Page 2 of 3

Posted: Mon May 18, 2009 8:00 pm
by Jeremy
andylaurence wrote:My thinking is that the batteries are a limiting factor. Even at 40mph, whilst only 10hp may be needed to maintain speed, a much larger amount will be needed to overcome the grip of the tyres. That will be limited by the batteries, unless a large number of them are used.


But I've just shown that, even for cheap batteries, this simply isn't the case. For acceleration equivalent to an ICE car you need maybe 3 times the mean power, which equates directly to 3 times the battery discharge rate. If you are running the batteries at 3C for the mean power, then this means you are running them at 9C maximum for the peak power. As this is much less than your notional 30C cells, then the batteries quite clearly aren't the limiting factor.

In practice, the motor will only have a mean to peak rating of perhaps 2.5 to 3:1. Heat in the motor will most probably be the thing that limits performance, rather than the discharge capability of the batteries. Motor heating isn't a function of power, it's proportional to torque, or more accurately, directly proportional to the cube of the current. As current is linearly proportional to torque for most motors, then it's easy to see that power isn't really that relevant.

andylaurence wrote:You need both. They are directly linked by definition. Power is equal to the torque and the speed. As such, higher torque for a given road speed equals more power. Yes, that means it takes less power at lower speeds, however, it also means large amounts of power at higher speeds.


Power equals torque x rpm, so for the case of a standing start, with the wheels loaded to the maximum torque that they will take without exceeding their ability to maintain grip, at the instant of application of throttle, the power is zero. As the vehicle accelerates the power increases to that needed to maintain the chosen speed, but the rate of acceleration is wholly dependent on the torque that is available (and can be used), not the maximum power.


andylaurence wrote: That to me says I need lots of power to make a fast EV and that's what I'm planning (comparatively). Perhaps I'll revisit this when I have got to the point where the drivetrain needs to be powered.

Cheers,
Andy


To match the full throttle top speed on the level with electric power will require the same power as the ICE produces, but this size of electric motor will produce massive levels of acceleration when compared to the ICE version. You need to decide what you actually want, then design for it, taking account of the very big difference in torque characteristics between the two types of motor.

The big challenge with high power electric motors is torque control. Even the heavy, low powered, Prius has much, much more motor torque than the front wheels can handle, to the extent that the electronic torque control is active all the time to stop the wheels from spinning, even under very modest throttle settings.

Jeremy

Fast Car

Posted: Tue May 19, 2009 2:07 am
by Deker
Andy.

I feel that you should slow down a lot, have a considered look at what your'e trying to achieve.

I understand what your'e getting at with flywheel/s, however, Jeremy is quite right, let the batteries do the work.

How are you going to power the flywheel/s ?

You want high acceleration ? then you will need a torquey motor, one that won't overheat when pushed hard at low RPM/s.
I might be wrong, but I think those small model motors would become very ineficient heaters before they catch fire.
In the use for which they are designed, they have a big fan in front.

Then there is control, consider the power a controller would have to handle, them big-uns cost.

Perhaps 3 phase may be best for your needs.

Deker.

Posted: Tue May 19, 2009 9:52 am
by andylaurence
Jeremy wrote:
andylaurence wrote:My thinking is that the batteries are a limiting factor. Even at 40mph, whilst only 10hp may be needed to maintain speed, a much larger amount will be needed to overcome the grip of the tyres. That will be limited by the batteries, unless a large number of them are used.


But I've just shown that, even for cheap batteries, this simply isn't the case. For acceleration equivalent to an ICE car you need maybe 3 times the mean power, which equates directly to 3 times the battery discharge rate. If you are running the batteries at 3C for the mean power, then this means you are running them at 9C maximum for the peak power. As this is much less than your notional 30C cells, then the batteries quite clearly aren't the limiting factor.


Yes, you've shown that if you are running the batteries at 3C for mean motor power, then peak will be 9C. However, what if mean power is 20C? I'm not planning to commute to work in this car, I'm planning to get from A to B as quick as possible.

Jeremy wrote:In practice, the motor will only have a mean to peak rating of perhaps 2.5 to 3:1. Heat in the motor will most probably be the thing that limits performance, rather than the discharge capability of the batteries. Motor heating isn't a function of power, it's proportional to torque, or more accurately, directly proportional to the cube of the current. As current is linearly proportional to torque for most motors, then it's easy to see that power isn't really that relevant.


OK, that makes sense. Afterall, you size cables fo the current, not the voltage and power is current x voltage. I've just taken a look at the official rating of the components and made a blind assumption that they are true. I'd not considered that those ratings assume some cooling effort. Clearly, they will need cooling (they're just wasting electricity as heat), yet I had assumed that sitting in the wheel well would be good enough. It's certainly a breezy area!

Jeremy wrote:
andylaurence wrote:You need both. They are directly linked by definition. Power is equal to the torque and the speed. As such, higher torque for a given road speed equals more power. Yes, that means it takes less power at lower speeds, however, it also means large amounts of power at higher speeds.


Power equals torque x rpm, so for the case of a standing start, with the wheels loaded to the maximum torque that they will take without exceeding their ability to maintain grip, at the instant of application of throttle, the power is zero. As the vehicle accelerates the power increases to that needed to maintain the chosen speed, but the rate of acceleration is wholly dependent on the torque that is available (and can be used), not the maximum power.


Whoops! There's a typo to make. I meant to say that power is equal to torque x speed (RPM). It's a rate of work done. My point that I was trying (rather unsuccessfully) to make is that taking speed as fixed, it doesn't really matter whether you're thinking in Nm, lb-ft or bhp.


Jeremy wrote:
andylaurence wrote: That to me says I need lots of power to make a fast EV and that's what I'm planning (comparatively). Perhaps I'll revisit this when I have got to the point where the drivetrain needs to be powered.

Cheers,
Andy


To match the full throttle top speed on the level with electric power will require the same power as the ICE produces, but this size of electric motor will produce massive levels of acceleration when compared to the ICE version. You need to decide what you actually want, then design for it, taking account of the very big difference in torque characteristics between the two types of motor.

The big challenge with high power electric motors is torque control. Even the heavy, low powered, Prius has much, much more motor torque than the front wheels can handle, to the extent that the electronic torque control is active all the time to stop the wheels from spinning, even under very modest throttle settings.

Jeremy


If the Prius struggles with grip so much, why does it take more than 10 seconds to complete a 0-62 dash? Agreed, it probably does struggle to maintain grip at low speeds, yet it will struggle to spin the wheels at higher speeds. The torque from the motor hasn't changed during this time, but the power required to spin the wheels has...

Cheers,
Andy

Re: Fast Car

Posted: Tue May 19, 2009 10:05 am
by andylaurence
Deker wrote:Andy.

I feel that you should slow down a lot, have a considered look at what your'e trying to achieve.

I understand what your'e getting at with flywheel/s, however, Jeremy is quite right, let the batteries do the work.

How are you going to power the flywheel/s ?

You want high acceleration ? then you will need a torquey motor, one that won't overheat when pushed hard at low RPM/s.
I might be wrong, but I think those small model motors would become very ineficient heaters before they catch fire.
In the use for which they are designed, they have a big fan in front.

Then there is control, consider the power a controller would have to handle, them big-uns cost.

Perhaps 3 phase may be best for your needs.

Deker.


It seems to me like the small motor idea may not be as viable as I had thought it might be. My thoughts on the flywheel were simply that it would be attached to a motor. When power is applied to to the motor, it spinsthe flywheel. Essentially, regenerative braking of that motor supplies power back again. It's a short term, high power, low capacity electrical storage device. The advantages being that it can produce large amounts of power in a short space of time and that it doesn't suffer from being constantly "recharged".

Perhaps I'm best off going back to my original plan of two larger motors in the engine bay attached to the front wheels via driveshafts. I saw the power density of the little motors and thought they seemed ideal for cutting weight, both of the motors and of the drivetrain (by mounting them on the hubs). Of course, they still may be suitable as they won't see more than 90 seconds of use at one time.

Cheers,
Andy

Posted: Tue May 19, 2009 10:35 am
by microman
Andy, I think Deker is right, you need to consider carefully what your specification actually is. You say you want 10 mile range and you might be operating the cells at 20C average. 20C corresponds to 2 or 3 minutes discharge time. For 10 miles that's an average speed of 200-300 mph. And you want extra acceleration as well? In a Mini (even with 4WD)?

If you really do want to augment the peak battery capability I would suggest that supercapacitors are a easier (and much safer) route than flywheels.

Off topic really but as a Prius driver I think Jeremy's comment is correct in that the traction control is always active, but I don't believe the Prius will lose grip accelerating from standstill on a good dry surface. In the wet or on a friable surface the traction control certainly can intervene.

Posted: Tue May 19, 2009 10:53 am
by andylaurence
microman wrote:Andy, I think Deker is right, you need to consider carefully what your specification actually is. You say you want 10 mile range and you might be operating the cells at 20C average. 20C corresponds to 2 or 3 minutes discharge time. For 10 miles that's an average speed of 200-300 mph. And you want extra acceleration as well? In a Mini (even with 4WD)?

If you really do want to augment the peak battery capability I would suggest that supercapacitors are a easier (and much safer) route than flywheels.

Off topic really but as a Prius driver I think Jeremy's comment is correct in that the traction control is always active, but I don't believe the Prius will lose grip accelerating from standstill on a good dry surface. In the wet or on a friable surface the traction control certainly can intervene.


OK, here's the requirements in full. I plan to use the car for AutoSolo and Sprint events.

AutoSolo involves up to 12 attempts at courses upto about 0.6 miles in length. Speeds are limited to about 50mph at the faster events although the last round ranged from 13mph to 38mph (wheel speeds are higher, due to wheelspin). Admittedly, outright power will be less important here! The key thing about AutoSolo is that the car must be driven to events. This won't be feasible if I go to far away events (the last event was a 200 mile round trip) without a hybrid solution of some description. I was thinking of a series hybrid, whereby the engine/generator unit can be removed trackside to save weight. I won't need it for 1.8 miles of competition; 3 runs are completed back-to-back with a gap of around 5 minutes between each run and 90 minutes between each batch of 3 runs.

Sprinting is like AutoSolo but faster. Speeds (after the standing start) are likely to remain over 30mph and the top cars go over 150mph, although I have no aspirations to go that fast! Grip is clearly less of a problem but power and gearing are likely to be. I may not be able to build a car that is suitable for both without adding a multi-ratio gearbox, which I'd rather not do. I'm not cutting options yet, though.

Cheers,
Andy

Posted: Tue May 19, 2009 1:18 pm
by Jeremy
andylaurence wrote:Yes, you've shown that if you are running the batteries at 3C for mean motor power, then peak will be 9C. However, what if mean power is 20C? I'm not planning to commute to work in this car, I'm planning to get from A to B as quick as possible.


If your mean current draw is 20C, then your endurance is only going to be 3 minutes, so this is hardly a practical vehicle for any purpose, is it? Even a ten mile range will require a much greater endurance than this. Ignoring acceleration time and assuming that your ten mile journey is all on motorways (and that you stick to the speed limit!) then the absolute maximum mean battery discharge rate would be about 7C, and your endurance would be just 8.6 minutes before the battery was flat, assuming no losses and a unity Peukert factor. This is clearly an impractically high battery discharge rate.

If you're opting to build a burst acceleration race car, say for hill climbing, then lead acid batteries are the way to go. These will deliver massive amounts of current for short periods.i

I think you need to think things through a bit more carefully, to see how these factors all relate to each other.



andylaurence wrote:OK, that makes sense. Afterall, you size cables fo the current, not the voltage and power is current x voltage. I've just taken a look at the official rating of the components and made a blind assumption that they are true. I'd not considered that those ratings assume some cooling effort. Clearly, they will need cooling (they're just wasting electricity as heat), yet I had assumed that sitting in the wheel well would be good enough. It's certainly a breezy area!


Motors certainly do need a lot of cooling, curiously more so when they are delivering low power/high torque. All PM motors have a torque constant, which gives the torque per amp. As heat generated in the motor is proportional to the square of the current flowing through it, they need more cooling when delivering maximum torque (which is at low speeds) than they do when running at maximum rpm. This is another fundamental difference between an ICE and an electric motor, where max heat generation is generally at max rpm.



andylaurence wrote:Whoops! There's a typo to make. I meant to say that power is equal to torque x speed (RPM). It's a rate of work done. My point that I was trying (rather unsuccessfully) to make is that taking speed as fixed, it doesn't really matter whether you're thinking in Nm, lb-ft or bhp.


True, but if you're after "performance" in terms of responsiveness, acceleration etc, then the only thing that really matters is having enough excess (and usable) torque available at the wheels, at any particular speed. It's a fact that few 100hp cars ever use anything like 100hp on the road, we've just got used to quoting power figures in a rather blase way, without understanding the principle of work being done.


andylaurence wrote:If the Prius struggles with grip so much, why does it take more than 10 seconds to complete a 0-62 dash? Agreed, it probably does struggle to maintain grip at low speeds, yet it will struggle to spin the wheels at higher speeds. The torque from the motor hasn't changed during this time, but the power required to spin the wheels has...

Cheers,
Andy


The Prius runs narrow eco tyres, with a relatively low friction coefficient, so cannot put the torque from the motors to good use. It's also a heavy car, with front wheel drive,which makes the problem a lot worse when accelerating. The Prius will briefly spin the wheels (before the traction control kicks in) at speeds up to 30mph or so and virtually always tries to spin the wheels when pulling away from junctions, particularl if the wheels are at an angle. I once had it momentarily lose grip on both front wheels when accelerating hard up a hill, at probably a bit over 30mph. Not bad for a direct drive car with no gears............

For best performance you need a matched system, where the torque available at the wheel is just less than the tyre needs to lose traction. The higher up the speed range you can maintain this, the more sprightly the performance will be. In all probability you will find that the maximum power needed will be substantially less than that needed for the same performance with an ICE.

Jeremy

Posted: Tue May 19, 2009 3:23 pm
by andylaurence
Jeremy wrote:
andylaurence wrote:Yes, you've shown that if you are running the batteries at 3C for mean motor power, then peak will be 9C. However, what if mean power is 20C? I'm not planning to commute to work in this car, I'm planning to get from A to B as quick as possible.


If your mean current draw is 20C, then your endurance is only going to be 3 minutes, so this is hardly a practical vehicle for any purpose, is it? Even a ten mile range will require a much greater endurance than this. Ignoring acceleration time and assuming that your ten mile journey is all on motorways (and that you stick to the speed limit!) then the absolute maximum mean battery discharge rate would be about 7C, and your endurance would be just 8.6 minutes before the battery was flat, assuming no losses and a unity Peukert factor. This is clearly an impractically high battery discharge rate.


I'm looking at a maximum of 5 minutes run time. The fewer cells I can use, the lighter the battery pack and therefore the better performance the car will have. 5 minutes is 12C mean current. So, perhaps 20C is higher than I'd realistically need. Assuming a 50% duty cycle, that means I need a peak of 24C. Can an average lithium cell take that sort of abuse? How about when I add in regenerative braking?

Jeremy wrote:If you're opting to build a burst acceleration race car, say for hill climbing, then lead acid batteries are the way to go. These will deliver massive amounts of current for short periods.


Really? From what I'd seen, the peak power density wasn't as high, leading to big, heavy packs to provide enough power.

Jeremy wrote:I think you need to think things through a bit more carefully, to see how these factors all relate to each other.

andylaurence wrote:OK, that makes sense. Afterall, you size cables fo the current, not the voltage and power is current x voltage. I've just taken a look at the official rating of the components and made a blind assumption that they are true. I'd not considered that those ratings assume some cooling effort. Clearly, they will need cooling (they're just wasting electricity as heat), yet I had assumed that sitting in the wheel well would be good enough. It's certainly a breezy area!


Motors certainly do need a lot of cooling, curiously more so when they are delivering low power/high torque. All PM motors have a torque constant, which gives the torque per amp. As heat generated in the motor is proportional to the square of the current flowing through it, they need more cooling when delivering maximum torque (which is at low speeds) than they do when running at maximum rpm. This is another fundamental difference between an ICE and an electric motor, where max heat generation is generally at max rpm.


Useful info, thanks.

Jeremy wrote:
andylaurence wrote:Whoops! There's a typo to make. I meant to say that power is equal to torque x speed (RPM). It's a rate of work done. My point that I was trying (rather unsuccessfully) to make is that taking speed as fixed, it doesn't really matter whether you're thinking in Nm, lb-ft or bhp.


True, but if you're after "performance" in terms of responsiveness, acceleration etc, then the only thing that really matters is having enough excess (and usable) torque available at the wheels, at any particular speed. It's a fact that few 100hp cars ever use anything like 100hp on the road, we've just got used to quoting power figures in a rather blase way, without understanding the principle of work being done.


I think we're at cross-purposes here. I have no intention of this car being a road car. I already have a daily driver for that. I'd agree with everything you said for a road car.

Jeremy wrote:
andylaurence wrote:If the Prius struggles with grip so much, why does it take more than 10 seconds to complete a 0-62 dash? Agreed, it probably does struggle to maintain grip at low speeds, yet it will struggle to spin the wheels at higher speeds. The torque from the motor hasn't changed during this time, but the power required to spin the wheels has...

Cheers,
Andy


The Prius runs narrow eco tyres, with a relatively low friction coefficient, so cannot put the torque from the motors to good use. It's also a heavy car, with front wheel drive,which makes the problem a lot worse when accelerating. The Prius will briefly spin the wheels (before the traction control kicks in) at speeds up to 30mph or so and virtually always tries to spin the wheels when pulling away from junctions, particularl if the wheels are at an angle. I once had it momentarily lose grip on both front wheels when accelerating hard up a hill, at probably a bit over 30mph. Not bad for a direct drive car with no gears............


My Smart Roadster has 80bhp. It spins the wheels quite nicely from junctions. It would also be a whole load quicker if it had 160bhp.

Jeremy wrote:For best performance you need a matched system, where the torque available at the wheel is just less than the tyre needs to lose traction. The higher up the speed range you can maintain this, the more sprightly the performance will be. In all probability you will find that the maximum power needed will be substantially less than that needed for the same performance with an ICE.

Jeremy


I think I see where you're coming from here. Because the torque of an electric motor is so great from low revs, the peak power is less important as the motor produces more power across the rev range. It's a very good point.

Cheers,
Andy

Posted: Tue May 19, 2009 5:15 pm
by Jeremy
Doubling the power of your Smart Roadster would increase the top speed by about 28mph, or about a 26% increase in top speed.

If that power increase was obtained by increasing the engine max rpm, without changing the rpm/torque characteristic, then the acceleration times would remain similar to those with the 80hp engine, as acceleration is only dependent on the usable torque available at the wheels.

In practice, this means that performance at lower than maximum speed wouldn't improve much, unless the gearing was also changed to increase the torque available.

The big advantages that electric power provides for your application is massive low speed torque, which will reduce your times by enhancing acceleration, and the ability to accurately control the torque available at the wheel, to avoid wheelspin and get maximum traction. The simple way to do this is to mimic the Prius controller (not it's traction control system). The Prius controller monitors the motor phase current in order to control the motor torque, which is an ideal way for you to get maximum traction without being accused of using a traction control system.

Lithium batteries have a high energy density, but lead acid batteries are often better in terms of power density. Even cheap lead acid batteries are capable of delivery extremely high power levels for short periods, just the characteristic that you are looking for. I haven't looked closely at the current state of the art with lithium, but last time we debated this here (in the context of a hill climb competition car) we concluded that lead acid still had the edge over lithium.

A bit of work to estimate the sort of maximum torque your chosen wheel/tyre combination will take before breaking loose, together with the sort of speed profiles that you're likely to be running, should allow the optimisation of the design to be as quick as is physically possible, apart from the vagaries of conditions and the drivers skill! In many ways, an electric drive system is ideal for this type of competition, because it allows so much control, relatively simply.

Jeremy

More batteries

Posted: Wed May 20, 2009 2:05 am
by Deker
Andy.

I have to agree with Jeremy, you want power density, then lead acid is the choice, traction batteries though heavy retain their integrity when heavy currents are demanded.
Lithium, me not know enough about them, but I don't think they stand abuse.

You have to weigh up the cost of batteries for your application, put enough Lithium in your car, and they will perform nicely, cost a fortune, require cubic volume, including the number of bank notes, volume-wise, that is.

As for traction, wheel spin and all that........ What's the right pedal for ?

Think 3 phase. Sure, lower torque at low RPMs, but they can maintain torque at higher RPMs.

PM motors have a voltage to peak RPM cutoff point, in which the torque curve starts dropping off when the motor has gone past the halfway RPM to voltage RPM peak, (hope you understand this)

If its of any help, I have constructed a spreadsheet to calculate motor RPMs to road speed via ones chosen transmission type, and tyre diameter. In English.

Look in Services in this forum.

Deker.