Torque curve inflection point

[E-HP]

When accelerating from stop to full speed with my rear direct drive motor, I can feel a point where the acceleration suddenly decreases and acceleration is slower for the last few miles per hour to reach top speed. Running the Grin motor simulator, I can see this graphically, where the torque curve, as well as the power and efficiency curves, experience an inflection point at a certain speed then drops off more dramatically.

In this example, it happens at 21mph:
https://ebikes.ca/tools/simulator.html?motor=MGRIN2708_SA&batt=B5220_GA&cont=C25&hp=0&axis=mph

My question is, what is happening at that speed that causes things to suddenly change?

I noticed by allowing the controller to provide more current, that the inflection point moves to a lower speed, for all three curves:
https://ebikes.ca/tools/simulator.html?motor=MGRIN2708_SA&batt=B5220_GA&cont=C40&hp=0&axis=mph

Continuing to increase current leads to a smoothing of the power curve, but the torque curve still exhibits the inflection point, at an even lower speed, or 2mph in this example:
https://ebikes.ca/tools/simulator.html?motor=MGRIN2708_SA&batt=B5220_GA&cont=cust_100_200_0.03_V&hp=0&axis=mph

Finally, there appears to be a point where the torque curve smooths out completely, along with the other curves:
https://ebikes.ca/tools/simulator.html?motor=MGRIN2708_SA&batt=B5220_GA&cont=cust_200_400_0.03_V&hp=0&axis=mph

 

[AHicks]

Radio control modeling (they use battery powered miniature 3 phase motors for some applications) has taught me you can use a straight line for power input to a motor (linear), yielding a curve on the motor's output, OR, you can use a curve on your voltage input (exponetial values) to get a straight line on the motor's output. Your call! Make yourself crazy!!!

In radio control, we have the ability to build a curve on our throttle sticks. We can install different voltages of our choice at 5 or more points along a line that represents our stick travel, building a curve to provide straight line/linear motor response, or modify anywhere along that line as required for our purposes.

KT controllers, I recently read somewhere, use 5 exponentially based percentages for their 5 PAS levels. Pretty sure it's one of the reasons their PAS software works so well. Check them out for an idea of what the hell I'm talking about. PAS 1 is set for 13% of available power, Pas 2 is 20%, Pas 3 is 33%, PAS 4 is 50%, and PAS 5 is 100%. Those PAS level values provide for fairly linear motor output.

Messing with a Bafang Ultra purchased recently, and decided to try those same KT values. Worked REALLY well, providing PAS values that yield a smooth power increase with no holes or big jumps between PAS levels going up or down. You can see this easily with the real time watt meter built into the display.

If you need 9 values, install the KT values at PAS 1,3,5,7,and 9, then split 1 and 3 to get the value for 2, 3 and 5 to get 4, etc. This will maintain the expo for the curve. FWIW, -Al

 

[ZeroEm]

Was talking about what you are seeing in another thread. My trike is set up where the drop off point is at my speed limit. My trike will go 5 mph faster if I take off the speed limit but I will feel what you feel the smallest grade and I will lose speed.
If you want to ride at 25 mph you need a Lower T motor or more Volts!

This is my trikes curent set up. My speed liimit is at 46 kph.
https://ebikes.ca/tools/simulator.h...AC&cont=C40&kv=6.56&frame=semi&mass=140&hp=25

I'm going to swap out my 7T for a 5T and move that point past where I ride at.

https://ebikes.ca/tools/simulator.h...&cont=C40&frame=semi&mass=140&hp=25&wheel=26i

It should have a nice steady pull, but will lose some torque.

 

[xfrankie]

Take a good look at the numbers below the graph. Then try to move the cursor line and observe what is happening.

As I understand, there are 3 main limiting factors:
1) phase amps (motor amps) limit
2) battery amps limit
3) bEMF rising with motor rpm

Good example is with PhaseRunner (hot), where you can all three limits in action:

https://ebikes.ca/tools/simulator.html?motor=MGRIN2708_SA&batt=B5220_GA&cont=PR_H&hp=0&axis=mph

First you have a flat torque line (blue line), because you are running into the motor amps limit. As the speed increases, so does the demand for power, to keep up with increasing RPM while at constant torque. Up until the 10mph mark.

When the blue line falls off is when you hit the Battery amperes limit, the battery side cannot supply more current to match the increased power demands. Thus torque must fall off as RPM increases to keep the power constant. Up until the 15mph mark.

Then is the third limit (second inflection point) where the RPM of the motor generates enough "backwards-pushing" voltage, that the controller cannot overcome it in order to keep the high amps flowing.
If you increase the amp limit, this inflection shift to the left. Because to keep the high amps you must supply high enough "voltage difference". And any motor movement (ie. bEMF) will cut away from this "voltage difference", thus limiting the possible amps.
There the consumed power falls off as a whole, because you simply cannot generate enough amps because of the limited voltage difference.

In your last example, the requested amp limit is so insanely high, that **any** motor movement will decrease the flowing amps. Thus giving a "smooth" curve. Essentially shifting the point to the left so far, that it moves beyond the left end of the graph.
If you increased the battery voltage, while keeping the amp limits from your last example, the inflection point will appear again.

EDIT: here is the increased voltage example:
https://ebikes.ca/tools/simulator.html?motor=MGRIN2708_SA&batt=cust_140_0.2_8&cont=cust_200_400_0.03_V&hp=0&axis=mph

 

[serious_sam]

As I understand, there are 3 main limiting factors:
1) phase amps (motor amps) limit
2) battery amps limit
3) bEMF rising with motor rpm

This.

 

[ZeroEm]

by E-HP » Oct 20 2020 2:59pm

When accelerating from stop to full speed with my rear direct drive motor, I can feel a point where the acceleration suddenly decreases and acceleration is slower for the last few miles per hour to reach top speed. Running the Grin motor simulator, I can see this graphically, where the torque curve, as well as the power and efficiency curves, experience an inflection point at a certain speed then drops off more dramatically.

Why did you get away from 72V?

 

[E-HP]

Take a good look at the numbers below the graph. Then try to move the cursor line and observe what is happening.

Thank You! Your explanation explains all of my observations and some insights into what can be done if you want to improve a particular aspect of performance within the power band. The battery current limit is particularly insightful. I never noticed that it remains constant between the first and second inflection points. I'm assuming that based on how the simulator works, the battery current is attained through the controller parameters instead of by selecting a battery from the drop down, or at least that was seems to be the case. It also looks like by raising the current, although the inflection point moves, the available torque is greater regardless, but does nothing to help available torque after the original (second) inflection point, with the solution being to raise the voltage.

 

[E-HP]

Why did you get away from 72V?

I'm not really getting away from it. My original planning for upgrading my battery pack started a while back. I wanted the main pack to provide more current, but after testing with various voltages, I decided to stick with 14S for normal riding (e.g. riding to the train station during the week, etc.), and strap on the lipo packs for higher voltage for fun during the weekends. Because of the lockdowns, the latter is the norm now, with the purpose being fun, exercise, and escape from "working from home", so I never even tried my new pack without the lipos.
I may be optimistically thinking that I can eventually go back to plan A, so I'm trying to maximize my main pack performance, after being spoiled by running at 72V, and specifically the torque/acceleration that's available after the second inflection point on the graph, where the voltage matters.

 

[ZeroEm]

I'm getting older so speed is still exciting but all this wisdom tells me I don't need or want it. 28 mph is sorta legal and plenty fast on a bike or my trike. I am looking for distance but want my trike to hold the speed and past the inflection point it slows on the hills.
I'm just want to move that point up 5 mph. Don't care about the extra speed or spinning my wheels.

 

[E-HP]

I'm getting older so speed is still exciting but all this wisdom tells me I don't need or want it.

I don't really need to speed either, just want better acceleration past the inflection point, all the way to the old top speed on my slow wind motor, 23 mph @ 52V. But, at 52V there is very little acceleration between 17 mph and 23 mph. I was able to tweak the Cycle Analyst settings this morning to get better response running 52V, since I had messed them up so much that there was virtually no response in the last half of the throttle range. After tweaking the response was way better. I then undid the current and power limits to see how much the motor would pull from the pack. 78A leaning way over the handlebars. Under the pack's max current, but over the 40A continuous, and noticeably more responsive. I decided to leave it that way and just trust my throttle hand to keep from exceeding the continuous rating.

But, when I added my lipo packs to get to 72V, I was surprised by how the new setting affected the mid range responsiveness. The throttle was responsive in any part of the range. Now when cruising at 17 mph and pegging the throttle, the response is enough to straighten out my arms, really noticeable. It was also pulling the wheel off the ground when doing the same at 10 mph; kind of crazy. It's fun to ride this way, but with all my hot rodding around, I was getting 2 miles/Ah, per the CA readout! I guess power takes power. I think if I were sticking to my original goal, I could set the max speed to 23 mph, and have powerful acceleration from zero to 23, but going fast occasionally is a nice option too.

 

[ZeroEm]

I speed across intersections at times when the light turns green and when a left is needed and ride with the cars then speed is needed. My trike will get up to 20 before I can make completely make a left turn. 15 mph is about all my trike can do without sliding the front tires. Tried some narrow front tires, increased my speed and get better watts per mile but not enough traction.

 

[E-HP]

I speed across intersections at times when the light turns green and when a left is needed and ride with the cars then speed is needed. My trike will get up to 20 before I can make completely make a left turn.

I was curious, so I charged my batteries up to ~81V to see what my 0-20mph times look like for different settings. I have my CA set to unlimited power/current but using the CA to control the ramp at 2V/sec, for the "Off Road" setting. My Unlimited setting is the same, but with bypass throttle mode, which is pretty unusable, but what should give me the best launch. I leaned over the front bars, then pegged the throttle. With the Unlimited setting, with the controller's 3-position switch set to normal/middle, I hit 20 mph with 3 consecutive launches in 2.7 seconds. Doing the same in Off Road mode, my times were 3 seconds.
Now the crazy part, by leaning forward in order to put the power to the ground, I pulled an AMax of 104A . I won't be doing that anymore, even if it only hits that for a millisecond, because it probably hammers my main pack, although the lipos likely handle it in stride. The motor was quite warm after 6 launches, so even more reason to keep my settings tame. Seems like for normal riding, setting the 3-positions switch to low is probably the way to go for long term survival of my motor and to have decent range.

 

[ZeroEm]

by E-HP » Oct 27 2020 3:55pm

ZeroEm wrote: ↑Oct 21 2020 8:50pm
I speed across intersections at times when the light turns green and when a left is needed and ride with the cars then speed is needed. My trike will get up to 20 before I can make completely make a left turn.

I was curious, so I charged my batteries up to ~81V to see what my 0-20mph times look like for different settings. I have my CA set to unlimited power/current but using the CA to control the ramp at 2V/sec, for the "Off Road" setting. My Unlimited setting is the same, but with bypass throttle mode, which is pretty unusable, but what should give me the best launch. I leaned over the front bars, then pegged the throttle. With the Unlimited setting, with the controller's 3-position switch set to normal/middle, I hit 20 mph with 3 consecutive launches in 2.7 seconds. Doing the same in Off Road mode, my times were 3 seconds.
Now the crazy part, by leaning forward in order to put the power to the ground, I pulled an AMax of 104A . I won't be doing that anymore, even if it only hits that for a millisecond, because it probably hammers my main pack, although the lipos likely handle it in stride. The motor was quite warm after 6 launches, so even more reason to keep my settings tame. Seems like for normal riding, setting the 3-positions switch to low is probably the way to go for long term survival of my motor and to have decent range.

My description may look grander than it really is, have my max power set to 2600w, depends on charge but is below the 36A max have set for my batteries for longevity. It takes me more than 3 seconds to get around the turn. I ride on the outside of the turn lane as not to cross traffic, this is the long track around a corner and try not to slide into the curb. I'm also peddling hard, riding a trike don't need to worry about lifting the front just not spin the rear which is not an issue on dry, clean pavement.

20 mph under 3 sec is fast is it not for bike. I know some here at ES talk about leaving cars at the stop light... Have thought about time trials just to see what it does.

I have set my ramp up different than yours, will need to look at it. Think it is like a Volt per .03/sec so it hits it peak just over a sec but does not jerk.

 

[goatman]

will your controller let you put about 20amps into the field weakening to boost top speed

 

[ZeroEm]

Grinfineon 40A but have it limited to 36A can take off all the power limits and max out the controller. Going to swap out the 7T leafmotor for a 5T Edge. This will up my top end but don't think it will help my 0-20mph times. Better get an acceleration test before the swap so I can have some info on it.