What's the deal with variable transmissions? Is this going to take over?

[Howard]

When torque is multiplied, power is reduced.

ZV

How do you figure?

 

[LTC8K6]

I only noticed one neat thing while driving my dad's ford 500 with a CVT. If you are going down a hill steep enough to gain speed while coasting it will automatically pick the right gear ratio to maintain you current speed using engine braking.

Conventional automatics do the same thing, they just have fewer ratios to choose from. My Jeep downshifts on downhill runs to use engine braking to maintain the set speed.

 

[futuristicmonkey]

Think of it like this. If I want to accelerate as fast as possible because I'm hauling rocks with a Corolla or I'm just trying to get up to highway speed and I don't want to die, where should the engine's RPM be? Some will say max torque but the real answer is max power.

The above statement doesn't represent any real-world situation nor is it useful to think in such terms. You don't exercise such control over your engine anyway (how would you accelerate if you maintained your RPM?). In order to accelerate as fast as possible you would go WOT. When you fully-open the throttle you are requesting the maximum torque of your engine at that given RPM.

The engine will be accelerating the fastest at its peak torque RPM even though it is putting out more power at its max power RPM. As your engine accelerates its torque characteristics change.

The torque output of the engine subtracting the torque requirements of the load (ie drivetrain friction, wind resistance, gravity if you're going up a slope) will determine the vehicle's acceleration.

If you were asking "if you wanted to maximize your instantaneous acceleration" the answer is at maximum torque. Here's a little thought experiment to show why:

Lets say you had two equivalent engines (ie equivalent torque curves) each connected to a dynamometer via a transmission. One engine is brought up to the RPM where it makes max torque and the other engine is brought up to where it makes max power. Suppose the transmissions connected to each engine are designed to have the dynamometers rotating at some given speed.

If you suddenly fully-open the throttle for each engine, which dynamometer do you think will be accelerating the quickest?

The max torque engine will accelerate the dyno the fastest. This is because the gear ratio for the other engine will be taller as well as the engine won't be able to produce as much torque.

 

[reallyscrued]

How do you figure?

Imagine you are riding a mountain bike up a hill. You downshift to make it easier to pedal, why?

Pedaling becomes easier because through a ratio of cogs, you have amplified the amount of leverage you have to turn your rear wheel (torque), however, now for each rotation of the pedals, the rear wheel turns a smaller distance, doing less work, therefore each rotation is effectively less powerful.

That's my understanding anyway.

 

[The Boston Dangler]

Imagine you are riding a mountain bike up a hill. You downshift to make it easier to pedal, why?

Pedaling becomes easier because through a ratio of cogs, you have amplified the amount of leverage you have to turn your rear wheel (torque), however, now for each rotation of the pedals, the rear wheel turns a smaller distance, doing less work, therefore each rotation is effectively less powerful.

That's my understanding anyway.

Power = Work / Time

in your example, Work is the turning of the rear wheel like you said, and Time is measured as the turning of the crank.

 

[Mike Gayner]

I'm still waiting for this system to pan out.

http://www.youtube.com/watch?v=F6zE__J0YIU

 

[zacmil]

Is there some kind of problem with variable transmissions that would cause so many car companies to avoid them like the plague? Cars like the Toyota Prius and Honda Civic Hybrid use variable transmissions, but every other Toyota and Honda non-hybrid uses a conventional fixed ratio automatic. Why is this?

Look into the Subaru Justy. It was a little, 3cyl. subcompact they put out in the late 80s and it had an option of a cvt. They seemed to have serious reliability issues with them and (to my knowledge) it was a pretty big failure.

Until recently, cvts were considered unsuitable for engines larger than about 1.2 liters, and, thus, were only really seen on scooters and snowmobiles. Despite being around for well over 100 years (I think the cvt came on the scene somewhere around the 1880s) cvt technology has only recently seen some real advances. To my understanding, most of the cvts equipped on cars today are perfectly fine and reliable.

It does, however, seem like most vehicles equipped with a cvt deliver a less than inspiring driving experience. I read a great example of this in a review about the Nissan Juke. The reviewer initially drove the cvt version and was less than impressed, but after reluctantly driving the manual, she decided that Nissan must have "hid the horsepower in the transmission". I do believe that the cvt is excellent for some sort of econobox that is just meant to get you from A to B. I do, however, fear that this may start to creep into cars that sometimes find themselves in the realm of the enthusiast. There's just no replacement for banging through the gears of a manual transmission.

 

[ShawnD1]

So basically my vision for the CVT is unrealistic because I'm asking for the equivalent of an 8 speed automatic in a cheap ass Sentra. Does that sound accurate?

I'm still waiting for this system to pan out.

http://www.youtube.com/watch?v=F6zE__J0YIU

bignate already mentioned this on the first page. Toyota calls it a Hybrid Synergy Drive. By using a gasoline engine and an electric motor, both power sources can be used at the same time and it can act like a variable transmission.

http://eahart.com/prius/psd/

The Toyota Prius is packed with some pretty high-tech stuff, but at the heart of the Hybrid Synergy Drive (HSD) is a simple little device called the Power Split Device, or PSD. The PSD is a planetary gear set that removes the need for a traditional stepped gearbox and transmission components, and also the familiar rev-lurch-rev-lurch of acceleration in an ordinary gas powered car. It acts as a continuously variable transmission (CVT) but with a fixed gear ratio.

What's really neat about this planetary gear system is that it's not using belts or chains or anything weak like that. It looks a lot more like a conventional automatic transmission; the kind of transmission that be hooked up to V8 engine without breaking.

Lets say you had two equivalent engines [two toyota corolla engines ] each connected to a dynamometer via a transmission. One engine is brought up to the RPM where it makes max torque [4000 rpm] and the other engine is brought up to where it makes max power [6000 rpm]. Suppose the transmissions connected to each engine are designed to have the dynamometers rotating at some given speed.

If you suddenly fully-open the throttle for each engine, which dynamometer do you think will be accelerating the quickest?

The engine at max power will accelerate faster because it has greater torque multiplication through the transmission.

Follow the thought experiment with numbers. Your first engine will run at max torque. Max torque in my toy car is at 4000rpm, and let's just say the gear ratio to maintain a certain speed is 1 engine rpm to 1 transmission rpm to keep it simple. My car's max power is at about 6000rpm. To make the dyno spin at the same speed, the gear ratio would then be 6000/4000 = 1.5.
The engine at maximum torque puts in 1 ft-lb and gets out 1 ft-lb.
The engine at maximum power puts in 1 ft-lb and gets out 1.5 ft-lb

Basically what you're asking is which gear accelerates faster. Does the car accelerate faster in second gear (high rpm at max power) or does it accelerate faster in third gear (mid rpm at max torque). IMO, my car accelerates fastest in lower gears.

 

[ViviTheMage]

Test drove a Subaru Outback with CVT today. Drove up some steep hills on the highway and CVT was really slick. Hope to test drive an Audi Q5 soon which has a 8 speed automatic.

mmmmmmmm, Q5, my next 'suv'.

 

[Zenmervolt]

How do you figure?

Original thought was:

HP = TQ * RPM/5252

To multiply torque, RPM goes down. Reduce RPM and you reduce power.

However, now that I think about it, the factor by which TQ increases is the same factor by which RPM decreases, so power should remain constant. Oops.

ZV

 

[thomsbrain]

They may continue to have a place in purely economy-oriented cars, but for every other class of vehicle, manufacturers and consumers have soundly rejected them in favor of regular transmissions with higher numbers of gears.

Why? They sound and feel terrible.

 

[Vic Vega]

Several manufacturers went with CVTs in the mid 2000s and have since gone back to traditional units. That should tell you something.

 

[Howard]

Oops.

ZV

 

[kornphlake]

Original thought was:

HP = TQ * RPM/5252

To multiply torque, RPM goes down. Reduce RPM and you reduce power.

However, now that I think about it, the factor by which TQ increases is the same factor by which RPM decreases, so power should remain constant. Oops.

ZV

The funny thing about that equation is that when RPM<5252 HP is always less than torque, when RPM>5252 HP is always greater than torque assuming torque is constant. Torque isn't constant at all RPMs, nor is horsepower. When you see torque vs horsepower curves they are actual measured values from a dyno, not calculated results, right, otherwise every curve would show the HP and TQ cross at 5252 RPM

 

[ShawnD1]

TWhen you see torque vs horsepower curves they are actual measured values from a dyno, not calculated results, right, otherwise every curve would show the HP and TQ cross at 5252 RPM

Actually that's exactly why every dyno chart does cross at 5252.

Any image you find, they always cross at 5252

 

[Howard]

The funny thing about that equation is that when RPM<5252 HP is always less than torque, when RPM>5252 HP is always greater than torque assuming torque is constant. Torque isn't constant at all RPMs, nor is horsepower. When you see torque vs horsepower curves they are actual measured values from a dyno, not calculated results, right, otherwise every curve would show the HP and TQ cross at 5252 RPM

Shawn's right about every dyno showing HP=torque at 5252RPM... there's no reason why it wouldn't be the case.

The math is what the dyno does anyway. It doesn't separately measure HP and torque.

 

[kornphlake]

Maybe I just found a poor example of a horsepower curve, this certainly does not cross at 5252 rpm

 

[overst33r]

Maybe I just found a poor example of a horsepower curve, this certainly does not cross at 5252 rpm

Because the scales are different

 

[kornphlake]

Because the scales are different

RPM is RPM for either scale It's not easy to find a torque curve that shows both HP and ft-lb, at least using google images, most are kw and nm or hp and nm or kw and ft-lb, different units make the graphs look substantially different.

Regardless of which graph is correct and which isn't, I believe the above equation is misleading. Consider an electric motor:

As RPM increases torque remains fairly constant but current decreases, current can be converted to watts by multiplying the current by the input voltage, which results in a curve that follows the current curve to zero at high RPM. KW can be converted to HP by multiplying by 1.34, so the horsepower curve would follow the KW curve to zero as RPM increases. The horsepower curve calculated by measuring the current looks very different than the curve calculated by converting torque to hp.

 

[IcePickFreak]

Horsepower is work performed over time. For mechanical (ie. rotational) horsepower it's 33000ftlbs/min, which is to say a draft horse, on average, could move 33,000 lbs one foot in one minute. It's strictly calculated from torque, and will always cross at 5252rpm as that's defined by (33,000 ft·lbf/min)/(2pi rad/rev) because torque is a twisting force. Keep in mind in engineering it's called a "moment" because there is no time defined in torque, hence the ft-lbs with no time unit - no work actually being done. They came up with horsepower to have a comparison of how much actual work steam engines (rotational force) could do relative to a single draft horse.

 

[ShawnD1]

RPM is RPM for either scale

Look at the Y axis. The dark line uses the right side and it's near 296 at 5252. The light line uses the left side and it's slightly below the 299 line at 5252.

Consider an electric motor:

As RPM increases torque remains fairly constant but current decreases, current can be converted to watts by multiplying the current by the input voltage, which results in a curve that follows the current curve to zero at high RPM. KW can be converted to HP by multiplying by 1.34, so the horsepower curve would follow the KW curve to zero as RPM increases. The horsepower curve calculated by measuring the current looks very different than the curve calculated by converting torque to hp.

Motor power is low for the left part of the graph because the motor's power factor is highly inductive at low speeds. The frequency of electricity going through the thing is thought of as f*s where f is frequency of input (usually 60Hz) and s is slip. From a dead stop, slip is 1, so frequency is 60. When the thing is running at 0.95x synchronous speed, the frequency of electricity in the armature is only 0.05x60 = 3Hz. The reason the frequency makes a difference is because the impedance of an inductor is directly proportional to the frequency. At 60Hz frequency, the circuit is strongly inductive, so large current does not mean large power. At something like 3Hz, the circuit is mostly straight resistance, the power factor is a lot better, so the real power is higher even though the current is less.

In the picture you posted, maximum power would be just a little higher rpm than the breakdown torque.

 

[Zenmervolt]

RPM is RPM for either scale

Yes, but when the Y-axis is different, the lines won't cross when they are equal. If you look at your graph, the lines are crossing at around 175 hp and 240 ft-lbs of torque (both at around 3,750 RPM) because the lines are using different Y-axis scales. So the lines cross, but the values aren't equal.

If you look at the values for 5252 RPM on that graph, you get about 299 ft-lbs and 299 hp. The values are equal at 5252 RPM.

Regarding electric motors, current is current consumed, not power put out by the motor. My laser printer pulls 700 watts, but that doesn't mean it's putting out 0.94 hp of work. At low RPM and high load, large amounts of the input current are turned into heat instead of useful work. It's a fundamental mistake to assume that KW in = KW out for an electric motor, especially under load.

ZV

 

[Capt Caveman]

Any ideas what kind of maintenance that a CVT requires? Belts/Chains need to be replaced at xxxxx miles?

I wonder if I go with a Subaru Outback with a CVT, will the CVT be something to worry about 5+ years down the road.