for a given outside diameter, the airgap is at a greater radius in an outrunner (the magnets take up less space than the armature) - so the turning moment, and therefore the torque produced, is greater.

 

[bandaro]

so the turning moment, and therefore the torque produced, is greater.

yeah, that would be my point, because the outrunners have a larger radius from centre to point of power they create far more torque and generally less rpm, and that is exactly what us bikers need. After gearing, the inrunner and outrunner motors both give the same speed and torque, but the inrunner motors will spin much, much faster and therefore require more gearing. less gearing means simplicity and efficiency.

when it comes to hub motors its a no brainer, to have an in runner design you would need brushes, defeating the point of brushless...

 

[liveforphysics]

RC folks discuss this topic. There are a number of differences and tradeoffs. Outrunners tend to have lower KV and are lighter in weight IIRC. Inrunners have higher RPM capability and are easier to cool. Search the RC forums.

Let's perpetuate more of the inrunners cool better myth.

 

[Alan B]

RC folks discuss this topic. There are a number of differences and tradeoffs. Outrunners tend to have lower KV and are lighter in weight IIRC. Inrunners have higher RPM capability and are easier to cool. Search the RC forums.

Let's perpetuate more of the inrunners cool better myth.

Hi Luke. Easier to add water cooling and big cooling fins to an inrunner. At least industrial motors claim that. Haven't seen too many industrial outrunners, but my experience there is fairly limited. We use them to control massive magnet arrays to micron precision.

I understand your point about keeping the magnets cool. But having them in the way of the heat path is not great either. We also never run magnet coils hot enough to endanger the permanent magnets, so it is a different operating regime.

 

[GrayKard]

when it comes to hub motors its a no brainer, to have an in runner design you would need brushes, defeating the point of brushless...

Some hub motors are inrunners. My geared "Cute" motors for example.

 

[John in CR]

Aren't inrunners less flexible in that you can't go much higher than the rated voltage/rpm or their rotor can fall apart? With outrunners your limitation is pretty much the bearings.

Is that BMC really an outrunner? If so I may have to pick one up and liberate it from the housing.

 

[Miles]

edit: StudEbiker is about to adapt a 600W BMC motor of conventional in-runner configuration onto a bike.

Spinning,

I think you'll find that BMC motor is an out-runner....

Here is a thread with some good pics of the 600W BMC internals: http://endless-sphere.com/forums/viewtopic.php?f=16&t=11772&start=0

 

[spinningmagnets]

Thanks for those BMC pics! I know some manufacturers carry both types. I ASSUMED that because of the shell fins on the one originally pictured...that it must be an inrunner. It doesnt seem that those shell-fins would have a good heat path for effective cooling, but...I guess anything is better than nothing.

I work around a lot of conventional motors that have the stator on the outside so the shell-fins give good ambient cooling with no moving parts. Giant heavy things meant to be bolted to the ground, and running 24/7.

 

[John in CR]

Thanks Miles, I missed that one.

With thermal issues being the primary limitation, it baffles me why these manufacturers insist on sealing these motors so tightly. Haven't they ever seen an alternator? Outrunners are especially easy to ventilate, and in the case of that BMC the power handling could easily be double or tripled. As-is the waste heat is transferred out primarily through the magnets. eg I have little doubt that the small change I'm looking at on my Turnigy outrunner will easily push 100cfm through the motor at just 2-3krpm even through the restriction of a filter, and that's low-balling the potential in the interest of noise.

 

[Wheazel]

There is no secret that the most efficient RC motors generally are inrunners. This is much due to tradition, outrunners are coming.
There are great outrunners out there already, even if they are few. But due to an outrunners many advantages to system efficiency for an ebike, its often the prefered choice.

 

[liveforphysics]

There is no secret that the most efficient motors are inrunners.

Holy chit.

Where do you guys get these things?

Inunners are more efficient? Inrunners cool better?

It's like saying inline 6's are more efficient than v6's or something equally foolish.

 

[Wheazel]

There is no secret that the most efficient motors are inrunners.

Holy chit.

Where do you guys get these things?

Inunners are more efficient? Inrunners cool better?

It's like saying inline 6's are more efficient than v6's or something equally foolish.

Why would you call it off like that? If you look at permanent magnet brushless motors that are used for these bikes (in the nonhub motor drives forum)
its heavily leaned towards RC motors.
Among these, there are loads of efficient inrunners of good quality, and very few outrunners that match or even comes close in efficiency.
Most high end rc models use inrunners for competition. Pylonracing, powergliders etc. Why? cause they are simply the best motors for it. (in some cases spaceresticted, but most often not)
Even big scale electric aerobatic models used geared inrunners for a long time until outrunners of decent quality appeared, now its all about outrunners.

If you want to look at motors that doesnt apply advantages or disadvantages to en electric bike, yes there are super efficient outrunners.
Any motor of decent engineering can be efficient. For example the ones used in the solar challange cars. Outrunner wheel motors.

With that said I think that my previous statement should have been more specific.
I can make a long list of ebike-compatible efficient inrunners(if one would overcome the gearing dilemma), but the list will not be nearly as long for outrunners.
Chinamotors not considered, as they simply aren't of the quality and efficiency seen in for example plettenberg outrunners. (One of the few that can compete)

Answers can be interpret as one like I assume, just as the bible, holy chit.

 

[liveforphysics]

Wheazel- It's as simple as this. One has the magnets on the inside of the flux gap and the iron on the outside, one has the magnets on the outside and the iron on the inside. That's it.

The efficiency will be determined by design of the lam teeth, flux gap and flux path containment, end-turn ratios, copper fill, lam thickness, avoiding excessive back-iron hysteresis losses, and if you're going for broke, electrically isolated magnets (to avoid magnet/tooth eddy losses) made into halbach arrays, slot-filled winds (and/or silver windings) etc.

At no point does motor efficiency care if the magnets are on the inside or the outside.

If you have them on the outside, since the magnets are thinner for a given flux value than the teeth/copper, you can have a higher sustained torque value for a given motor OD package limitation. That's the only difference.

 

[liveforphysics]

If you make a good inrunner design, you have a good inrunner.

If you make a good outrunner design, you have a good outrunner.


If you make a crappy inrunner design, you have a crappy inrunner.

If you make a crappy outrunner design, you have a crappy outrunner.




Now, for our application, which is a system perspective, when you lose around 1.5-3% on each gearing stage, with steeper reductions having higher chordal pitch losses for a given space limiation (from needing low-tooth-count gears or sprockets), an inrunner of the same power level typically needs an additional stage of reduction due to the trade off it makes between torque and RPM to generate that same power level as an equal power outrunner (as a generalization). This makes outrunners typically most efficient, and in an un-geared application like a hubmotor or something, it makes them massively more efficient of course due to highest magnet-tooth speeds for a given vehicle speed (the efficiency of ANY motor at 0 magnet/tooth speed is 0%).

This makes an outrunner of equal design/build quality almost always end up being a higher efficiency system, but over all the difference is a mute point compared to the reduction in drive complexity and noise of additional high RPM gearing stages.

Also, even the low quality el-cheapo hobbyking 80-100 130kv outrunner gets to something like ~94-95% efficiency on ~80v systems. You can rewind it yourself for about $10 and a day of labor and gain an extra percent or so, not only matching the best inrunners (and at a more system friendly RPM), but making the margin of difference between the pletty's and a $100 motor a matter of bearing quality and magnet adhesive etc.

 

[Wheazel]

I think the solar car motor example I made underlines your reply, yes the efficiency depends on engineering not some magical layout.
Even if a HK outrunner might be at 94% efficiency (will believe it when i see it) at 80v, its quite irrelevant.
WHat matters is what a motor performs under the conditions it will be used.
These motors will be used with 10s and 12s by the vast majority. I think its a clear trend that higher voltage systems often incorporate a hubmotor.
Maybe because the hall sensor hassle with rc motors and hall sensor controllers.

From a RC enthusiast for 20years perspective, inrunners are still used in many applications because there are no outrunners that can match them.
This has probably more to do with tradition than anything else, still it is the reality.
Take Lehner motors or Neu motors, two high end motor brands. Still only inrunners in the larger sizes. Biggest being a 1650g beast.

I was not intending to step on your physics toes for any reason, I will edit my first post so you find it less insulting.

As you mentioned lamthickness I remembered a question I have had for a long time. What lamthickness is optimal for efficiency on a sub 50v motor?
I have built a small amount of motors(outrunners) with 0.2mm laminates, which is told to be a good compromise. What does your knowledge say? What is the turningpoint? (Sorry to author for threadhijack)

 

[recumpence]

OK, I think I will chime in here. Bear in mind, I am not an electrical engineer. However, I do have alot of experience with both inrunners and outrunners in RC and bike use. These are my findings from my own experimentation;

My Plettenberg Terminator is listed (or was when I bought it) at 80% efficient. They claim 7kw continuous. This is a $1,100 motor. I am sure it is actually more efficient than that number. But, that is what it was listed at. Even with the fan is has, it runs very warm at anything above 2kw (almost too warm to touch). My Astro Flight inrunners are claimed to be 95.4% efficient (I have the dyno test charts from Bob for a few motors he has build and tested). That efficiency is sustained over a relatively high RPM range.

It seems, as Luke eluded to, that the outrunners have higher torque and are happier at a lower RPM than inrunners of the same diameter. That is my personal experience.

Now, as for my personal choice of Astro Flight inrunners for the majority of my builds, that has nothing to do with their inrunning design as much as their overall quality, performance, efficiency, etc. Also, a spinning can presents its own problems of installation that are avoided with an inrunner. Inrunners are convenient, and easier to mount. Also, in my case, I make my own reduction units. So, high RPM is not an issue. Also, I personally build and sell alot of drives for customers. The reliability of the Astro motors is something I need when selling to the public. I cannot risk the cheapo china motors that have such sketchy reliability (this may improve, however).

The negative of high-end inrunners (Astro, Neu, etc) is cost. Of course, my Plettenberg is, by far, the most costly motor I have ever run and it is an outrunner. Anyway, I have been talking to Astro Flight about the price of their motors and the possibility of reducing it to a more reasonable level. The issue with the 32 series motors is the fact that they are mostly hand made, individually configured, and individually tested before release. This costs alot. Most motors that are inexpensive are machine made and, therefore, much cheaper.

To a certain extent, this is like the whole "Ford versus Chevy" debate. There are good points on both sides. The fact of the matter is, people will buy and run what they perceive will work best for their particular application, period.

Matt

 

[Wheazel]

I agree on the astro, in my book it probably is the best nonhub motor for ebikes.

I am abit surprised about your pletti findings tho, thought they would produce less heat than that. (at those powerlevels)
Got any direct comparisions between the pletti and the astro?

 

[johnrobholmes]

From my limited understandings, taking two motors of identical frame size and construction, the outrunner will tend to be less efficient because of the larger air gap as % of the motor cross section. Sure this can be improved upon, and many other design elements can offset this disadvantage. Talking with the head engineers at the 3 factories I use for my R/C motor business, they all agree that an inrunner can be built to higher efficiency levels holding other design and construction aspects constant. 1mm airgap on an inrunner will be less volume than a 1mm airgap on an outrunner, simply speaking.

I have six electrical engineers that disagree with you Luke, including the owners/ engineers of Nuetronics, Castle creations, and Bob Astro. Maybe one day you can have your own motor production facility and prove them wrong!

 

[liveforphysics]

John-

The goal for any motor is to have as much active flux area (diameter and length) as you can fit in a given motor size constraint with out compromising the magnetics path (this doesnt mean the gap distance from tooth to magnet should be large, just flux area). It is never a bad thing to have more flux area, as it means to create a given torque, you can have lower flux intensities, and it carries ZERO penalty.

Look at the most efficient motors in the world. The CISRO motor for example with 98.8% efficiency, and it's got a flux-gap to motor volume ratio about 4x higher than most inrunners, and 2x higher than most outrunners. Same with the 8hp continuous 1.4lbs and 97% efficient launchpoint motor. Getting a boat load of flux area for a given motor volume is kinda the whole idea of being able to make more torque for a given motor volume, as the only other option is to use more exotic lam materials with higher flux saturation levels (like Honda did with the new alloy it created for its KERS motor).

Possible torque will saturate at a value roughly equal to the flux area of the motor times the lamination alloy's flux saturation values (with things like annealed iron cobalt alloys having highest saturation numbers). But this doesn't mean the same lam materials can't be used for outrunners or inrunners.

If you don't have a lot of flux area, it doesn't mean you will be down on power from a given amount of active motor materials, it simply means since your torque limit (the point at which that flux gap saturates) will inherently be lower (for a given quality of magnetic materials used), so to make up the power you have to do it with higher RPMs.





John- To further show how silly it is to say one design is more efficient or less efficient than another, tell me what happens if you took your favorite inrunner, chopped up the magnetically active parts and swapped which sides they are on. Would the motor perform any differently?


Think about a motor as if you could lay the circular device out in a big flat line making it a linear motor(like the track of a mag-lev train or something). At that point, to say inrunner is better than outrunner is to say when the track is right side up or upside down it performs better or worse. Obviously it doesn't matter if the track is upside down or right side up any more than it matters in a motor for which side of the flux area is holding the magnet and which is holding the copper/iron side.



Wheazel- Your lamination thickness isn't a function of voltage or anything, it's a function of the rate at which flux lines bisect the lams and the conductivity of the material in the lams (how much Si they dope in the iron to make it conduct as poorly as possible). Each time a flux line passes through the lam, by induction is has to induce a voltage in the lam, so the goal is to create as little path as possible to develop large voltages that result in large eddy currents (which are realized just as parasitic thermal gain). To say 0.2mm is thin enough or un-neededly thin depends on this rate the flux lines are cutting through it in your application (which of course depends on stator radius and motor rpm and the number of poles etc). The real no theoretical BS way to check if your lams are too thin or not is to simply spin the motor up before it's wound, and plot the torque vs RPM curve of spinning the rotor past the flux gap. It will eventually start to look like an x^2 function (climbing rapidly in parasitic torque for a given RPM increase). The goal is to make the operating range of the motor stay in that very low section before the curve takes off, and that means you've got thin enough laminations.

 

[johnrobholmes]

John- To further show how silly it is to say one design is more efficient or less efficient than another, tell me what happens if you took your favorite inrunner, chopped up the magnetically active parts and swapped which sides they are on. Would the motor perform any differently?

You say this, but first you state a few axial flux motors that are clearly superior to radial flux motors in every way. Efficiency, torque density, power to weight, flux area...

If design didn't matter, then why are none of the largest motor manufacturers getting as high of efficiencies with outrunners as compared to the inrunner lines? I visited Castle a few months ago and got to see some of the new outrunner designs, and they commented that the max efficiency and power density was still below the inrunners after three years of R&D. 0.2mm laminations on both, same magnet strength, same physical sizes and weights, for all intents and purposes as close to identical construction as possible. The outrunners made more torque per RPM, the inrunners spun faster to make up the difference. Basic stuff.


If I chopped up an inrunner and swapped sides, we would either have a stator the same size as the original rotor with the same airgap area (and smaller overall diameter), or we would have a totally different motor with a larger stator, flux area, and airgap area. Too many changes to theorize on anything, only testing or extensive modeling could really show us how they stacked up.


Everything I have read and every motor designer I have talked to preaches about air gap reduction for best performance. To quote Mohammad Modaras' article "Study on Axial Flux Hysteresis Motors Considering Airgap Variation" from April 2010:

Furthermore all simulation results show that smaller air gap (g) how much increases machine efficiency, Power factor, maximum flux density, torque and stator winding turns and reduces the terminal current and hysteresis ring thickness for constant load. So by considering the mechanical constrains, the air gap is assumed minimum possible value.

What gets me with the outrunner design is the lack of rigidity in the rotating can, especially considering longer or larger diameter motors needed to produce suitable torque and speed for ebikes. A ring bearing must be used, or else the airgap must be large to reduce the likelyhood of a rotor crash. Neither are good design aspects when considering a perfect motor. Neither are design aspects that the inrunner faces until the rotor becomes long and skinny enough to flex like a bow.


I'm not saying that you are wrong about the theory, but in the real world I just haven't seen a radial flux outrunner beat a radial flux inrunner on the dyno. In application, I can say that a well designed outrunner could beat the inrunner if a reduction stage can be omitted or if reduction is not possible.


What would I love to see? A 6" axial flux motor with a planetary reduction integrated. Spin that sucker to the moon and get gobs of power in a lightweight and thin package. Too bad launchpoint hasn't gotten money invested yet.

 

[liveforphysics]

I found the problem in what you're hearing and reading John, and what I'm trying to tell you.

Everything I have read and every motor designer I have talked to preaches about air gap reduction for best performance. To quote Mohammad Modaras' article "Study on Axial Flux Hysteresis Motors Considering Airgap Variation" from April 2010:

Furthermore all simulation results show that smaller air gap (g) how much increases machine efficiency, Power factor, maximum flux density, torque and stator winding turns and reduces the terminal current and hysteresis ring thickness for constant load. So by considering the mechanical constrains, the air gap is assumed minimum possible value.

When they say the flux gap decreasing increases torque and efficiency etc, they mean the distance between the magnet surface and the stator tooth. This is a no brainer.

What I'm talking about is the flux area, meaning the length x circumference of the area encompassing the flux gap. This is completely irrelevant to the flux gap distance itself (like what that quote is talking about).



Maybe it will help if you don't even think about it as a spinning device, but break it down to it's base function, the magnet and tooth interaction. You have some quantity of flux in the magnet and in the tooth pushing against each other to provide a force. The radius this force acts upon determines the torque (because any amount of force, no matter how small, can become any amount of torque, it's just dependent on the radius). If you've got some given area with x lines of flux pushing on it. It doesn't matter what made these lines of flux, if it's a permanent magnet on each side, a copper coil on each side, the end of an iron tooth with a coil 5ft away on the end of it. It doesn't matter what makes those lines of flux (or what side it's on), they all result in the same amount of force if the area and flux lines are the same. Then this force is converted to a torque when you give it a radius, the amount of torque changing directly with the amount of radius. You can see that as a single tooth/magnet interaction, it makes no difference which side it's on. A motor is just a cluster of these same tooth/magnet interactions arranged in a circular group, and once again, it makes no difference which side they are on (inrunner or outrunner).

The only difference is that an outrunner can have a larger amount of flux area (note: this is UNRELATED to flux gap distance) for a given amount of motor diameter. This enables a greater torque for a given amount of flux intensity generated.