If CVTs are supposed to set the gearing ratio at the ideal (exactly as much torque as the car needs at that moment to drive the wheels, with the rest of the engine power producing horsepower), why are CVTs in practice so shitty/low power compared to traditional gearboxes?
Also, how do electric cars get away with not having any transmissions? How does power get from the electric engine to the wheels?
Other urls found in this thread:
youtube.com
youtube.com
youtu.be
youtube.com
twitter.com
>why are CVTs in practice so shitty/low power compared to traditional gearboxes?
Because they can't handle any appreciable amount of torque so they either
>get put on a torqueless shitbox engine
or
>bet put on something with a little bit of torque and self destruct
yes, but why is this the case - do the belts just not have the ability to transmit enough power?
>do the belts just not have the ability to transmit enough power?
Yes. They have limited surface area and are designed to be quiet instead of to hold a lot of torque.
>Also, how do electric cars get away with not having any transmissions? How does power get from the electric engine to the wheels?
They don't have gearboxes because electric motors have a linear torque output instead of a curve, and they don't need transmissions because the motors are generally fitted right to the wheels because it's easier to just have four motors than one big one -- similar to why every "drone" these days is a memestastic quadrotor. It's not as efficient, but it's easier to manage and features fewer potential failure points.
Some electric cars have high/low ranges like automatic car shifters do but it's more for climbing hills or start-stop traffic vs constant highway driving than being there to regulate torque like a CVT or traditional gearbox does
Electric motors have a wide power curve,enough so that you don't need more than one gear ratio (usually there's one somewhere). You just adjust the amount of power by switching max power on and off really quickly, with the percentage of on/off time being the percentage of power.
A
FUCKING
B
E
L
T
EV's still have a transmission, It's just a single speed gear reduction. They get away with it because motors have torque and rev to 20 000 rpm so they can pull off low gearing.
I have a question.
Since hp is a function of torque and most engines today are produce all their torque at 2k rpm, surely with infinite gears, the car should always stay at 2k rpm and still produce maximum hp to the wheels.
>it's easier to just have four motors than one big one -- similar to why every "drone" these days is a memestastic quadrotor.
>mfw picturing a quadrotor with two motors each with and 2 independent drives with collective pitch instead of 4 motors with variable speed
where is the question
but yes, a CVT is basically an infinite number of gears and is capable of sitting at peak horsepower or peak fuel efficiency, whichever is in demand
Because electric motors have much greater flexibility. An internal combustion engine must be run at a minimum rpm in order to prevent it from stalling, along other things like not having enough torque to move a heavy mass from a standstill without torque multiplication.
You can have gears on an electric motor, it's just simply not really necessary so it isn't done often. Adding a gearbox introduces mechanical losses, weight, packaging problems (less room for battery), abd additional maintenance.
For the sake of simplicity, imagine cvt's as a f friction belt that uses friction to grip and move a pulley sort of thing. If you add too much torque all at once, you overcome the friction coefficient between the belt and the pulley (because there's wheels and the cars weight that wants to prevent the pulley from spinning freely) and now the belt is basically just grinding away at the pulley like a wire-saw at really high rpms.
You can see this in the WRX cvt. It but actually gets LESS mileage than the manual because I bet Subaru had to a bunch of crap to reinforce the transmission to make sure it can handle the torque of the twinscroll turbo.
>dude what if like lmao you attached like two rotors to one engine haha
NEVA BEEN DUN BEFO
this said I think some early electric cars actually did just have one big electric motor that powers a normal drivetrain, im pretty sure its how recreational shit like golf carts operate too. dont quote me on this tho.
I straight up forgot to phrase it like a question. Lemme do it right quick
>Shouldn't cars be staying always at around 2k rpm for maximum efficiency and power?
thats basically the point of a cvt, yes
the problems arent so much theoretical stuff as that it's really hard to make one that won't disintegrateafter four miles of casual sunday driving
~2k is where most gas engines are most fuel efficient so having a cvt keep the engine there at exactly where it is most efficient is the idea
max power really depends on the engine. My '91 miata makes the most power at 6500rpm. power = torque over distance = acceleration. If I had a cvt that could handle the mighty torque of my 1.6, it would accelerate the fastest if the cvt would keep my engine at 6500
CVT is for fuel economy. Others are for fun.
That said it isn't impossible to have a cvt race car.
youtube.com
>hp is a function of torque
>most engines today are produce all their torque at 2k rpm
Wrong and wrong
>hp is a function of torque
>wrong
come on, give that user a break, it's only half wrong
I think you meant to post this
In practice, no one does a CVT right.
In a car with a CVT, the accelerator pedal should have no control over the engine. Instead, it should directly control the CVT. The car's computer automatically runs the engine at its peak torque whenever the driver is attempting to move the vehicle, while the driver uses the accelerator pedal to change the gearing ratio of the CVT to change speeds while engine speed remains constant.
CVT is number 1
Snowmobiles use CVT's and they're pretty fun. However belt breaks are common which is something you kind of don't want in a car.
After thinking about it for a while, it seems that I am mistaken. I mean, hp is a function of torque, but staying at 2k would not translate into best acceleration.
I do still stand behind my ~2k rpm max torque, though.
>I do still stand behind my ~2k rpm max torque, though.
maybe for a diesel powered container ship
Google
>[insert model name] 2017 dyno
look at results.
snowmobiles are so gay
Really just most small turbo 4 cars and SUVs.
Not all CVT's use belts. Nissan has a pretty good one that use traction oils to transmit the force between the input and output gear ratios.
Also, note that your torque curve controls how much you can get away with peak power-wise.
Also, thank you user who pointed out how CVT's reverse what the accelerator controls. You get it.
nearly infinite ratios between two points available. yet they are still programmed to use a set number of them like a planetary gear automatic.
Which electric cars have 1 motor per wheel? Definitely not Teslas or the Chevy
Well yes, but a snowmobile has a much larger power-to-weight ratio than your normal car, and they dont reduce the gearing in the same way that a car-CVT does, having the RPM be much more consistent with the speed
Please no. I don't wanna be gay
Definitely Teslas, user.
>If CVTs are supposed to set the gearing ratio at the ideal (exactly as much torque as the car needs at that moment to drive the wheels, with the rest of the engine power producing horsepower), why are CVTs in practice so shitty/low power compared to traditional gearboxes?
CVT's are usually programmed for maximum efficiency in regards to fuel consumption. There are other modes that allow a CVT to maximize torque and power
I saw these dudes do an electric conversion to a car and they left the old transmission in. Was pretty cool watching the dude silently row through gears.
because most just can't handle the torque. a lot of big tractors using them these days tho so they CAN be made to handle it but likely not affordable enough for cars and such. in fact those transmissions will cost you about the same amount as your average car
considering peak torque is around 6000 rpm (depending on engine) that'd be pretty fucking inefficient plus most cooling systems aren't designed to keep up with an engine running like that
Not at all. Teslas have 1 motor per axle, and they have differentials too.
Thats not how you control an electric motor there bud.
Electric cars dont need gearboxes because they have enough torque to move the vehicle at zero revs and so don;t need to multiply output to the wheels like ICEs do. The main tradeoff is top speed with most electric cars topping out at relatively low speeds
Which is exactly why even EVs should have at least two gears. One regular, and one that acts like an overdrive. A dual gear gearbox is small and shouldn't introduce a whole lot of friction losses.
>why are CVTs in practice so shitty/low power compared to traditional gearboxes
Just look at the power band in a Subaru CVT and you wonder how on earth does such a flimsy-looking piece of braided metal wire hold up for 80,000 miles. And of course, how does it develop enough friction to those shiny metal surfaces without wearing away all the metal in the links or spalling the smooth metal surface?
It also makes you wonder if a stop/start engine causes extra stress to this CVT belt as it has to go from full stop to being in motion over and over and over each time the engine stops running at a stop sign or traffic light.
Was that youtu.be
He's trying to describe pulse width modulation, and it is the current standard for motor control.
Rimac has 4 electric motors, Tesla has 1 per axle on AWD models, wich saves them a driveshaft and a center differential.
The point is that adding a differential to a 2-step reduction gear is not complicated or expensive, but a additional driveshaft from front to rear with a additional center differential is.
Also the rimac uses its 4 motors for torque vectoring and faster slip controll, theoreticly it could turn while staying at the same place like a tank with it or put down the exact amount of power on each wheel midcorner.
That would be impossible with conventional differentials, but Tesla is not exacly a performance oriented brand.
>The car's computer automatically runs the engine at its peak torque whenever the driver is attempting to move the vehicle
That would make no sense at all, also most CVTs change their gear ratio controlled by a computer and the throttle controlled by the same computer to run the engine at maximum thermal efficiency for the required power output.
Pic related, a prius BSFC and the line on wich the CVT keeps the engine.
>hp is a function of torque
Not exactly, power is a function of torque and rpm:
P=M*n/9550
So for example 100 Nm at 2000 rpm is the same as 50 Nm at 4000 rpm
>most engines today are produce all their torque at 2k rpm
Not exactly, most engines produce about 80% of their torque at 2000 rpm
>stay at 2k rpm and still produce maximum hp to the wheels
No, since most engines produce only a fraction of their power at 2000 rpm, for example:
Assuming a 2,3 L gas engine produces 200 Nm at 2000 rpm, how much power does it produce?
200Nm*2000rpm/9550=41,9 kW
41,9 kW are verry little power for acceleration, less than your average microhatch produces.
Now assume it revs 6000 rpm at the same torque:
200Nm*6000rpm/9550=125,7 kW
That is already three times as much power.
That is also the reason you shift down when you accellerate onto a highway.
PWM is not exactly the standard for automitive electric motors.
Usually the use a variable step-down to controll the voltage going into the motor.
More voltage -> more amps -> more power
(the controller limits the amps to a preset maximum, just like a wastegate on a turbo)
If you increase voltage and amps on a electric motor you can do crazy stuff, like a moped that accellerates like a literbike, a Datsun shitbox with a forklift motor that accellerates faster than a P100D
Only issue is temperature und on brush fire on brushed DC motors.
>do the belts just not have the ability to transmit enough power
In the beginning, automatic transmissions also had power problems. When CVT can also get rid of the belt coupling and move to fluid coupling, that may usher in development of a less problematic CVT which doesn't have so many belt failures.
Not true. Some are, because morons thought the CVT operating properly felt weird. But the CVT in my Civic doesn't do this at all.
>When CVT can also get rid of the belt coupling and move to fluid coupling
That would no longer be a CVT but a hydrostatic transmission, which is inherently less efficient than a CVT and therefore not going to happen in the modern auto industry
>therefore not going to happen in the modern auto industry
It sucks that CVT is used to generate improved MPG stats for the car makers. It basically shifts the costs of meeting MPG onto the consumers who get less reliable automatic transmissions that will eventually need an expensive replacement (since dealers want to replace instead of repair in situ).
This is the problem of the "torks wins races" meme.
It doesn't. Horsepower is quite directly "work done".
Here is a basic example of this. You have two guys. One guy who can only carry a 20 lbs box and one other guy who can carry two 20 lbs boxes. They are tasked with moving boxes from one side of the warehouse to the other. Now big man can carry more boxes at once but he is slow to begin with and generally doesn't care for this job much so he only moves two boxes every 5 mins. Hungry skellington man on the other hand is nimble and needs this job so bad that he hustles to make 3 trips in 5 mins, delivering 3 boxes for bro-dude's 2 boxes.
The end result for this scenario is obvious, the one moving more boxes getting more work done (actual wheel horse power) and the other dude is fucking fired for being a lazy cunt AND harassing the admin girls.
>how do electric cars get away with not having any transmissions?
You can't stall an electric motor and you can "rev" up an electric motor much much higher
You're almost there. I drive a CVT daily, in my experience for normal acceleration the car will "shift" because that's what people expect to hear/feel, so manufacturers made it seem like a normal car. If I floor it, my car will sit at max power, ~5800 rpms, and the gear ratio will slowly decrease as the wheels speed up. When I let off the gas, back down to ~1900 rpms because fuel efficiency. Benefit here over manual/automatic is I don't have to choose between 2 gears, neither of which might be *optimal* for my current speed, my car provides the correct final drive ratio.
Hp and torque in an ICE are both functions of rpms, so they can't both be at max at 2k rpms. IIRC I do think they can match at ~5800 because of the math, but I'm not exactly qualified to get into that. Think of diesels, low rpms with high torque and low HP, versus a V12 with high HP and low torque. It all has to do with how fast you can rev. What you're thinking of is a car sitting at 2k rpms and proving enough power to the wheels for any speed(up to mechanical limit) with the help of *infinite gears*, which yes, CVT's can do
I'll answer the electrical bit since almost every response for it is wrong.
This is a low level explanation but quite accurate.
Electric motors produce max torque through their whole rpm range up until you want to 'overdrive' them. At that point you trade off torque for more rpm. It's like getting more speed for free out of a lower spec motor since it's rare to require peak torque at massive rpms.
Motor control is largely PWM or a twist on PWM so he's right
No you burger.
>step down to control voltage
No that would incur losses before you even start to send power to the motor. You avoid doing this and build batteries that supply the same voltage as the motor requires.
>moar voltage means moar amps
Nah man, PWM is what controls the power. You pulse 'on' for longer to allow more current to be drawn by the motor.
In electric motors, voltage is proportional to speed and current is proportional to torque. Since you load a motor the variable has to be torque and therefore current.
Remember that you can apply anything up to max torque over the whole rpm range, so voltage will be constant and speed will increase up to the maximum dictated by voltage thanks to torque.
Voltage*current=power
Speed*torque=power
Also brushed dc motors are a load of fuck and only used on shit like drills nowadays. It's like using wooden wheels at this point.
>electric motors can't stall
No, electric motors can stall, and will catch fire unless you shut them down immediately.
Electrical and electric engineer here btw, I work in the auto industry.
>If I floor it, my car will sit at max power, ~5800 rpms
Does it immediately go right up to that high RPM amount? In my normal automatic transmission car, I am so used to seeing the RPMs creeep up more slowly because my muscle memory has me trained to "floor it" only to the point where I get to 4000 RPM and no more. Beyond that I start to feel nervous about stressing out the engine.
But if the CVT always shoves it to the 5800 RPM point (or whatever the equivalent is for my car model), then my usual driving method of "flooring it" would not work. The engine would simply go to max power (5800 rpm) and let the CVT do the actual adjusting. Maybe I am a dinosaur, but I would rather do the adjusting myself. Or if the CVT performance package for the car allowed me to do so, I would set a max RPM limit of 4500 rpm unless the pedal was truly pressed all the way to the floor. But up to pressing it all the way, the accelerator would allow a semi-linear access to the max limit of 4500 rpm to maybe the 80% pedeal depress point. If it allowed that, then maybe I would like CVT more. I don't want the engine to be gunned all the time to max RPM each time I press the pedal halfway down.
>why are CVTs in practice so shitty/low power compared to traditional gearboxes?
The picture you posted already suggested the answer to you. All the power has to flow through that belt. If too much force is applied to the belt, it or the surfaces it contacts will be damaged and thus wear out prematurely.
>electric engine
>electric engine
An electric engine would be quite a Rube Goldberg device. It's far more efficient to stick to electric motors than creating a fanciful electric engine.
If OP didn't mean "engine" but meant "motor" instead, his post is an example of how careless use of words results in a waste of posts answering the wrong problem.
>Rube Goldberg device
youtube.com