[brian_]

As I'm driving my Leaf, I am really wishing that there was a way to set a fixed amount of power being delivered to the motor.
...
This way there would be a gradual acceleration down hill, and a gradual deceleration uphill. Maybe even better would be if there was a way to specify the efficiency that we wanted to achieve?

That would be almost trivial to do... buy why? If you're thinking of keeping the motor at the same point in its performance map, and thus at constant efficiency, it's not likely that this one point will be more efficient than ramping the power up and down to maintain constant speed (except for allowing speed to pick up on steep descents rather than regenerative braking).

Think Captain Pecard: "Warp 2 - Engage."

Not really, because in the Star Trek world "Warp 2" is a set speed (Warp 1 is the speed of light and higher warp factors are faster, on a non-linear scale), not a set power level. The equivalent of Picard commanding Warp 2 is a driver pushing the "set" button on any cruise control.

Star Trek's model of ship control (for the starships, not shuttles and the like) resembles a traditional oceangoing ship (or even better, a submarine) rather than an aircraft. The captain's commanding of speed is an update to the traditional ship practice of calling for an engine speed, which effectively meant a power level.

It would function more like the throttle on an airplane. . . why am I feeling like this would be more appropriate for an EV? And potentially inspire a more efficient mode of driving? It's not so important exactly what speed your going this second. . . what's more important is how much power are you using as you get there~

Yes, aircraft use power levers (or push-pull knobs like an old choke control), but in the case of a basic piston engine they are just like a car's accelerator pedal, except that they stay where they're put rather than springing back to zero when released. In practice, advanced aircraft are flown most of the time on autopilot, which controls speed like an automotive cruise control.

I don't see how constant-power (rather than constant-speed) driving is any more appropriate for an EV than for an engine-driven vehicle. It actually seems likely to be the other way around, since a modern high-voltage AC motor in an EV has a reasonably broad range of efficient speed and load combinations. Yes, "hypermiling" enthusiasts drive at constant power and let the speed increase, then shut off the engine and coast, then repeat the cycle - this is incompatible with safe and effective driving.

I do like the idea of "soft" cruise control (allowing some speed variation with road grade), with sensible limits, but it would be less valuable for efficiency to a typical EV than to a typical automatic-transmission engine-driven vehicle.

 

[brian_]

I can already hear the safety nuts coming out . . . what about emergency breaking?!? Well, you hit the brake and you pull the throttle way back~!

I think any reasonable person would want any application of the brake pedal to cancel (return to zero) the power lever effect, whether by overriding the power lever signal, or by physically releasing it to spring back to zero. I wouldn't even want it just temporarily overridden, because I wouldn't want the set power level back when the brake pedal is released; think about how applying the brake pedal cancels cruise control in every stock vehicle.

There's no direct equivalent to the brake pedal in an aircraft*, so applying the aircraft control model here is inappropriate.

* Yes, I know that aircraft have wheel brakes, but they are not used in flight. I also know about spoilers and dive brakes, but they are not routinely used to control speed, and are not suddenly applied in an emergency.

 

[Meni Menindorf]

it's not likely that this one point will be more efficient than ramping the power up and down to maintain constant speed (except for allowing speed to pick up on steep descents rather than regenerative braking).

This is an excellent point.

Not really, because in the Star Trek world "Warp 2" is a set speed (Warp 1 is the speed of light and higher warp factors are faster, on a non-linear scale), not a set power level.

Of 'course! How could I forget this fact?? Glad to know I am not the only Star Trek fan in the room

The captain's commanding of speed is an update to the traditional ship practice of calling for an engine speed, which effectively meant a power level.

This is probably where I am taking my inspiration from. A sense of luxury and command, just setting the motor to a particular power output, and then kicking back and enjoying the ride. . . with less general concern about getting anywhere fast or at a particular speed..

I don't see how constant-power (rather than constant-speed) driving is any more appropriate for an EV than for an engine-driven vehicle. It actually seems likely to be the other way around, since a modern high-voltage AC motor in an EV has a reasonably broad range of efficient speed and load combinations. Yes, "hypermiling" enthusiasts drive at constant power and let the speed increase, then shut off the engine and coast,

Yes! You make an excellent point that any gains here would be far less than a petrol motor might have using the same theory. However. . . something about making the switch to electric has made me even more concerned about efficiency. It seems like an interface, or possibly auto-piloted speed could be designed to do something more like this "hypermiling" approach.

then repeat the cycle - this is incompatible with safe and effective driving.

This is somewhat debatable

I do like the idea of "soft" cruise control (allowing some speed variation with road grade), with sensible limits, but it would be less valuable for efficiency to a typical EV than to a typical automatic-transmission engine-driven vehicle.

Yes!! And very true about making more of a difference to combustion engine vehicles. But on our EV's. . . how amazing would it be to set the "cruise" to 6 mi/kwh? And the car could at least attempt to get this level of efficiency, at the sacrifice of speed/acceleration.

I think any reasonable person would want any application of the brake pedal to cancel (return to zero) the power lever effect, whether by overriding the power lever signal, or by physically releasing it to spring back to zero. I wouldn't even want it just temporarily overridden, because I wouldn't want the set power level back when the brake pedal is released; think about how applying the brake pedal cancels cruise control in every stock vehicle.

This is all very true, and I totally agree. It is kind of hillarious to think about the fact that most commercial EVs are designed to give a small amount of power to the drive-train simply be releasing the brake. . . in order to resemble an automatic-transmission engine vehicle.

 

[PStechPaul]

A "set and hold" lever as you propose might be appropriate for a vehicle that does not require much quick adjustment of speed, like a tractor, train, boat, or plane (except fighters and aerobatic types). Many of those vehicles are actually speed-controlled using governors, so that they automatically adjust the throttle according to need.

It would certainly be possible to design an algorithm to maximize efficiency, but it would need additional parameters, such as minimum and maximum allowed speed, which would depend on terrain, wind, speed limits, traffic, and other factors.

I made an interface that allowed me to use a joystick to control an induction motor via a VFD.

 

[brian_]

A "set and hold" lever as you propose might be appropriate for a vehicle that does not require much quick adjustment of speed, like a tractor, train, boat, or plane (except fighters and aerobatic types)...

Agreed, but even fighters and aerobatic aircraft use "set and hold" power levers, right?

 

[Meni Menindorf]

A "set and hold" lever as you propose might be appropriate for a vehicle that does not require much quick adjustment of speed, like a tractor, train, boat, or plane (except fighters and aerobatic types). Many of those vehicles are actually speed-controlled using governors, so that they automatically adjust the throttle according to need.

It would certainly be possible to design an algorithm to maximize efficiency, but it would need additional parameters, such as minimum and maximum allowed speed, which would depend on terrain, wind, speed limits, traffic, and other factors.

I made an interface that allowed me to use a joystick to control an induction motor via a VFD.

That is rad! Yes, that is what I'm talking about . . . though, of 'course that joy-stick has a return-to-center feature, which removes the "cruise-control" feeling I'm going for. If we could get vehicle manufacturers to start developing "efficiency-cruise" algorithms using the parameters you mention. . . it seems like people would love it! As we are hot on the trail of autonomous vehicles, it seems like this would be a natural feature of their driving algorithm? We wouldn't want bot-drivers that just hold the same speed all the time. . . we would want them to vary speed on corners and hills to maximize efficiency, within reason~

 

[brian_]

We wouldn't want bot-drivers that just hold the same speed all the time. . . we would want them to vary speed on corners and hills to maximize efficiency, within reason~

Maybe we want them to coordinate speed with other vehicles for most efficient traffic flow, not to drive considering only their own individual efficiency.

 

[Meni Menindorf]

Maybe we want them to coordinate speed with other vehicles for most efficient traffic flow, not to drive considering only their own individual efficiency.

Yes! Exactly!


This is all making me wonder about efficiency rather deeply. I know the improvements we get following electric motor efficiency curves is nill compared to their petrol counter-part, but . . . it still just makes me curious!


Is there an ideal speed to drive to hypermile in an EV? All conditions being the same (flat, no-wind, etc~)? I am assuming it depends on some lower peak of the efficiency curve of the motor that would place the car around 25-30 mph? But is this just a bad guess based on my experience with petrol vehicles?



According to this, my Nissan Leaf should get the best efficiency at around 5000 RPM? Maybe on up to 6000?

So then ideal cruising speed would be around 65 mph. . . . but wait! That makes sense for our crazy horrible society where getting there faster is always better. . . Excuse my American sentimentality. But there is so much wind resistance at that speed! And it's totally unsafe to drive that fast. I would much rather travel between 30 - 45 mph most of the time.



Should I be re-gearing my rig???? LOL!!


I think the idea of a fixed efficiency control system feels luxurious to me BECAUSE the speed would vary. . . as I glide up a hill the vehicle would naturally slow down a bit, and give me a sense of being in no rush. This way when it tipped over the top of the hill, it would gradually start to gain momentum again up to a safe and efficient driving speed. It is reminiscent of being a rich person with a chauffeur. Personally I would much rather have a chauffeur with a relaxed attitude, so that I always felt at ease while traveling



OK, I'm going to really get off-topic for a moment. But brian's comment inspires me. As long as we are coordinating all of our traffic flow on an algorithm for efficiency of energy use (which seems like it would be fascinating to engineer, but also, disappointingly simple?) Then we might as well put all these travelling-units into a different and more efficient infrastructure grid like a hyperloop type system (but more dynamic, with small letter sized tubes which could merge/nest/cluster with larger/faster tubes?) where we don't have to deal with nearly as much wind resistance.

Then I can imagine a world where it was rather inexpensive to travel or ship anything anywhere, and the cost factor would simply be how fast you wanted it to get there. The bigger the hurry, the more it's going to cost to move. So it would be something like "Ok Google, I want to be downtown in an hour." And if this required a reasonable/efficient speed for the system, it would be extremely inexpensive, but there would be sharp pricing curve as we asked for shorter travel times.

 

[brian_]

Is there an ideal speed to drive to hypermile in an EV? All conditions being the same (flat, no-wind, etc~)? I am assuming it depends on some lower peak of the efficiency curve of the motor that would place the car around 25-30 mph? But is this just a bad guess based on my experience with petrol vehicles?

According to this, my Nissan Leaf should get the best efficiency at around 5000 RPM? Maybe on up to 6000?
http://www.electricvehiclewiki.com/images/a/a1/LEAFPowertrainEfficiencyCurve.jpg

So then ideal cruising speed would be around 65 mph. . . .

The highest efficiency area for that inverter and motor is about half way up the motor speed range, but it's not as simple as driving at that speed. As the graph shows, motor efficiency depends on both speed and load. If you at the optimal point for efficiency you're at essentially the full torque that the motor can produce at that mid-range speed (peaking at perhaps 6,000 rpm), so you're producing far more power than it takes to drive a Leaf at that road speed. If you really do drive at a steady highway speed you're way down on the torque scale, and far off the optimal efficiency point.

The curved top of the shape of that chart is a line of constant power (remember that power = torque X speed), at 80 kW. You can draw a whole set of those lines to make a third scale (e.g. at 10 kW, 20 kW, 30 kW, etc).

For each road speed, you can determine the output power (or torque) needed to maintain that speed (such as by coast-down testing) and put a mark on the graph - the line connecting them is the set of conditions that correspond to steady-speed flat road driving. You can choose the point along that curve where the colour indicates that the powertrain is most efficient, and that might be a good speed to drive, but it won't likely provide the lowest energy per distance travelled. For each speed along that line, you can take the output power, divide it by the efficiency to find the input power, and divide the input power by the road speed to get energy per distance. The speed for which the energy per distance is lowest is the most efficient traveling speed.

. . . but wait! That makes sense for our crazy horrible society where getting there faster is always better. . . Excuse my American sentimentality. But there is so much wind resistance at that speed! And it's totally unsafe to drive that fast. I would much rather travel between 30 - 45 mph most of the time.

Should I be re-gearing my rig???? LOL!!

It is unlikely that the speed for lowest energy consumption per distance travelled will be as high as 65 mph with stock gearing. Yes, the reason is aerodynamic drag.

While 30 to 45 mph is far to slow for highway travel, the most efficient speed for a Leaf might be even lower than that.

 

[PStechPaul]

For an electric motor vehicle, the most efficient speed is likely to be close to zero, whereas for an ICE that needs to idle, this would not be the case. Also, an ICE tends to be most efficient at a specific point in its power/torque/speed range, while an electric motor's efficiency is more determined by losses compared to output power.

Most of an electric motor's losses are due to resistance and current, so they would increase by the square of the current through the resistance. A larger motor would have less resistance, so if used at the same power (or torque), it may have lower losses and thus greater efficiency. Conversely, a larger ICE would likely be less efficient at the same power level.

A larger inverter or DC controller would also likely have lower losses when operated at a smaller percentage of rated power.

As speed increases, losses due to aerodynamic drag and drive train friction increase exponentially, so lower speed is favorable to efficiency. If time is not a major consideration, low speed is a viable option, but usually there is a time-cost factor, which might figure into real-world efficiency.

 

[brian_]

For an electric motor vehicle, the most efficient speed is likely to be close to zero, whereas for an ICE that needs to idle, this would not be the case.

You might expect that for the electric motor, but as the example for the Leaf shows, that's not true at all. Low speed is death to efficiency... although "low" for a modern motor is still 85% (if you believe that's really the bottom of the efficiency scale in this graph).

Yes, engines need to be running at a decent speed to justify the mechanical friction, and in the case of spark-ignition throttle-controlled engines they need to be at high load. On the other hand, too fast is too much internal aerodynamic loss, so moderately speed at high load is ideal. The usual rule of thumb is to run at whatever speed gives peak torque, at or near wide-open throttle.

Also, an ICE tends to be most efficient at a specific point in its power/torque/speed range, while an electric motor's efficiency is more determined by losses compared to output power.

Both types have a substantial range of near-optimal efficiency. I think the big difference is how badly efficiency drops off when away from the range for an engine, while well off the best conditions still gives decent efficiency with a modern electric motor.

Most of an electric motor's losses are due to resistance and current, so they would increase by the square of the current through the resistance. A larger motor would have less resistance, so if used at the same power (or torque), it may have lower losses and thus greater efficiency.

I get the logic, but it doesn't prove to be quite true. Again using the Leaf example, if you used the 80 kW rated Leaf motor at half power to replace a 40 kW rated motor, the Leaf motor would never be operating within a few percent of the peak efficiency.

I mentioned that a grid of power level could be added to the Leaf graph; I'll attach an example, without the efficiency colours but showing the shape of the maximum power of the Leaf motor, and the torque-speed combinations to produce various power outputs. Staying at low power means never seeing the motor operated at peak efficiency.

Conversely, a larger ICE would likely be less efficient at the same power level.

There's always an optimum level. If the larger engine is too large, it will be inefficient... but if it is just large enough to run in the optimum range and replaces a small engine which is working too hard, efficiency improves. This is readily apparent in fuel economy of vehicles towing trailers, but in typical use most cars and light trucks are so massively over-powered that yes, a larger engine would be even worse.

As speed increases, losses due to aerodynamic drag and drive train friction increase exponentially, so lower speed is favorable to efficiency.

Aero drag force (and therefore energy needed per unit distance) varies as the square of velocity, but friction drag is relatively constant, so the energy needed per unit distance is relatively insensitive to speed. By highway speeds, aero drag dominates and the cost of higher speed is readily apparent.