20kw continuous motor enough for small 2000lb 40-45mph car ?

[neptronix]

http://www.goldenmotor.com/eCar/HPM10KW (120V) Curve.pdf

http://www.goldenmotor.com/eCar/HPM10KW (120V) Data.pdf

I want to put this motor in a small ~2000lb car like a Toyota Tercel ( or something like it )
I only need the car to do about 45mph for 50 miles - basically a vehicle to commute to work with and use to get groceries.

Looks like it can do 20kW continuous on 120v. Heat would be around 3000W continuous at that level, but can be liquid cooled, so maybe with the right water pump and radiator, it could do this indefinitely? I heard that 40mph requires around 10kw continuous on a small car like this. Maybe 45mph then requires 12-15kw?

The idea of using a 17kg ( 37lb? ) motor is super appealing. Do you think i'm stretching this motor too far?

 

[amberwolf]

What car? If you know it's wind resistance, you can figure out it's power usage at various speeds, with the online calculators for that kind of thing, and then know if that car will need a bigger motor or not.

What RPM is that motor going to be at? Will it be compatible with the car's orignal transmission? Or with a different (manual, for best efficiency) transmission that will still fit in the car?


If the motor will be at it's max continous power all the time, you'll have to be sure you can get rid of that 3KW (or however much it is) of heat thru the cooling system, and keep cooling it even when the car is at a stop in traffic, even in worst-case hot weather.

So the cooling system has to be able to remove that 3KW of heat at any instant from the motor, and move it to a radiator that can get rid of it, too.

I'd expect that the car's original radiator could do the heat-shedding, but you'll likley need a high-volume pump that is very reliable, and can run all the time during the trip (and after you stop until it cools the motor down enough to shut off automatically).

 

[neptronix]

According to some calculators i've seen online, to cruise at 50mph would require 7KW constant on a 1999 Toyota Tercel ( 0.32cD, 2000lbs, and low frontal area ). That does seem optimistic though! I'd be doing some pretty classic aero mods, like a smooth underbody plate, grille blockoff, rear wheel skirt, etc. Then i'd also reduce the weight however i can.

So i figure that a 20kW powertrain would be rather decent for climbing hills at lower speeds, etc.

RPM? i'd keep the car at as high of an RPM as possible of course. It'd be a manual transmission. I have a local guy who can make a 7/8" shaft to transmission adapter for me, as he's done a geo metro with a doublewide etek. ( how well that motor performs on a car of that size is up for debate )

The golden motor unit can be liquid cooled, so i think that shedding 3kW continuous might actually be possible. However in this flat area, i would be cruising along at 45mph on level or slightly sloped ground, so for the majority of the time, i'd be pushing about 10kW, which would only produce ~1500W of heat.

For removing the heat, i was thinking of using a valve that could either run the hot liquid through the original heater core in the winter, or redirect it through a 50cc motorcycle sized radiator. I'm not sure what kind of pump i'd use; maybe 500 watts could flow the coolant pretty well, and i could run sort of an electric thermostat that would run the pump at a certain temperature threshold.

The question is, what's the real world power usage hauling about 2500lb of driver and car up a 5% grade at 40mph? that's the worst case scenario i'll face. Some rudimentary calculations say that my power usage would be in the 20-25kw range, but i can't find a calculator online that has the ability to add the grade, etc.

 

[nieles]

not sure if this is the correct way to calculate, someone please check the physics.

i am going to calculate how much energy is needed to lift an object (vertically) to i higher point.

E=m*g*h

P=E/t

P=(m*g*h)/t

40mph = 17.9m/s



delta h (vertical) = 17.9 / 100.1 * 5 = 0.89m (per second)

m= 2500lb = 1134kg

g= 9.81 m/s^2

t = 1s

P= ( 1134 * 9.81 * 0.89 ) / 1 = 9901W = 9.9kW

so add the power needed to travel a 40mph to this and it should give you a fairly accurate number.

 

[hydro-one]

whats the point of having an electric drive if your going to waste 3000w of energy when driving it, no starter for me .. Its like driving my bike with the cro vs the ninec at the same power level. cro goes way farther/cold at end of ride.

but then again the cro cost 800$, i guess economics could dictate undersizing the motor, but then you are going to need batteries to supply that 3000w of waste heat , that will cost a few pennies

 

[Jeremy Harris]

I think my main concern would be whether the motor rating is realistic and whether or not the liquid cooling can handle the heat load at maximum power. At max power it's going to be trying to get rid of around 4kW of heat, which seems reasonable with liquid cooling. You're going to be pulling around the rated power from that motor when going up the gradient at that speed, so it all hinges on how long the motor takes to get hot and whether or not the cooling system can get the heat out of the motor fast enough.

My experience has been that advertised motor ratings seem to often be just a little optimistic, but having said that this motor looks, on the face of it, to be quite nicely made. I vote that you be the guinea pig for ES, buy one, test it and let us all know whether it meets the advertised spec.........

 

[Jeremy Harris]

whats the point of having an electric drive if your going to waste 3000w of energy when driving it, no starter for me .. Its like driving my bike with the cro vs the ninec at the same power level. cro goes way farther/cold at end of ride.

but then again the cro cost 800$, i guess economics could dictate undersizing the motor, but then you are going to need batteries to supply that 3000w of waste heat , that will cost a few pennies

It's a lot better than having a car with a petrol or diesel engine and wasting around 60% or more of the power being fed into it, though. A pretty efficient car engine delivering 20kW is going to be using around 60kW or more of fuel, and wasting most of it as heat.

 

[hydro-one]

Anything is better than a petrol driven car i wholeheartedly agree. 3000w/20000w 15% wasted energy. i guess thats 85% efficent, alright for a motor i guess. its seems like alot of waste, but i guess everythings alot when its a car size.

Thats why we need super light cars. that look like enclosed street luges

 

[neptronix]

whats the point of having an electric drive if your going to waste 3000w of energy when driving it, no starter for me .. Its like driving my bike with the cro vs the ninec at the same power level. cro goes way farther/cold at end of ride.

Your cromotor will do about 85% efficiency too. The reason it's cold for you is that you aren't likely pinning it at 3000W-4000W continuous, and the motor is huge.. so it can shed what heat it generates really fast. It still generates around 15% of it's power as heat. I can't get it to do 90% efficiency on the ebikes.ca simulator.

That's 3,000W of waste at 20,000W of power, or 15% energy wasted in a 85% efficient motor. If i am cruising at 10kW, i can expect to shed about 1500W of waste heat. That's actually not too bad.

If i had a 90% efficient motor, i'd be shedding 2000W. But i'd have to pay 2-3 times as much for the motor, to get a 5% efficiency gain. Such a motor would also be 100lbs heavier. So i'd have to haul around more weight, which would increase my power usage.

For comparison;

Here's a netgain warp9 ( commonly used in budget EV conversions ) - $1,740

Peaks at about 88% efficient, so would be making 2400W of heat in this instance.

Another motor.. is the Motenergy ME1002. It's 190lbs and can put out around 20kW continuous or so at 90% efficiency. But to get that efficiency gain, i have to haul around 153lbs of extra motor, compared to the golden motor unit. It will take extra power to lug that extra motor up a hill.. lol

http://www.electricmotorsport.com/store/ems_ev_parts_motors_me1002.php

And this motor is double the cost of the golden motor unit, at $1,200 ( @ cloudelectric ).

http://www.thunderstruck-ev.com/ac-35.html

There's the AC-20 Kit with a ~50lb motor, and the motor + controller is $2,700. It's got higher peak power output, but the company reslling these states no continuous power level, and has no efficiency graphs whatsoever. They don't even own a dyno. so i won't even compare this one.

http://www.thunderstruck-ev.com/ac-20.html

Weight comparison:
Warp9: 143lbs
ME1002: 190lbs
GM 20kW: 37.5lbs

It might really just be more cost effective to carry around a bigger battery!

 

[neptronix]

I think my main concern would be whether the motor rating is realistic and whether or not the liquid cooling can handle the heat load at maximum power. At max power it's going to be trying to get rid of around 4kW of heat, which seems reasonable with liquid cooling. You're going to be pulling around the rated power from that motor when going up the gradient at that speed, so it all hinges on how long the motor takes to get hot and whether or not the cooling system can get the heat out of the motor fast enough

Yeah, that's the question - what's the continuous heat dissipation? I wish i knew. I can't find anyone else who's attempted this. I think at worst, i could replicate luke's deathbike on a motorcycle chassis :lol:

One thing to think about is that liveforphysics pushes 40kw during acceleration out of this motor in a non liquid cooled config. He's raced the thing around >60mph, probably pushing the 7kW-10kW continuous. So maybe this motor isn't so bad in a car that, sans motor, would weigh about 1700-1800 lbs.

 

[Jeremy Harris]

Yeah, that's the question - what's the continuous heat dissipation? I wish i knew. I can't find anyone else who's attempted this. I think at worst, i could replicate luke's deathbike on a motorcycle chassis :lol:

One thing to think about is that liveforphysics pushes 40kw during acceleration out of this motor in a non liquid cooled config. He's raced the thing around >60mph, probably pushing the 7kW-10kW continuous. So maybe this motor isn't so bad in a car that, sans motor, would weigh about 1700-1800 lbs.

I say go for it if the price is reasonable. If nothing else I'd be really interested to see how this motor performs, as I've had a back burner (and slightly barking mad) project in mind for some time where this motor would be a reasonable choice.

 

[neptronix]

Hey nieles; i think your calculations are very close to reality

http://www.evalbum.com/2575

Here is one of many who have converted his tercel to electric power.
He states ~145A at 96v. under load, his batteries are probably at 90V.
90 X 140 = 13050W to move along at 55mph.

That's pretty good considering that his model has worse aerodynamics than the 1996-1999 model i am looking to buy.

He is using a motor that is about 25% - 33% more powerful than the GM ~20kW motor.
That motor is 110lbs.

http://www.evparts.com/products/str...-to-192-volt-street-vehicle-motors/mt2116.htm

The motor would be $1800 and the controller would be $1720. That's ~$3500 ...

 

[neptronix]

Controller thoughts...

The golden motor is under $800 and the HPC700 is ~$900. A $1700 powertrain is nice, but..

The golden motor HPC700 controller has a peak input current of 280A, so if the max continuous current is 140A, then it can only push 16.8kW at 120v.. that is too low of power for me.. :/

Sevcon gen4's are capped at 116v voltage. That means i could run a 31 cell lifepo4 pack ( 113.15v fully charged ), which would be 99v nominal. That sucks. But the size 6 controller can do 210A continuous for 1 hour.
Maybe the lifepo4 batteries could be charged up to 3.5v? Then i can run 33S lifepo4 at a fully charged 115v, 105v nominal?

These would cost about $1199,

http://www.electricmotorsport.com/store/ems_ev_parts_controllers_sevcon_gen4_72-80_550.php

The Kelly KHB12601 can do 300A continuous, that's pretty good. I could dial it back down to 200A for a 24kW maximum output power.

http://kellycontroller.com/khb1260124-120v600aopto-bldc-controllerwith-regen-p-830.html

Those are $1,299. Warranty is one year. I don't have much faith in these controllers though.

I wish the sevcon wasn't so hard to program. I'd buy one and deal with the lower voltage if that wasn't the case.

 

[nieles]

just to make sure, what i calculated is only the power needed to make the climb. you need to add the power needed to travel at a speed of 40mph on a flat road.

so 9,9kW + ?5kW? = total power needed

 

[Electroddy]

Neptronics: I was looking for a larger motor for a possible Dodge P/U project. I found this KOREAN co. http://www.higenmotor.com/eng/product/product.asp They have some very nice specs, but I have not tried to make contact yet, so no idea of price/availability..

BTW I was looking at the 60-120KW motor 94Kg 164-426Nm torque.

 

[o00scorpion00o]

Hey Nep,

how about buying a high mileage MK II Prius ? it's got a 50kw motor and already has all the electronics and cooling you need. And the engine if you need more range !

The ECU can be hacked to allow 65 mph (don't quote me exactly on that )

I was considering converting the Prius, but the hassle of converting is putting me off and it wouldn't be economical esp for the range I need. I think I'll wait for a decent 2nd hand leaf, I've put in a request at work to install charge points, who knows, maybe they will. Leaf range is around 85 miles at 60 mph, in Winter -10% Irish average winter climate, so I need a charger at work for my 84 mile commute.

Pity North Americans can't buy a Twizy, it's Europe's best selling E.V, would probably work out perfect for you!

 

[neptronix]

nieles: whoops, i misunderstood you. It should be 9.9kW + about 8kW ( to be pessimistic ) then. That's doable.

Neptronics: I was looking for a larger motor for a possible Dodge P/U project. I found this KOREAN co. http://www.higenmotor.com/eng/product/product.asp They have some very nice specs, but I have not tried to make contact yet, so no idea of price/availability..

BTW I was looking at the 60-120KW motor 94Kg 164-426Nm torque.

Oooh, that's gonna be an expensive project. Why convert a pickup? continuous power is going to be double what a small car like this will need. I dunno where you'd get those motors though. If you can get a quote, let us know, those are appealing.

http://ecomodder.com/wiki/index.php/Vehicle_Coefficient_of_Drag_List

Here's a fantastic list from ecomodder, showing the CDA ( which can be used to calculate the entire aerodynamic drag of a car ) of various cars and trucks. The Ford F-150 is 11.36, the tercel i want is 6.26..!

 

[neptronix]

It might have a 50kW motor, but that is not a continuous figure. Continuous is much lower. People modding priuses have tried to do that, and it seems that homemade plugin prius conversions max out at about 50mph.

http://boulderhc.com/products.html <-- here is one kit i've seen before.

Those cars are $9,000 used here in the states, and that's with an old battery. Add lithium, BMS, etc. on top of that. You're up to $16,000 fast.

The tercel is lighter ( 2000lb versus the Prius's 2900lb ), a fraction of the price, and very close in aerodynamics ( that's why these old econo-beaters are great ). I want to come up with a nice formula for a cheap DIY electric car that other people can copy and do on their own, en-masse. That's the idea here!

Hey Nep,

how about buying a high mileage MK II Prius ? it's got a 50kw motor and already has all the electronics and cooling you need. And the engine if you need more range !

The ECU can be hacked to allow 65 mph (don't quote me exactly on that )

I was considering converting the Prius, but the hassle of converting is putting me off and it wouldn't be economical esp for the range I need. I think I'll wait for a decent 2nd hand leaf, I've put in a request at work to install charge points, who knows, maybe they will. Leaf range is around 85 miles at 60 mph, in Winter -10% Irish average winter climate, so I need a charger at work for my 84 mile commute.

Pity North Americans can't buy a Twizy, it's Europe's best selling E.V, would probably work out perfect for you!

 

[neptronix]

I just noticed something..

Looks like this motor can run 2 controllers at once?

If this is the case, then two HPC500 controllers would be fantastic.
If one blows up, i could hobble home.

If this is a 4 pole motor, maybe two greentime 36FET 4115 controllers on a 95A amp limit would do.
190A total X 115v nominal = 21850W = 23 hp continuous to the shaft.
Maybe i could figure out a water cooling setup for those too. That is, if they could tolerate the motor.

Now that would buck the power:dollar:weight ratio

 

[hydro-one]

go for it , greentime FTW!! i love the controller redundance, that is something that has left me walking my bike many many times, would have loved to have a 6fet in my pocket, who cares if its slow , better than walking or pushing

 

[boostjuice]

Assuming the HPM10000B shares the same design principles as the HPM5000B.....

.....The thermal resistance between the phase coils and the liquid cooling is terrible with this motor design, just as it is with the fan cooled motor. The whole motor needs a case redesign to efficiently get the heat out of the coils and into the casing. The best way I can see is using vacuum impregnated thermal epoxy (unfortunately from the reading I have been doing is way too hard to DIY properly with basic equipment).

The path phase coil heat must take is coils > air gap > Nomex paper (or like) > tooth laminations > a few small tack welds joining the tooth laminations to a spiraled disc of backing laminations > non thermal greased and unlapped clamping to cast Aluminium backing plate > liquid/air = Piss poor thermal conductivity. In fact im sure that the thermal path though the air contained within the motor dominates in the thermal transfer rather than through any solid medium.

Anyway, bottom line is that to bring the real-world continuous power figure closer to the unsaturated burst power potential of this motor, the thermal managment design needs serious revision.

 

[neptronix]

Hmmmmm, yeah that's something to consider.

The interior of the stator would be really hot trying to ditch 3000W of heat on hills.
Only one side of one stator would get cooled.

Well, you kinda took a crap on my rainbow to be honest :lol:

Do you have any other pics of the interior by the way?

 

[neptronix]

The more i read, the more it seems like i'm gonna end up with an impulse 9.
This motor is 129lb, which i don't like, but it's one of their lighter motors, and can do >30hp continuous, so it will climb those hills with some extra headroom.

I wonder if the 90-94% efficiency is real?
That would negate the extra weight.

If i wanted to stick with the golden motor 10-20kw units, i could run two.
I'd get two 96V models and run them at 120v x 120A. This would produce 14,400W each x 2 = 28,800W total, or 39 horsepower.
Each motor would then only need to shed 2160W at peak.
If i split the load between the two motors while cruising at 10kw, they run only 5kw each, and only need to dissipate 750W.

I think this could be reasonably done, but the price would be high.

2x HPC500 controllers = $1380
2X 10-20kw motors = $1570
Total: $2,950
Weight: 87.78lbs, but i will need to build mounting plates / coupler, so more likely >100lbs.
Max output: 28,800W on 120v x 240A.

Let's do the math on the impulse 9 setup;

1x Netgain Impulse 9: $1,700
1x Alltrax SPM-72500 500A 72v ( OK up to 90vdc, so let's run it at 25S lifepo4 aka 80v nominal ) : $688
Total: $2,388
Weight: 136lbs.
Max output: 30,400W on 80v x 380A

Hmmm.. if the netgain efficiency corresponds to that graph, then the netgain motor definitely wins here on price and value.

I wish i could find a more affordable 120v-144V capable brushed controller. More powerful controllers start at $1,500 and go up to $2,000 to max out the netgain impulse 9. Oh well. 30kW is pretty damn good peak power!

 

[Arlo1]

Chances are if you blow 1 controller the other will go too. When controllers pop the fet usaly short eventualy if left hooked to hi power long enought the shorted fets will burn what they need to to open the short, BUT in this time there is full voltage flowing all over and usaly coming out of the gate into the rest of the low voltage circuit. If you have 2 "dumb" controllers hooked to it you might find they both blow... Or you could get lucky and they will be ok but with something with phase current monitering it may have just enough smarts and a quick enough reaction to save it self in the event of other other controller blowing... I would like to see more data on this motor. I would love to test on... HMMM

 

[o00scorpion00o]

It might have a 50kW motor, but that is not a continuous figure. Continuous is much lower. People modding priuses have tried to do that, and it seems that homemade plugin prius conversions max out at about 50mph.

http://boulderhc.com/products.html <-- here is one kit i've seen before.

Those cars are $9,000 used here in the states, and that's with an old battery. Add lithium, BMS, etc. on top of that. You're up to $16,000 fast.

The tercel is lighter ( 2000lb versus the Prius's 2900lb ), a fraction of the price, and very close in aerodynamics ( that's why these old econo-beaters are great ). I want to come up with a nice formula for a cheap DIY electric car that other people can copy and do on their own, en-masse. That's the idea here!

I'm nearly sure the Prius can be hacked for 60-65 mph and you would get over 100 mpg in blended mode if you need more range .

Sure it wouldn't be 50 kw continuous but at 45-50 you wouldn't be pulling 50 kw on level ground .

It's going to be expensive anyway if you want 50 miles range at 45-50 mph !

A 2nd hand Leaf would be much better to wait for ?

Actually with the ultra cheap lease deals you can get over there would it not be better to lease a Leaf for 200 a month V spend thousands Dollars for batteries, motor and car ?

By the time you spend on a lease what you would on a conversion, your batteries would probably be near dead anyway.