Conclusive proof gearboxes are awesome.

[liveforphysics]

Perhaps think of this. Each gearing stage that increases torque trades motor shaft speed for torque. This same trade can be had through motor geometry alone, and then it doesn't have parts being added that add weight and frictional losses and failure modes.

If you want to match the power to weight of a small RC motor, you simply need to use the iron and copper at the same rates and current densities, but configured in a geometry that provides the desired output drive torque as a function of a single stage.

 

[John in CR]

I guess all your motors are at peak efficiency at all RPMs.
All the motors in the real world have an efficiency curve through the RPM band.

Not only is the graph one of a terrible motor like Luke said, a controller setup to result in that graph is improper for our use. Current limiting comes into play that greatly broadens and flattens the efficiency and power curves. That's not the biggest problem with using those graphical examples to make decisions about a real world setup. The biggest problem is that it shows the result of only one throttle position from 0 rpm to the top, but our variable throttle gives us an infinite number of graphs, each with different peak power and efficiency points.

If you want to ride something that is so horribly under-powered that you need to ride at WOT most or all the time, then sure when you get to a hill you need to be geared lower.

Our argument has always been that the added weight is better spent on more motor instead. For me the only exception is a dual use vehicle that needs highway speeds on flat road, and also needs low enough gearing for efficient low speed high load use off road. I couldn't pull off Miles' elegant Retro-Direct 2 speed (and don't want to give up regen or the ability to roll backward anyway), but the gear change would be seldom enough that I'd do it the simple and just as durable way by adding only an extra front and rear sprockets or pulleys along with a quick release lever for the idler/tensioner. ie Stop, release the idler, and move the chain/belt to the other gearing by hand.

Motomoto, do yourself a favor and build a bike with an appropriately sized motor that does what you want using as little WOT as you would with a gasser. Once electrics spoil you with all the torque you want over such a large band, then you'd never even consider a shift to 2nd gear that reduces torque.

 

[Miles]

Does anyone disagree with using this as an example?

This shows a torque band requirement of 4:1

 

[macribs]

Some explanation to the example would go a long way.

 

[Miles]

Your cruising along at 45 kph on the flat. You come to a 15% gradient. What is the ratio of the difference in torque requirement, assuming you don't want to slow down.

I've no need for spectacular acceleration but I certainly don't want to slow down much when I come to a hill.....

 

[Miles]

I chose 45 kph because it's between bicycle and motorbike speeds. Obviously, the ratio of torque requirement, for the change in gradient (0-15% in this case) will vary depending on the vehicle speed and CdA. The torque required to climb the hill is constant. The torque required to overcome drag goes up as the second power.

 

[toolman2]

I chose 45 kph because it's between bicycle and motorbike speeds. Obviously, the ratio of torque requirement, for the change in gradient (0-15% in this case) will vary depending on the vehicle speed and CdA. The torque required to climb the hill is constant. The torque required to overcome drag goes up as the second power.

Dam, its started again, why motomoto? .
Totally agree Miles if we want more than opinions or language to decide this we need a concrete example, but a 15% grade and a desire for no loss of speed sounds like the perfect scenario for a well sized motor and definately no gearbox.
However i regularly have to climb a 71% grade (yes i measured it with engineers digital angle gauge on a 1 meter flatbar) hill on an electric bike (dirt track) that current 4wd's and motorbikes can slightly more readily climb. A substaintial drop in speed is acceptable though.
And i would expect luke to also want a further example of a 4 fold + increase in cruise speed to also be part of the equation :lol: , maby these as further example variations, if you are laying the concrete that is, to have all real possibilities covered?

 

[Miles]

Luke's case is already covered because increasing the speed will only decrease the torque ratio.

I'll run your case as the extreme variant though... :wink:

 

[Miles]

toolman2's case requires a torque ratio of 37:1

I ran it a 25 kph. Is that reasonable?

 

[Miles]

Maybe run it again with 25 kph up the 71% gradient and 45 kph on the flat?

 

[Miles]

That lowers the torque ratio to 13:1

 

[toolman2]

toolman2's case requires a torque ratio of 37:1

Well yes, you want your hiking boots to just to stand there, and looking down makes you hesitate like it may be a cliff.
But in OZ we are a land of bogans who like it for 4wding.

 

[Miles]

This shows eta for a joby motor using the 4:1 torque ratio of the inital example.

Not much need for variable gearing there........

 

[toolman2]

I think i get where your coming from luke, that ultimately in theory and with little to no limitations to the scaling problems of getting a motor to be say the rim diameter of the same weight and efficiency then yes, for sure it cannot be beaten.
So grab yourselves a magic pie chaps cos apart from a few prototypes that i doubt you can buy at present, its the closest thing we have and already has a good dose of these scaling problems as its Achilles heel.

The reality however, is that the gear ratio used for motors we do actually have for nearly ANY vehicle/machine electric or not is ALLWAYS a compromise.
And not to ever consider changing it at any time leaves an advantage not gained.

the varied terrain and conditions one encounters negotiating this planet mean that we are likely to have an advantage in one way or another to taller or shorter ratio when conditions suit.

-reasonably unavoidable fact that.

And even with a rim motor (at 1:1) then the real world challenge/compromise will continue in the form of a different tooth length or copper/iron fill to change its torque vs rpm losses (just like a ratio adjustment) that can still NEVER be perfectly correct for all conditions, so it will never end.

 

[Miles]

Here's the last example using a 13:1 torque ratio

 

[Miles]

And even with a rim motor (at 1:1) then the real world challenge/compromise will continue in the form of a different tooth length or copper/iron fill to change its torque vs rpm losses (just like a ratio adjustment) that can still NEVER be perfectly correct for all conditions, so it will never end.

Right!

A bigger PMSM motor will be more efficient at greater torque levels but less efficient at lower torque levels.

The diameter to length ratio of the airgap of motor is yet another compromise to be made.....

 

[Miles]

Here's the extreme case with a torque ratio of 37:1

 

[Punx0r]

Is anyone else noticing a very flat efficiency from moderate to high load... The copper losses might not great, but on a smaller, high speed motor with gear reduction you'd just be trading them for comparable core losses...

 

[John in CR]

Damn Toolman, I wouldn't want to walk on a 71% grade that didn't have stairs, much less ride up or down one unless it was really short. Of course my bike is much more capable than I'd dare. With big fat me aboard it has no problem accelerating past 80kph on 100-200m straights coming out of very low speed curves going up a solid 20% grade road, and that's with fixed gearing at a top speed of over 170kph. That means with you aboard and a fixed gearing for a healthy off-road top end of say 100-120kph, I think it would laugh at your 71% grade.

 

[toolman2]

Damn Toolman, I wouldn't want to walk on a 71% grade that didn't have stairs, much less ride up or down one unless it was really short. Of course my bike is much more capable than I'd dare. With big fat me aboard it has no problem accelerating past 80kph on 100-200m straights coming out of very low speed curves going up a solid 20% grade road, and that's with fixed gearing at a top speed of over 170kph. That means with you aboard and a fixed gearing for a healthy off-road top end of say 100-120kph, I think it would laugh at your 71% grade.

Yep, may-be John.
But right there with what i assume is a VERY large motor you have demonstrated both the need and understanding of the benefits of an adjustment in gear ratio.
Good to have you on-board.

 

[nutnspecial]

What is even the argument here? I am missing a clear argument against the efficacy of non-fixed gearing in 'E'. It seems obvious to me (a layman), advantage shifts toward variable gearing under appreciatively harsh circumstances.

For example I recently considerd Tesla VS . . . Depending on goals (such as cruising or competing with a supercar over about 4 seconds or 1/8mi, eventually energy storage and application would benefit from variable gearing.

Consider how well model S does against similar HP cars up to only 100mph before 'falling flat'. With HP similar, it seems the fixed gearing is it's weakpoint.

[youtube]_nQDU7HOStc[/youtube]

Sure, you could drastically increase the size of the motors, controllers, batteries to offset taller fixed gearing. Or you could consider variable gearing?

I just have no idea what the argument is against variable gearing. Seems obvious in my mind that some circumstances will always lead to benefit of variable gearing over fixed gearing in any contemporary application of 'E' ??

[youtube]y7j8BfIrfpY[/youtube]

 

[liveforphysics]

Tesla's are battery power limited. More gear options gives them more loss and less performance.

They started out with a 2spd. Then they went 1spd and improved range, dropped 0-60time AND got a higher top speed.

What Miles was trying to show folks here is the futility in attempting a mechanical solution to increase dynamic performance envelope when its already less loss than a tranny adds alone over the whole useful power profile.

 

[nutnspecial]

Aside from the fact I'm curious how 0-60 time, top speed, and range could possibly ALL be improved by going from 2spd to 1spd. .. .

What Miles was trying to show folks here is the futility in attempting a mechanical solution to increase dynamic performance envelope when its already less loss than a tranny adds alone over the whole useful power profile.

Does this mean: 'it's futile to attempt variable gearing over direct drive because more energy will be lost'?? Does that apply to my post, or anyone's argument here? It's interesting for further thought/discussion, but how is it relevant?
_______________________________
IMO one will obviously sacrifice some weight and energy efficiency in power delivery with variable gearing, but in some circumstances these sacrifices would still certainly be preferential and yield better specific results than alternative penalties (like increased total system size).
Tesla attempting to compete with a comparable HP ice beyond ~1/8mi and ~80mph was a good example imo, with the evidence in the gearbox ice spanking a slightly blockier and heavier tesla @ similar ~700HP.

. . . . .So I'm still curious if there is any disagreement to this idea. . . . .
-benefits can be found in variable gearing over fixed gearing for ice or 'E' in some circumstances-

-because E is more efficient than ICE doesn't mean variable gearing suddenly becomes completely irrelevant, but does evidence benefits do lessen as system 'efficiency' increases-

?therefore the only way to negate any/all benefit from variable gearing in any/all circumstance would possibly be to achieve 100% efficiency?

 

[toolman2]

What Miles was trying to show folks here is the futility in attempting a mechanical solution to increase dynamic performance envelope when its already less loss than a tranny adds alone over the whole useful power profile.

That, i assume is your opinion Luke.
Not sure your correct and im not sure if Miles wants to show folks that or not.

 

[Miles]

I was just trying to pin things down a bit

In my opinion and as I said, there's absolutely no point in having variable gearing for the first case, which I presume corresponds reasonably well with motomoto's expectations.

I think toolman's rather extreme case would probably benefit from a wide ratio 2 speed with one of them being a direct drive.