I'd like to use ultracapacitors for capturing regenerative braking energy and then assist the main battery pack under acceleration. The best ultracapacitors are not very energy dense, so they wouldn't work well for a main battery pack. However, they have very low internal resistance and can take a charge very quickly without that energy turning into heat.
They can capture about 80% of regenerative braking energy as opposed to less than 30% for most batteries. With that sort of efficiency they might just be worth the extra weight. It's something I'd like to experiment with in any case.
-Dave
electrathon said
Mar 15, 2010
As far as I know, they are currently allowed. They would have to be on the motor side of the controller and not the battery, so that you did not start the race with them charged.
Pomodori Corsa said
Mar 15, 2010
Ah, ok, I didn't see anything about them in the rulebook so I thought I'd ask to make sure.
Thanks!
electrathon said
Mar 15, 2010
I moved this post here hoping others might see it and get some ideas.
Cent-ECAR-man said
May 3, 2010
I didnt think there was any re-generation allowed in a race. I could be wrong though
electrathon said
May 3, 2010
Cent-ECAR-man wrote:
I didnt think there was any re-generation allowed in a race. I could be wrong though
Yes, re-gen is allowed. Regen stops the ability to freewheel, so most teams do not do it.
Anonymous said
Sep 12, 2010
Pomodori Corsa wrote:
They can capture about 80% of regenerative braking energy as opposed to less than 30% for most batteries.
In another thread on ReGen I posted a bit more technical data on ReGen. Lead-Acid batteries are actually about 80% efficient, boost caps are above 95% efficient, depending on the currents that you are using the batteries and boost caps at.
A good ReGen design would require placing the caps on the supply side of the motor controller, probably in parallel with the batteries, or switchable, or both. As the admin notes, they may require a means to verify they are discharged when the car leaves the pits for the staging area, so you are not trying to sneek a few more WHrs into the race. It would make better sense to put them AFTER the cut-off switch, so they will quickly discharge ... but there will be a LARGE in-rush when the switch is activated, so it will take some charge limiting circuit to keep from frying the switch and wiring as the boost caps will look like a dead short for a little while. More like, requiring a current limited pre-charge circuit before switching the battery on.
Other than that, boost caps can replace the 80% efficiency of the battery for a ReGen charge storage, with something better than 95% efficiency, greatly improving ReGen. They can also flatten out the current spikes on the lead acid batteries, providing more power to the driver on demand, and improving battery life.
Beware though, that these boost caps will explode violently if charged very far above their limit. So just like Li-Poly and LiFePO4 batteries, you will need the equivalent of a BMS to protect the caps from over voltage when they are placed in series. The energy released when they internally short/arc can be significant ... so use with VERY deliberate caution in an EA car.
It will take a fair number of boost caps to store the KE in a 300lb EA car traveling 60mph, or even 35mph for a parking lot race. But not out of the question at all. You will probably have to build your own high efficiency motor controller to do it really well. This is a great student physics problem to solve.
Anonymous said
Sep 12, 2010
electrathon wrote:
Cent-ECAR-man wrote:
I didnt think there was any re-generation allowed in a race. I could be wrong though
Yes, re-gen is allowed. Regen stops the ability to freewheel, so most teams do not do it.
Only for VERY stupidly design ReGen controllers. Just as accelleration current/force is controllable by a good motor controller, so should decelleration, or ReGen, current/force be controllable. Freewheel should also be an option, when the motor controller is in a control deadband as the car is at a specific selected speed. Something along the lines of cruise control, but including braking.
electrathon said
Sep 15, 2010
Anonymous wrote:
electrathon wrote:
Cent-ECAR-man wrote:
I didnt think there was any re-generation allowed in a race. I could be wrong though
Yes, re-gen is allowed. Regen stops the ability to freewheel, so most teams do not do it.
Only for VERY stupidly design ReGen controllers. Just as accelleration current/force is controllable by a good motor controller, so should decelleration, or ReGen, current/force be controllable. Freewheel should also be an option, when the motor controller is in a control deadband as the car is at a specific selected speed. Something along the lines of cruise control, but including braking.
I guess I should have said it this way: Regen stops the ability to use a bike freewheel.
Anonymous said
Sep 15, 2010
electrathon wrote:I guess I should have said it this way: Regen stops the ability to use a bike freewheel.
That may be true for many existing simple controllers that have a single control input ... the throttle.
A more advanced controller can/should implement an at-speed dead-band to coast, and only initiate accelleration or ReGen decelleration when the throttle is moved away from the current at speed dead-band, proportional to the difference between current speed and requested throttle speed. If ReGen can create panic braking forces, this algorithm has some safety problems with a single control, unless ReGen is limited to a safe decelleration rate.
Our first generation prototype controllers did this. Works ok, but the throttle is a little sloppy for speed control when the dead band is modestly wide. A narrow dead-band has a better feel ... more like what you get from a gas engine, using engine back pressure to decelerate, but requires actively backing the throttle off slowly to keep it in the dead-band as you coast down. ReGen proportional to the difference between actual speed and requested slower speed makes this easier to actually use a fairly narrow dead band, because small speed differences create a very small ReGen braking force.
A better design for the controller is with a second control input for braking. With two inputs you can coast when the throttle is released (possibly with a dead-band), and ReGen braking is limited and proportional to a light or modest brake control activation pressure. Harder braking requires a brake input.
Our current custom motor controllers design works this way, using a narrow dead-band on the throttle that is overridden with the brake. Backing off the throttle all the way will also ReGen brake, just not as strong as with a heavy brake ReGen activation. Riding experience is much closer to that of a motorcycle, but using ReGen to opportunistically recover KE without explict rider thought.
Since many storage and traction systems lack the efficiency to implement ReGen effectively, many designers haven't put enough thought into good designs for ReGen. As better batteries like LiFePO4's are common, plus high Joule capacitor systems getting cheaper, then pressure will be for traction motor systems to also become highly efficient, Then more companies will do ReGen systems better. Simply because it will be expected to be done right. Maybe in another year or so, EA teams will have access to that cheaply. Small EV systems being driven by the bicycle market will evolve quickly to improve range currently limited by expensive LiFePO4 battery systems - ReGen is an important part of improving range in most markets.
Anonymous said
Sep 15, 2010
If a traction motor and controller pair are both efficient and good at ReGen you should be able to size a Maxwell Boost Cap bank charged to about 2-3X the joules of the KE for your car/driver at some target speed - say 30mph. You should then be able to accellerate and decellerate to/from something near that target speed several times just on the energy stored in, and recovered to, the Boost caps. There will be losses, and after a few times, the peak speed will get slower as you start/stop, till there is no energy left to move at all.
I suspect the motors and best ReGen controllers available to EA teams today will be lucky if they reach the target speed even once, and are very likely to reach less than 1/3 of the target speed a second time using only Boost Caps sized to 3X the KE of their car/driver.
There are several reasons for this. The largest is that EA cars without transmissions will operate their traction motors at well below 50% efficiency for an extended period as they pull away from a stop.
It would be interesting to restructure a class of EA races to require a stop at the finish line every pass around the track to simulate stop and go city driving. A very different traction motor and controller design would be necessary to boost efficiency from dead stops, and do high efficiency ReGen properly.
Using a NuVinci CVP rear hub transmission is probably the quickest off the shelf fix today for this problem. It would allow quickly bringing the motor up into it's most efficient part of the power band, then using the CVP to bring the car up to the desired speed maintaining the most efficient motor RPM. With a good controller, then the CVP and motor speed can be adjusted to maintain peak efficiency depending on speed and loading. This is very similar in concept to what street cars to today for gas engines - using a CVP transmission to keep the engine in it's most efficient power band.
Adding a NuVInci CVP to many EA cars today would decrease the time to reach optimal speed, and decrease the energy consumed doing so, allowing more laps during the limited race time. It would also improve ReGen, by "down shifting" to steadily increase motor speed to decellerate, keeping the motors back EMF at/above the battery charge voltage and improving ReGen efficiency. At would also eliminate teams having to juggle chain drive sprocket ratios to fine tune efficiency for a particular track. Possibly improving typical performance by 10-20%.
electrathon said
Sep 21, 2010
I guess my answer to this is that the majority of the Electrathons in this area are very student built. We have issues getting the kids to tighten the battery bolts, no rocket science is being done here. Our hope is not that we necessaryly reinvent the wheel, but rather that we excite a student to a level that they will continue on in life so they may reinvent the wheel. The information you posted above is far above the design level of most (likely all) of the students we have racing in this area.
One thing that we do that you talked about though is we have road courses here. There is a lot of speeding up an dslowing down. It is one of the primary differances between us and the NE area. They like big circles, we like parking lots. We turn, speed up, avoid the other guys, slow down, change elevation, cross storm drains, etc. It is a much more real life situation then speed up turn left. Some people like one, some like the other. I for one would have lost all intrest in Electrathons if all we were doing was driving in big circles.
Aaron
Anonymous said
Oct 3, 2010
I very much understand. I've been coaching a FIRST FRC robotics team for several years, and appreciate the differences in experience and skill levels between various student teams. The last two years I've been working with mostly 9th and 10th graders, and scaling back expectations to match their skill levels. EA student teams have similar constraints.
My goal was only to provide an accurate engineering perspective on ReGen. It's a much tougher issue than just slapping a PWM boost circuit into the drive electronics, and expecting it to magically recover all the KE.
I'd like to use ultracapacitors for capturing regenerative braking energy and then assist the main battery pack under acceleration. The best ultracapacitors are not very energy dense, so they wouldn't work well for a main battery pack. However, they have very low internal resistance and can take a charge very quickly without that energy turning into heat.
They can capture about 80% of regenerative braking energy as opposed to less than 30% for most batteries. With that sort of efficiency they might just be worth the extra weight. It's something I'd like to experiment with in any case.
-Dave
Thanks!
Yes, re-gen is allowed. Regen stops the ability to freewheel, so most teams do not do it.
In another thread on ReGen I posted a bit more technical data on ReGen. Lead-Acid batteries are actually about 80% efficient, boost caps are above 95% efficient, depending on the currents that you are using the batteries and boost caps at.
A good ReGen design would require placing the caps on the supply side of the motor controller, probably in parallel with the batteries, or switchable, or both. As the admin notes, they may require a means to verify they are discharged when the car leaves the pits for the staging area, so you are not trying to sneek a few more WHrs into the race. It would make better sense to put them AFTER the cut-off switch, so they will quickly discharge ... but there will be a LARGE in-rush when the switch is activated, so it will take some charge limiting circuit to keep from frying the switch and wiring as the boost caps will look like a dead short for a little while. More like, requiring a current limited pre-charge circuit before switching the battery on.
Other than that, boost caps can replace the 80% efficiency of the battery for a ReGen charge storage, with something better than 95% efficiency, greatly improving ReGen. They can also flatten out the current spikes on the lead acid batteries, providing more power to the driver on demand, and improving battery life.
Beware though, that these boost caps will explode violently if charged very far above their limit. So just like Li-Poly and LiFePO4 batteries, you will need the equivalent of a BMS to protect the caps from over voltage when they are placed in series. The energy released when they internally short/arc can be significant ... so use with VERY deliberate caution in an EA car.
It will take a fair number of boost caps to store the KE in a 300lb EA car traveling 60mph, or even 35mph for a parking lot race. But not out of the question at all. You will probably have to build your own high efficiency motor controller to do it really well. This is a great student physics problem to solve.
Only for VERY stupidly design ReGen controllers. Just as accelleration current/force is controllable by a good motor controller, so should decelleration, or ReGen, current/force be controllable. Freewheel should also be an option, when the motor controller is in a control deadband as the car is at a specific selected speed. Something along the lines of cruise control, but including braking.
A more advanced controller can/should implement an at-speed dead-band to coast, and only initiate accelleration or ReGen decelleration when the throttle is moved away from the current at speed dead-band, proportional to the difference between current speed and requested throttle speed. If ReGen can create panic braking forces, this algorithm has some safety problems with a single control, unless ReGen is limited to a safe decelleration rate.
Our first generation prototype controllers did this. Works ok, but the throttle is a little sloppy for speed control when the dead band is modestly wide. A narrow dead-band has a better feel ... more like what you get from a gas engine, using engine back pressure to decelerate, but requires actively backing the throttle off slowly to keep it in the dead-band as you coast down. ReGen proportional to the difference between actual speed and requested slower speed makes this easier to actually use a fairly narrow dead band, because small speed differences create a very small ReGen braking force.
A better design for the controller is with a second control input for braking. With two inputs you can coast when the throttle is released (possibly with a dead-band), and ReGen braking is limited and proportional to a light or modest brake control activation pressure. Harder braking requires a brake input.
Our current custom motor controllers design works this way, using a narrow dead-band on the throttle that is overridden with the brake. Backing off the throttle all the way will also ReGen brake, just not as strong as with a heavy brake ReGen activation. Riding experience is much closer to that of a motorcycle, but using ReGen to opportunistically recover KE without explict rider thought.
Since many storage and traction systems lack the efficiency to implement ReGen effectively, many designers haven't put enough thought into good designs for ReGen. As better batteries like LiFePO4's are common, plus high Joule capacitor systems getting cheaper, then pressure will be for traction motor systems to also become highly efficient, Then more companies will do ReGen systems better. Simply because it will be expected to be done right. Maybe in another year or so, EA teams will have access to that cheaply. Small EV systems being driven by the bicycle market will evolve quickly to improve range currently limited by expensive LiFePO4 battery systems - ReGen is an important part of improving range in most markets.
I suspect the motors and best ReGen controllers available to EA teams today will be lucky if they reach the target speed even once, and are very likely to reach less than 1/3 of the target speed a second time using only Boost Caps sized to 3X the KE of their car/driver.
There are several reasons for this. The largest is that EA cars without transmissions will operate their traction motors at well below 50% efficiency for an extended period as they pull away from a stop.
It would be interesting to restructure a class of EA races to require a stop at the finish line every pass around the track to simulate stop and go city driving. A very different traction motor and controller design would be necessary to boost efficiency from dead stops, and do high efficiency ReGen properly.
Using a NuVinci CVP rear hub transmission is probably the quickest off the shelf fix today for this problem. It would allow quickly bringing the motor up into it's most efficient part of the power band, then using the CVP to bring the car up to the desired speed maintaining the most efficient motor RPM. With a good controller, then the CVP and motor speed can be adjusted to maintain peak efficiency depending on speed and loading. This is very similar in concept to what street cars to today for gas engines - using a CVP transmission to keep the engine in it's most efficient power band.
Adding a NuVInci CVP to many EA cars today would decrease the time to reach optimal speed, and decrease the energy consumed doing so, allowing more laps during the limited race time. It would also improve ReGen, by "down shifting" to steadily increase motor speed to decellerate, keeping the motors back EMF at/above the battery charge voltage and improving ReGen efficiency. At would also eliminate teams having to juggle chain drive sprocket ratios to fine tune efficiency for a particular track. Possibly improving typical performance by 10-20%.
My goal was only to provide an accurate engineering perspective on ReGen. It's a much tougher issue than just slapping a PWM boost circuit into the drive electronics, and expecting it to magically recover all the KE.
Have fun!
John