Conclusive proof gearboxes are awesome.

[DanGT86]

Makes perfect sense about some of the losses having a linear relationship with torque and speed and some being constant. Thanks.

That got me thinking about chains. My 219 chain feels really dense. It got me thinking about how when in motion around 2 sprockets that is a lot of weight that is being constantly flung in different directions. As it makes its way around a sprocket it is being effectively accelerated the whole time. Around a small sprocket it basically gets thrown one way then instantly makes a turn and comes back the other way. At first glance it seems like tons of wasted energy. The smaller the sprocket the closer to reciprocating mass it becomes...Maybe?

But after further thought I reasoned that because its a loop there is always another part of the chain effectively cancelling out that motion. For example the chain mass reversing direction as it passes the apex of the sprocket is being pulled by the inertia of the chain that already made that 180. And its being effectively pushed by the links that havent made the turn yet..... Then my head exploded.

So long story/question short: does the inertia from the chains mass pretty much cancel out between the load side and the slack side? If you have a bunch of directional changes like many sprockets and idlers do they all cancel just reducing to bearing and frictional drag? Does the physical path of the chain in space require different amounts of power to rotate?

 

[Hillhater]

This is an available today PM motor from a LEAF (not something exotic or ironless etc).

Show me what operating conditions it's going to be worth the gearbox's mass, cost and added friction. Even though no motors today are 100%, many are already so close to 100% efficient over such a wide envelope, it's tricky to see the area at which the transmission could possibly justify its own losses over simply selecting a logical motor choice.

Call me paranoid, but....
Wouldnt the logical motor choice have a much lower optimum rpm range ?
That graphic shows how much the Leaf needs a 10:1 reduction between the motor and the wheels ?
Wheels speed is rarely above 1000rpm and most shopping cars need around 1000+ NM of torque at the wheels .
So the leaf still has to have the weight & efficiency issues associated with a transmission.
Couldnt they at least have designed the motor for lower optimum rpm , higher torque, and use less reduction (3:1, 2:1) ? ...or again, is there some obvious reason they did not do that ?

 

[liveforphysics]

This is an available today PM motor from a LEAF (not something exotic or ironless etc).

Show me what operating conditions it's going to be worth the gearbox's mass, cost and added friction. Even though no motors today are 100%, many are already so close to 100% efficient over such a wide envelope, it's tricky to see the area at which the transmission could possibly justify its own losses over simply selecting a logical motor choice.

Call me paranoid, but....
Wouldnt the logical motor choice have a much lower optimum rpm range ?
That graphic shows how much the Leaf needs a 10:1 reduction between the motor and the wheels ?
Wheels speed is rarely above 1000rpm and most shopping cars need around 1000+ NM of torque at the wheels .
So the leaf still has to have the weight & efficiency issues associated with a transmission.
Couldnt they at least have designed the motor for lower optimum rpm , higher torque, and use less reduction (3:1, 2:1) ? ...or again, is there some obvious reason they did not do that ?

You would need to ask Nissan why. Ironically at around the same time period they designed the LEAF motor I also was still into high RPM's for the sake of high RPM's myself. It's not like there is a tight packaging requirement under the hood of the LEAF where they can only have a 6in diameter rotor or whatever it is they use, so certainly they could shift motor geometry around to eliminate the gearing weight/loss/cost and have a cheaper longer lasting more efficient drive solution.

The same question could be asked of Toyota and the newest gen prius engine that has 40% thermal efficiency today, when aluminum and steel bits were equally free to be shaped that way for the last half century Toyota has been trying to make fuel efficient vehicles and never broke the low-mid 30's for thermal conversion efficiency. Or Honda having many decades of 20-30mpg accords, then evolving the design into something that skips the gearbox and gets 50mpg while maintaining excellent performance.

For design evolution to occur, it requires someone who believes it's possible being willing to put the RnD work into it. Until that happens we have LEAF's with multiple stages of spur gears taking that beautiful motor efficiency curve and adding various types of unnecessary losses and cost and failure modes to the system.

EV's will evolve at a rate fueled by sharing ideas (like this thread) and leveraging the awesome tech of the abundant FEA simulations to find optimizations for each individual applications needs more rapidly than we've ever seen through the history of technology evolving.

ATB,
-Luke

 

[liveforphysics]

Makes perfect sense about some of the losses having a linear relationship with torque and speed and some being constant. Thanks.

That got me thinking about chains. My 219 chain feels really dense. It got me thinking about how when in motion around 2 sprockets that is a lot of weight that is being constantly flung in different directions. As it makes its way around a sprocket it is being effectively accelerated the whole time. Around a small sprocket it basically gets thrown one way then instantly makes a turn and comes back the other way. At first glance it seems like tons of wasted energy. The smaller the sprocket the closer to reciprocating mass it becomes...Maybe?

But after further thought I reasoned that because its a loop there is always another part of the chain effectively cancelling out that motion. For example the chain mass reversing direction as it passes the apex of the sprocket is being pulled by the inertia of the chain that already made that 180. And its being effectively pushed by the links that havent made the turn yet..... Then my head exploded.

So long story/question short: does the inertia from the chains mass pretty much cancel out between the load side and the slack side? If you have a bunch of directional changes like many sprockets and idlers do they all cancel just reducing to bearing and frictional drag? Does the physical path of the chain in space require different amounts of power to rotate?

Found this white paper on chain efficiency, but it's behind a paywall sadly.
http://link.springer.com/article/10.1007%2FBF02915918

Abstract
This paper investigates the optimisation of the chain drive system on sports motorcycles. Recently there has been the development of a chain transmission efficiency model that is suitable for motorcycles. The new model is used to predict the efficiency of 600cc sports motorcycle at different speeds. The transmission efficiency is estimated to be between 96 and 99% for speeds less than 75 mile/h. Between 75 and 150 mile/h the transmission efficiency can be as low as 85% due to inertial tension. The transmission efficiency model presented in this paper enables optimisation of sprocket and chain sizes. In general, large sprockets are better at low speeds and smaller sprockets are better at high speeds. The optimum chain size is the chain with the smallest pitch that can meet the torque and power requirement. The sprocket centre distance also has a big effect on efficiency and it is important to use an effective installation procedure. In particular, it is important to set a chain up when the rear wheel axle, front crank and swing arm bearing are all in-line.

Essentially, once you start to spin them fast (IE, faster than human bicycle pedaling speeds), chains suck power and still suck power even while transmitting a very light load due to inertial loading (essentially the frictional loading of forcing the chain that wants to be a circle to be an oval going around sprockets.) For low speeds at high loading chains rock (like a bicycle). For high speeds and low loading chains have far more losses from there own inertia induced friction than belts or gears.

This is why it's tricky to say a chain is or isn't efficient, because there is some operating perimeter for any chain that makes it look great, and some operating perimeter that makes it look awful. I was always amazed how hot my GSX-R1000's chain would be after high speed riding, considering while operating it has triple digit wind speed cooling it.

 

[DanGT86]

That concept of the chain being an oval when it wants to be a circle is exactly what I was talking about. We have a pretty active stunting comunity in St. Louis and I frequently get passed by groups of sport bikes on the highway that are geared primarily for wheelies. Always figured the chains were being stressed when they were running giant rear sprockets yet still insisting on 100-150mph speeds. That is a lot of chain speed but probably still within the design limits of the chain cosidering how many motorcycles flirt with 200mph from the factory.

 

[nutnspecial]

That's a very fair observation into chains imo, and heat from friction is the obvious evidence of energy loss. If it can be traded within desired use range for a better design, 'more power to ya' (literally lol, or 'more energy').

My attempted point is more broad. It seems to me the first obvious answer to gearing desirability argument is the case of *battery limited*, specifically in mentioned EV's/hybrids above. If this wasn't such an issue, there would be less benefits for hybrids or naked IC, considering the superiorities of modern electric motors. I find it important to consider the abundance, ease of refinement, and recyclability (ie the ultimate definition of 'cost') of elements going into power sources vs gearing designs that conserve said energy under intended goals.

Striving for balance in use of gearing over additional copper/lithium might be potentially more eco-sound/ less 'costly', but also has many already hopeful obvious physical benefits VS motor size and placement restrictions, perhaps most largely appicable to vehicles.

I love many aspects of 'E'. In electric motors for energy transformation in biological/steam/wind/hydro or from bouancy/density (sidenote: could the all aforementioned 4 be described more plainly in terms of bouancy and density??), and used for brushing teeth to lighting and communication to powertools to vehicles (and even electroshock therapy which I've been recommended ). So E also now gives the ability to get excellent torque across ~7hp in ~30lbs on my bicycle, with usable albeit limited range per charge. IC has virtually no competition there, especially when 'silence is golden'. E for the win.

Also no competition on my bike: a DD or geared hub. Even if more efficient DD could reasonably compete across my power and application preference, the motor would be much larger and/or my wheel much smaller, and would likely also go 'in the wheel'. Regardless if more battery would also be required, in this case single reduction > variable gear or DD.
Obviously the preferential benefit could be found for serial mid or DD under different circumstances, and that's exactly the point here some might avoid. Imo this example evidences remaining issues with scaling and general physicality of motors/ limited range of design application. Then of course there's still that power source to consider.

In closing, with hopefully little risk of sounding like a fanboy, but perhaps an asshole nonetheless:
Motors and batteries are not sufficiently advanced to rule out gearing's positive inherent attributes. Find a more efficient motor and power source under unity, and find greater efficiency @ greater application range and/or eliminate physical design limitations with various physical gearing- it's what gearing does. As I alluded to eariler, perhaps when basic freaking physics like the lever and the wheel are obsolete we can finally say byebye hahaha.
I challenge myself with this post, and also challenge and would highly prize the breakdown of it for any contrary argument by reader.

Ps.
Thankyou alan for mentioning "The Feynman Lectures on Physics"! It would seem my quandary at that aspect is not so far misplaced, and my range of study likely included some sources well aware of specifics contained therein. I'll be sure to check it out more specifically; the universal (literally) mystery of nature at quantum level and from photosynthesis to magnetics is considerable.

 

[alan]

Honda already figured out a clever way to skip the very meaningful low power cruising losses. The result is a huge and heavy full size accord that has better performance and economy than the tiny engine lightweight CRX.

This Honda hybrid is not a good comparison for this discussion. Honda started with a purely parallel hybrid system that require the engine to be turning 1:1 with the single electric motor all the time (the motor replaced the flywheel on the engine for a hybrid system that had great value on the highway scenario, but not so great in start and stop city traffic. This is the system they used on the first hybrid Insight through the CRX. For the latest Accord hybrid (not the first Accord hybrid which was a parallel system), they went to the parallel-serial hybrid model pioneered by Toyota and Ford which combined two electric motors with an ICE through a planetary gearset. This allowed far greater control for keeping the ICE in its most efficient operating window by effectively having an infinitely variable tranny whose ICE to wheel ratio was a function of the speed ratio between the two motors, and also easily allowed for pure EV mode.

Just a point of reference, the BMW i3 REX is the first (that I know of) purely serial hybrid where the ICE is only used to drive a generator to provide current to the motor/battery.

 

[John in CR]

Just a point of reference, the BMW i3 REX is the first (that I know of) purely serial hybrid where the ICE is only used to drive a generator to provide current to the motor/battery.

Diesel/electric locomotives have had it for decades.

 

[Miles]

Not to mention the Lohner Porsche....116 years ago.... et al...

 

[speedmd]

GE patented the diesel electric single lever control in 1914 to put in perspective.

Luke is spot on, on this development topic. If we want it bad enough, we need to make it happen. We know the simple way to get the low speed torque is simply increasing the diameter. Not a issue on a bike as we already have relatively large diameter wheels. Weight is a issue. Permanent magnets and Iron cause losses as speed increases killing top speed - efficiency while coasting on slight down hill grades at higher speeds. A suitable large diameter induction design best possibly for max efficiency and broad speed range?

 

[Miles]

I'm halfway through this book https://books.google.co.uk/books/about/The_Electric_Vehicle.html?id=5689qa8vrLQC Not an easy read but fascinating....

 

[liveforphysics]

Nutspecial- Just a minor misconception to clear up you may still be having. You mention the battery often.

From the perspective of a battery, the most efficient drivetrain will always put the least load on the battery for a given power output (or for any desired torque at any desired speed if you want to think of it that way).

This may be counter intuitive to some, but say you have an EV with a multispeed transmission. When you're in 1st gear starting on a hill or whatever, it enables you to discharge from the pack harder than 2nd gear could enable, and it's in this additional power drawn from the battery due to higher motor BEMF rise rate from higher motor shaft speed to vehicle speed.

I was playing around driving my friends Zero motor powered Insight a couple days ago and noticed even though the performance off the line sucked, just leaving the car in 3rd was definitely the most watt-sipping way to get around vs rowing through the shorter gears first, which would enable drawing around twice as much power from the battery (and hence also accelerating about twice as hard).

 

[liveforphysics]

Honda already figured out a clever way to skip the very meaningful low power cruising losses. The result is a huge and heavy full size accord that has better performance and economy than the tiny engine lightweight CRX.

This Honda hybrid is not a good comparison for this discussion. Honda started with a purely parallel hybrid system that require the engine to be turning 1:1 with the single electric motor all the time (the motor replaced the flywheel on the engine for a hybrid system that had great value on the highway scenario, but not so great in start and stop city traffic. This is the system they used on the first hybrid Insight through the CRX. For the latest Accord hybrid (not the first Accord hybrid which was a parallel system), they went to the parallel-serial hybrid model pioneered by Toyota and Ford which combined two electric motors with an ICE through a planetary gearset. This allowed far greater control for keeping the ICE in its most efficient operating window by effectively having an infinitely variable tranny whose ICE to wheel ratio was a function of the speed ratio between the two motors, and also easily allowed for pure EV mode.

Just a point of reference, the BMW i3 REX is the first (that I know of) purely serial hybrid where the ICE is only used to drive a generator to provide current to the motor/battery.

When the new accord is in the highway cruising getting its amazing highway fuel economy, it's hybrid system is entirely shut down, and the economy comes from a dry clutch directly coupling the engines output to drive the wheels without eating the losses of a box full of a dozen oil churning spinning toothed disks.

When its in the city, you're right that it's got a clever hybrid motor setup, but it's city mileage isn't remarkable vs it's amazing highway economy.

.• EV mode: The car is propelled only by an electric motor powered by the lithium ion battery pack. The top speed is about 60 mph, but the range is very limited, just a few miles. Once the battery is depleted, the car automatically switches to Hybrid mode.

• Hybrid mode: The Accord Hybrid works a lot like a Chevrolet Volt. The 2.0-liter gasoline engine powers a generator to create electricity for the electric motor. Honda calls the generator a motor, hence two motors. Only the electric motor, fed by the generator, moves the car in Hybrid mode.

• Engine mode: The electric motor is decoupled, and the Accord's Atkinson cycle gasoline engine drives the wheels directly through a one-gear transmission. In that highway gear -- similar in efficiency to a sixth gear on a manual transmission -- the gasoline engine runs in its sweet spot, delivering both strong acceleration when needed and high fuel economy.

 

[Rassy]

alan wrote:

This is the system they used on the first hybrid Insight through the CRX

Just a picky little detail, but I think that should have been CRZ instead of CRX.

I owned a couple of CRX HF's and they were terrific little cars that delivered 55 MPG highway with no electric hybrid equipment.

I also owned a CRZ, which seemed a real waste of lots of great technology and couldn't even reach 50 MPG highway.

 

[Hillhater]

Seems like tha Honda Accord uses similar driveline tech to the Koenigsegg Rugera... No gearbox ICE drivetrain...
https://www.carthrottle.com/post/engineering-explained-how-the-koenigsegg-regera-hypercar-drives-without-a-gearbox/

 

[Hillhater]

Deleted....double post..see next post.

 

[Hillhater]

Hillhater I think it is supposed to mean the amount of power the motor can produce per given amount of weight will stay the same so when you are making the motor bigger it is getting heavier but it is able to run more continuous power. ......

Sorry Arlo ( and others) i missread this comment last week.
Somehow i though the discussion was suggesting you could have a 8 times more powerful motor which weighed little more than the smaller motor
..but if this is the case, that motor power is basicly a function of weight, then to me that explains exactly why EV designers all seem to stick with compact , high rpm, motors , and a reduction drive.
There would be little point in Tesla / Nissan using a larger motor with 8 times the torque , if it meant the motor was 8 times heavier !...that would be orders of magnitude greater than a reduction box weighs !
( and yes, i understand in the real world the power/ weight ratios are not that simplistic)

PS.. The nearest estimation of the transmission ( reduction gearing and differential) of the Tesla is about 175 lbs with the motor at 350 lbs. http://www.teslarati.com/tesla-model-s-weight/
So, it would have to be a very slick motor redesign to increase the motor Torque 8+ times , with only a 50% increase in motor weight .

 

[liveforphysics]

Hillhater I think it is supposed to mean the amount of power the motor can produce per given amount of weight will stay the same so when you are making the motor bigger it is getting heavier but it is able to run more continuous power. ......

Sorry Arlo ( and others) i missread this comment last week.
Somehow i though the discussion was suggesting you could have a 8 times more powerful motor which weighed little more than the smaller motor
..but if this is the case, that motor power is basicly a function of weight, then to me that explains exactly why EV designers all seem to stick with compact , high rpm, motors , and a reduction drive.
There would be little point in Tesla / Nissan using a larger motor with 8 times the torque , if it meant the motor was 8 times heavier !...that would be orders of magnitude greater than a reduction box weighs !
( and yes, i understand in the real world the power/ weight ratios are not that simplistic)

The motor doesn't need to weigh anymore if the materials are utilized the same way as the small motor.

Hence, if you have 1kg of iron in a small stator getting used at 600Hz in an RC motor in the shape of a long skinny 4pole inrunner, or in a huge pancake ring outrunner with 100poles turning so slow you can see it, provided the mass and current density through the copper is the same, and the utilization of the magnetic materials is the same, the power density will have the potential to be the same as well.

 

[Chalo]

The motor doesn't need to weigh anymore if the materials are utilized the same way as the small motor.

Electrically speaking, you're right. But structurally, it takes a lot more beef to maintain rigidity and a uniform air gap in a large diameter motor. I have imagined a solution to this problem that would mount the ring of coils and back iron on a tension spoked framework like a spoked bicycle wheel inside a lenticular disc wheel (which contains the magnets).

 

[Hillhater]

?....if you have 1kg of iron in a small stator getting used at 600Hz in an RC motor in the shape of a long skinny 4pole inrunner, or in a huge pancake ring outrunner with 100poles turning so slow you can see it, provided the mass and current density through the copper is the same, and the utilization of the magnetic materials is the same, the power density will have the potential to be the same as well.

That implies you could make the Tesla motor ( for example) 3 times the diameter, but 1/3 the axial length, and retain the power output. But does that also mean it could be made to operate at 1/8 the rpms and 8 times the torque ?

Even if that is possible, weight is proportional to volume , and a motor 3 times the diameter and 1/3 the length will have 3 times the volume of its long thin brother.....and consequent weight penalty, unless some ingeneous construction methods are used.

Chalo, with Tesla, we are talking induction motors which would imply a lot of solid metal and cooling liquid filled spaces etc

 

[speedmd]

Roughly 3 times the diameter and 3 times the volume or weight (if no additional weight reduction changes incorporated) for 8 times the torque and 1/8 the rpm. Side benefit would be elimination of a major reliability issue. Lower pack voltages? Looks to be good trade offs.

 

[Chalo]

Chalo, with Tesla, we are talking induction motors which would imply a lot of solid metal and cooling liquid filled spaces etc

That in turn would imply a large hole through the middle of the motor with the "shaft" taking the form of a ring or hollow cone, and thin section large diameter bearings representing one of the technical challenges.

I think the limitations of slip rings for the rotating armature are every bit as serious as the limitations of permanent magnets.

 

[Arlo1]

Chalo, with Tesla, we are talking induction motors which would imply a lot of solid metal and cooling liquid filled spaces etc

That in turn would imply a large hole through the middle of the motor with the "shaft" taking the form of a ring or hollow cone, and thin section large diameter bearings representing one of the technical challenges.

I think the limitations of slip rings for the rotating armature are every bit as serious as the limitations of permanent magnets.

No slip rings in a Tesla motor... It has a stator that is more or less the same as a BLDC motor then a rotor with copper bars and end pieces to make a conductor.
I suggest goggling how it works. But the simple way to explain it is the field in the stator spins around the rotor just like a BLDC or PMAC (innerrunner) but the rotor does not spin the same speed. And the torque comes from a bit of slip... The rotor will spin slower then the stator field for accel and faster then the stator field for decal (regen).

 

[Hillhater]

Roughly 3 times the diameter and 3 times the volume or weight (if no additional weight reduction changes incorporated) for 8 times the torque and 1/8 the rpm. Side benefit would be elimination of a major reliability issue. Lower pack voltages? Looks to be good trade offs.

Im not the expert, but i suspect if you want to maintain the same total motor power, you would need to keep the same battery capacity ( volts x amphrs) . So unless you upgrade to more amphrs, you would have to retain the voltage.
Also, im not sure adding 500-600 lbs of motor weight would be considered a good trade for the reduction gears.
That would have to hurt efficiency, range, performance and handling etc.

 

[liveforphysics]

Roughly 3 times the diameter and 3 times the volume or weight (if no additional weight reduction changes incorporated) for 8 times the torque and 1/8 the rpm. Side benefit would be elimination of a major reliability issue. Lower pack voltages? Looks to be good trade offs.

Im not the expert, but i suspect if you want to maintain the same total motor power, you would need to keep the same battery capacity ( volts x amphrs) . So unless you upgrade to more amphrs, you would have to retain the voltage.
Also, im not sure adding 500-600 lbs of motor weight would be considered a good trade for the reduction gears.
That would have to hurt efficiency, range, performance and handling etc.

Once again, you still don't need any weight penalty no matter how much of a grude you may have against the obvious first principals optimization of driving a wheel.

Here is an example of very high power density and very high torque density (this one is ironless which helps keep it light).

http://www.thingap.com/high-power-density/#torq-three