Strantor's thoughts on how hybrid vehicles should be.

Thread Starter

strantor

Joined Oct 3, 2010
6,782
Long time EV enthusiast here, if you didn't know.

So I’ve been thinking a lot lately about hybrids and EVs as I drive my Prius around town installing AC & DC drives and such. I finally found the time to sit down and put all my ideas down in writing but found it a little hard to organize them into a linear thought process or exercise any brevity, so if you have any patience left in you at this hour, please read and critique.

I was thinking that gasoline powered automobiles are so inefficient for 2 reasons: (1) The engines must be sized for worst case scenario. I have read that it only takes ~20HP to drive full sized sedan down the highway. This could be accomplished with a single cylinder lawn mower engine, yet the sedan needs 5 extra cylinders and 200 extra HP so that the driver can achieve that 70MPH in haste, without being ran down by semi trucks. So once at freeway speeds, we have 5 cylinders banging away for no reason. I have read that some cars were built to use less than their full complement of cylinders when not needed, but that was a big failure so it hasn’t been tried again. (2) I have read that internal combustion engines have a peak theoretical efficiency of ~32%. That sucks, but it gets even worse; that 32% is only approached at a specific speed and torque, known as the peak of the power band. I will refer to this peak as “peak efficiency” henceforth. At all other speeds and torque loads (accounting for almost all normal driving conditions) the efficiency is much lower than the peak efficiency.

So my idea is that the hybrid vehicle should be purely electric drive, with a capacitor bank instead of a battery bank. The capacitor bank should be sized such that at nominal voltage, it will provide only enough power for one “pedal to the floor” acceleration uphill with full cargo load to freeway speed. The internal combustion engine should be sized only large enough to slowly recharge the cap bank after said heavy acceleration, once at freeway speed. Using the “20HP for a full sized sedan” rectally retrieved number as an example, perhaps the engine could be sized around 40HP.

The engine would be ran at a fixed RPM and coupled to a DC generator. A generator with a fixed field supply and ran at a fixed RPM would act as a constant voltage supply. Charging the cap bank with this constant voltage supply would result in a high load at the beginning of the charge, and tapering off to very little load once the bank approaches nominal voltage - allowing the engine to run outside it's peak efficiency . My idea is to intelligently regulate the field current to have the effect of turning the generator into a constant current source instead of a constant voltage source. Let’s say we have a cap bank capable of 400V, and we establish 200V as the nominal cap bank voltage. We establish 150V as the lower limit (generator turn on) and 300V as the upper limit (generator turn off). We could regulate the field for a desired “actual + 20V (rectally retrieved number)” output. So when the voltage present across the bank is 160V, the generator attempts to output 180V, and at 161V it attempts to output 181V, and so on until the actual bank voltage is 300V, at which point it shuts off. We reserve the remaining 100V for any energy that may be recouped by regenerative braking. By turning our generator into a constant current supply in this manner, and by fine-tuning the RPM and “actual + X volts,” (AKA current, AKA torque) we provide what appears to the internal combustion engine as fixed load that is exactly at the very peak of its power band (peak efficiency) 99% of the time. Also, the electric machine could be used as a starter for the engine.

So now on the traction motor side of things, we have a wildly fluctuating supply that could be as low as 150V or as high as 400V. How do we cope with that? Well, 2 things come to mind. We can employ the same intelligent field current regulating techniques as the generator when below rated voltage, and PWM current limiting when above. Let’s say we use a DC motor rated for 200V. When the available voltage is 150V, we can prematurely (below base rated speed) employ field weakening to achieve desired speed with less than desired voltage. This will be at the expense of higher than desired current draw, but this shouldn’t be a problem for the cap bank to supply, and as long as we have adequate cooling and high quality brushes, it shouldn't be a problem at all. When the voltage is 400V, 50% duty cycle PWM should have us right on the money. A combination or varying degree of these 2 techniques could be tailored for any possible bank voltage.

All this talk of intelligent field control leads me to regenerative braking. In the same way that we control the mechanical torque load applied to the generator by setting the target voltage at a prescribed voltage over and above cap bank voltage, upon braking, we can turn the traction motor into a generator and let the brake pedal set the prescribed “X volts above bank voltage” (AKA current, AKA torque). I believe that regenerative braking employed in this manner could provide significant braking power and power recovery, able to slow the car almost as well as mechanical brakes. Mechanical brakes would be engaged only in emergency circumstances, or when approaching zero speed, or when cap bank voltage is > 400V. Ordinary electrics and hybrids have batteries, which have a low “do not exceed” charge rate and therefore lose out on a lot of potential energy recovery when braking; with the cap bank, I think it would swallow whatever the regenerating motor could spit out.

The forward/reverse functions could be controlled by a low-power H-bridge on the field supply, minimizing losses. The armature could be switched by a single transistor instead of the 6 required for a 3 phase motor, again minimizing electrical losses. The high power regenerative braking could reclaim a much higher quantity of energy than what my Prius can. I have seen DC motors advertised with high-90’s percent efficiency. The field control switching in both the motor and the generator can be considered almost negligible, so I tally up very little losses in the electrical side of the house, and with the engine being several cylinders smaller than what’s found in a similarly sized internal combustion engine car, and ran exclusively at it’s peak efficiency, I foresee this being a VERY efficient car. The bulk of the electrical inefficiency would probably be in the step-down DC/DC converter for powering 12VDC loads. But then I am also dreaming up ideas to all-but eliminate the 12V circuit and power the larger accessories (A/C, lights, power steering, etc.) directly from the fluctuating bank voltage.

One down side I can think of, if you came out to your car to leave for work in the morning and found the cap bank at 0V, you might have to wait a little while in the driveway for it to charge before you take off. BUT, this could mitigated by a float charger that plugs into 120V power. 120V>step up transformer>rectifier, current limiter = low power 300V top-off. In the rare circumstance where maybe you went camping for a few days and your bank bled down and there’s nowhere to plug in, I suppose a 12V starter and backup battery wouldn’t hurt.

So, why is it not done this way? Why does my Prius have a 1.8L engine (same size as comparable non-hybrid), and why is my 1.8L engine coupled to the wheels mechanically with some convoluted sun gear and running outside it’s peak efficiency? Why am I carrying around 800Lbs worth of batteries when I could be carrying 100Lbs of capacitors? Why is it a $10,000 upgrade to get a cord I plug the thing in at night? Why does my Prius engage the mechanical brakes every time I come to a stop instead of recouping the dollars I invested to get it up to freeway speed?

I know, I know, everyone’s a critic. It takes teams of brilliant engineers to design these things and it’s not easy, and then everyone and their mother out there thinks they can do it better. But that’s why I started the discussion, to find out why my way isn’t better. I’m just here to learn.
 

WBahn

Joined Mar 31, 2012
29,976
Some of the things you mention have been explored and used. In particular, running the engine at a single speed to exploit its peak efficiency. Back in the early 90's Ford sponsored the Hybrid Electric Vehicle Challenge and schools that participated and met certain criteria in terms of funding and/or other sponsorship to have a realistic chance of completing their design were given a Ford Escort to modify as they saw fit. The team at CSM (and many others, I suspect) used a battery bank and DC motors as the sole drive and motorcycle engine turning a generator to keep the batteries at target levels chosen so as to have power available when needed but also regenerative braking capacity as well.

The notion of using capacitors looks nice on paper qualitatively, but in order to see some of the limitations of what you are proposing, you have to come up with representative numbers.

Let's put everything in the concepts favor and assume that we can use 100% of the energy in the capacitors and do not have any friction or other losses at all. How big would the capacitor bank need to be in order to get a car of mass M up a hill of height H?

The total change in gravitational potential is

E = MgH

The energy in the capacitor is

E = CV^2/2

Equating the two gives

C = 2gMH/V^2

Lets use your V=400V and lets use a mass of 1000kg and a hill height of 100m.

C = 2(10N/kg)((1000kg)(100m)/(160000V^2)

C = 12.5F

And that's a pretty light car going up a pretty small hill. What about a 2000kg car (still not a very big one at all) road that is climbing toward a mountain pass and ascending just 1000m to do so? Now instead of 12.5F you have 250F.

And that's assuming zero losses and the ability to use ALL the energy in the capacitors.

If you wanted to keep the charge between 100V and 300V, now that size goes up by a factor of (400V)^2/((300V)^2 - (100V)^2) which is 2. So now you are at 500F.

Turning to braking. A car moving at a speed S has kinetic energy of

E = MS^2/2

The energy in the capacitor is, again,

E = CV^2/2

Assuming we start off with an uncharged capacitor and end up at 400V, then we have

C = MS^2/V^2

Let's use S = 32m/s (~75mph) and M=1000kg. This means we need

C = (1000kg)(32m/s)^2/(400V)^2 = 6.4F

For a 2000kg car doing 106mph and going from 100V to 400V on the capacitor would take that up to right at 50F.

So the long hill climb is going to dominate, but you also have to look at the power handling requirements. A normal car can stop with about 1g of deceleration, so as it hits the brakes at 75mph (25m/s) the power output of the brakes is

P = dE/dt = (dE/dS)*(dS/dt)

dE/dS = MS
dS/dt = 1g

P = MSg = (1000kg)(25m/s)(10m/s^2) = 250,000W

For the heavier car doing twice the speed, it would be 1MW.

When you start seeing these kinds of numbers for even fairly small cars, you start really appreciating why riding your brakes going down a long hill is a recipe for disaster, literally.

Those are the kind of power levels that your electronics have to be able to handle in order to do regenerative braking. Even if you only use regenerative braking for decelerations 1/5 of the max, you are still talking huge power levels.

Another approach that could be considered is a hydraulic one using hydrostatic drives and accumulators.
 

bountyhunter

Joined Sep 7, 2009
2,512
I was thinking that gasoline powered automobiles are so inefficient for 2 reasons: (1) The engines must be sized for worst case scenario. I have read that it only takes ~20HP to drive full sized sedan down the highway. This could be accomplished with a single cylinder lawn mower engine, yet the sedan needs 5 extra cylinders and 200 extra HP so that the driver can achieve that 70MPH in haste, without being ran down by semi trucks. So once at freeway speeds, we have 5 cylinders banging away for no reason.
Many modern engines have systems that cut off fuel to cylinders not in use. They also have variable valve timing and variable ignition timing to "detune" the engine. The modern trend is very small 4 cylinder engines which are turbocharged to make very light weight and still have high peak power. This is a disaster as their actual gas mileage is far below their EPA ratings since the driver has to "rev up" to get any power at all. That means the engine is running in a very inefficient mode a lot of the time. But there are still lots of cars with good V6 and V8 engines that give good power and economy.

In reality, modern cars are FAR more fuel efficient. In my younger days (60's), a typical moderate performance car with a V8 got 12 MPG in the city and maybe 15 on the highway. Now you have cars that are twice as fast and get close to twice the gas mileage.
 

bountyhunter

Joined Sep 7, 2009
2,512
It takes teams of brilliant engineers to design these things and it’s not easy, and then everyone and their mother out there thinks they can do it better. But that’s why I started the discussion, to find out why my way isn’t better. I’m just here to learn.
The truth about hybrids is that the "hybrid" effect really has only a small effect on gas mileage. If you take a small subcompact of similar size and performance (all gas) the mileage is withing maybe 20% of a hybrid. The Prius is in the 45 mpg ballpark and the recent Motor Trend had a CIVIC all gas that returned about 35 mpg (combined). Highway was over 40.

Even at $4/gallon, the incremental money savings going from 35 mpg to 45 mpg isn't really all that much when calculated as cost per mile (and compared to all the other cost of ownership).
 

bountyhunter

Joined Sep 7, 2009
2,512
Why am I carrying around 800Lbs worth of batteries when I could be carrying 100Lbs of capacitors?
Because the capacitors would not store enough energy. The gas savings only happen if you can keep the gasoline engine from running as much as possible.

As for the weight of the batteries: yes, that's a problem and based on the last 30 years I was in the business, it's not a problem that has a solution.
 

THE_RB

Joined Feb 11, 2008
5,438
Long time EV enthusiast here, if you didn't know.
...
Me too. :)

...
So, why is it not done this way? Why does my Prius have a 1.8L engine (same size as comparable non-hybrid), and why is my 1.8L engine coupled to the wheels mechanically with some convoluted sun gear and running outside it’s peak efficiency? ...
What you are describing is basically called a "series hybrid" and has a number of benefits if done properly. That's my preferred way of doing it too.

A series hybrid basically is a pure electric vehicle, with a small gas engine running at good efficiency who's job is charging the batteries.

Car makers are slow to make series hybrids, although I think the new Tesla comes a lot closer? The big drawbnack to them is that the vehicle needs to stand on it's own as a purely electrically powered vehicle, so the electronics and electric motor all need to be able to power the vehicle under all conditions (peak loads etc) whcih means it is much more expensive in terms of the battery/electrics/motor which all now have to be large enough to power an entire car. Then you just stick a small gas motor and alternator on to keep the batteries charged.

The existing hybrids are a bit of a crock, they're just a gas powered car with a little bit of electrics added to "assist". So all the peak power (and peak consumption!) comes from the gas engine, then they stick an electric gimmick on top (which only does stuff when the motoring is really easy, ie at times when it doesn't save that much). :)

You should have a look at the solar racer cars, they have pioneered new technology in super efficient hub motors and streamlining and reducing running drags etc. Basically you need one of those, and add a tiny specialty-designed deisel engine which runs at one optimised speed and gives high efficiency conversion of fuel to electric charge for the batteries.
 
Last edited:

MrChips

Joined Oct 2, 2009
30,706
I have been thinking about regenerative braking for a long time.

With super caps now available today, a 1000F cap is very feasible.

What we need now is a variable 0-1000F cap. It would function like a syringe, to suck the energy out of the moving vehicle during deceleration and then pump it back into the electric motor for acceleration.
 

Markd77

Joined Sep 7, 2009
2,806
Some good ideas in there, it gets pretty complicated though. For city driving in heavy traffic there's an advantage to having a big battery, because you can turn the engine off. An engine idling while stationary uses quite a lot of petrol and has 0% efficiency. Of course carrying a big heavy battery also wastes fuel.
http://www.autospeed.com/cms/article.html?&A=112611
There are a lot of misconceptions about fuel efficient driving, for example I see a lot of people slowly accelerating away from traffic lights. They are using the engine in one of its most inefficient ways. If you have open road in front of you and you want to get up to 30mph, it takes the same amount of energy (ignoring all inefficiencies), so once you have let the clutch out, put a decent load on the engine.
Cars with small engines and turbochargers are more efficient at higher revs. A turbocharger increases efficiency so at low revs before it has kicked in, it is just as bad as a normal car.
 

panic mode

Joined Oct 10, 2011
2,715
one huge factor in efficiency is weight of the vehicle compared to weight of the driver... 2000kg car to move 75kg person. you burn lots of fuel to accelerate this huge mass, then you burn brake pads to slow it down.
 

crutschow

Joined Mar 14, 2008
34,280
Here's an interesting approach for a 170 mpg hybrid car using all hydraulics with a unique, free-piston Diesel engine and total regenerative braking. Unfortunately they apparently can't get any money to fund a prototype.
 

Thread Starter

strantor

Joined Oct 3, 2010
6,782
Turning to braking. A car moving at a speed S has kinetic energy of

E = MS^2/2

The energy in the capacitor is, again,

E = CV^2/2

Assuming we start off with an uncharged capacitor and end up at 400V, then we have

C = MS^2/V^2

Let's use S = 32m/s (~75mph) and M=1000kg. This means we need

C = (1000kg)(32m/s)^2/(400V)^2 = 6.4F

For a 2000kg car doing 106mph and going from 100V to 400V on the capacitor would take that up to right at 50F.

So the long hill climb is going to dominate, but you also have to look at the power handling requirements. A normal car can stop with about 1g of deceleration, so as it hits the brakes at 75mph (25m/s) the power output of the brakes is

P = dE/dt = (dE/dS)*(dS/dt)

dE/dS = MS
dS/dt = 1g

P = MSg = (1000kg)(25m/s)(10m/s^2) = 250,000W

For the heavier car doing twice the speed, it would be 1MW.

When you start seeing these kinds of numbers for even fairly small cars, you start really appreciating why riding your brakes going down a long hill is a recipe for disaster, literally.

Those are the kind of power levels that your electronics have to be able to handle in order to do regenerative braking. Even if you only use regenerative braking for decelerations 1/5 of the max, you are still talking huge power levels.

Another approach that could be considered is a hydraulic one using hydrostatic drives and accumulators.
You said a whole lot of intelligent things there, as usual (i was hoping to hear from you), and as usual, it will take me some time mulling over your words before I approach you level of understanding. But WOW! 1MW? I didnt realize what a modern marvel disk brakes are to be able dissipate 1MW into 4 dinner plate without turning themselves, the plates, and surrounding plastics into plasma.
 

crutschow

Joined Mar 14, 2008
34,280
You said a whole lot of intelligent things there, as usual (i was hoping to hear from you), and as usual, it will take me some time mulling over your words before I approach you level of understanding. But WOW! 1MW? I didnt realize what a modern marvel disk brakes are to be able dissipate 1MW into 4 dinner plate without turning themselves, the plates, and surrounding plastics into plasma.
There's a units problem here. Energy is given in Joules, not Watts (which is Joules/sec) and 1MJ is not that difficult to dissipate. For example the specific heat of iron is .473J / gm °k. Thus to dissipate 1MJ of energy with a 200°k(C) temperature rise requires 1E6 / .473 / 200 = 10.5kG or 23.3 pounds of iron. That requires four disks of less than 6 lbs each which is likely less than a typical automotive disc.
 
Lets use your V=400V and lets use a mass of 1000kg and a hill height of 100m.

C = 2(10N/kg)((1000kg)(100m)/(160000V^2)

C = 12.5F

And that's a pretty light car going up a pretty small hill. What about a 2000kg car (still not a very big one at all) road that is climbing toward a mountain pass and ascending just 1000m to do so? Now instead of 12.5F you have 250F.

And that's assuming zero losses and the ability to use ALL the energy in the capacitors.

If you wanted to keep the charge between 100V and 300V, now that size goes up by a factor of (400V)^2/((300V)^2 - (100V)^2) which is 2. So now you are at 500F.
To carry it a little further, ultracapacitors generally have a working voltage rating of 2.7 volts:

http://www.maxwell.com/products/ultracapacitors/products/hc-series

To build a 270 volt bank we would have to put 100 individual capacitors in series. If we need 500F then the individual caps would have to be 50,000F each. We would have to include circuitry to keep the voltage across the individual caps balanced.

This capacitor bank would only store enough energy to drive up a big hill one time. This is why there are no electric cars without batteries and only a capacitor bank.

Something else I noticed about ultracapacitors. Here's a 48 volt module:

http://www.maxwell.com/products/ultracapacitors/products/hc-series

Here's the spec sheet:

http://www.maxwell.com/products/ultracapacitors/docs/datasheet_48v_series_04302013.pdf

On page 1 an absolute maximum voltage rating of 51 volts is given, referring to note 3 on page 4. Note 3 says the absolute maximum voltage is non-repeated, and not to be applied for more than one second! Does this mean one time, for all time? Or can it be applied again for one second after it cools down? One must be careful about overvoltage it seems.

For electric cars to be equivalent in capability to existing gasoline powered cars, and therefore appealing to the general public, they need a battery that is inexpensive, has several times the specific energy density of existing lithium batteries, has a long life with repeated deep discharges, can be recharged quickly and is not too heavy or large.
 
Last edited:

WBahn

Joined Mar 31, 2012
29,976
There's a units problem here. Energy is given in Joules, not Watts (which is Joules/sec) and 1MJ is not that difficult to dissipate. For example the specific heat of iron is .473J / gm °k. Thus to dissipate 1MJ of energy with a 200°k(C) temperature rise requires 1E6 / .473 / 200 = 10.5kG or 23.3 pounds of iron. That requires four disks of less than 6 lbs each which is likely less than a typical automotive disc.
But power is in W, which is what was being calculated.

P = dE/dt

The time rate of change of the energy.

The total kinetic energy that must be dissipated is

E = MS^2/2 = (1000kg)(25m/s)^2/2 = 312.5kJ for the light, slow car and 2.5MJ.

And keep in mind that the majority (about 2/3 IIRC) of the energy is dumped into the front brakes and I think it is even higher in a max performance stop because so much weight is lifted off the rear wheels.

The bigger problem from a total heat standpoint isn't the heat to make a single high speed stop, but to make multiple stops in a short period of time or, even worse, to dissipate the gravitational potential energy while descending a long grade. That will be a big problem for regenerative brakes because there is no place to put it. That is why trains use dynamic brakes and not regenerative brakes. Ski lifts can use regenerative braking because they can dump energy back into the grid all day long since the grid is effectivelly an infinite energy sump.
 

t06afre

Joined May 11, 2009
5,934
I think turning on and of cylinders on demand. Have been used as a method of for saving fuel some time. I remember I once had failing fuel pump in the the middle of nowhere. So i lost one cylinder row. And ended up with a 2.5 litre 6 cylinder engine in a quite heavy car. But the car was still kind of drivable for the trip home about 200 km. On a flat road it kind of worked. But in hills the engine lost all its power
 
...to dissipate the gravitational potential energy while descending a long grade. That will be a big problem for regenerative brakes because there is no place to put it. That is why trains use dynamic brakes and not regenerative brakes.
Usually, if the vehicle is all electric (not hybrid, with a gasoline engine as well), the energy used to get up the hill will have come from the battery. Therefore, the battery will have enough free capacity to absorb the energy of regenerative braking when descending.

This assumes that the vehicle didn't spend the night at a hilltop inn, and fully charge the battery before making the descent!

The problem could also arise if gasoline power was used to climb the hill before the battery had been significantly depleted. But, don't the existing EV's, such as the Prius only start the gasoline engine when the battery has been significantly depleted?

Perhaps the vehicle's owner lives on a mountain top and always begins travel with a fully charged battery.

At any rate, blended brakes will be de rigeur, as with locomotives:

http://en.wikipedia.org/wiki/Dynamic_braking_(locomotive)
 

Wendy

Joined Mar 24, 2008
23,415
Something I've wondered about is why Stirling engines aren't used to drive a generator, period. Stirling engines have been tried by cars, but they have a lousy throttle control. They can burn anything, period, because they are a heat exchanger type engine, and a constant speed scenario used to drive a generator is ideal for them. They are also extremely efficient, but it hasn't been that long relatively that material technologies have caught up with their requirements.
 

panic mode

Joined Oct 10, 2011
2,715
how efficient is Stirling engine?
if i recall it is more efficient than steam engines (~12%) but no external combustion engine can match internal combustion engine which has limit of some 37%.
 

t06afre

Joined May 11, 2009
5,934
Something I've wondered about is why Stirling engines aren't used to drive a generator, period. Stirling engines have been tried by cars, but they have a lousy throttle control. They can burn anything, period, because they are a heat exchanger type engine, and a constant speed scenario used to drive a generator is ideal for them. They are also extremely efficient, but it hasn't been that long relatively that material technologies have caught up with their requirements.
As I understand the sterling engine is very complex to build and hence also hard to maintain.. Compered to say a steam engine. That is why it has not found it self any wide spread use yet
 

crutschow

Joined Mar 14, 2008
34,280
As I understand the sterling engine is very complex to build and hence also hard to maintain.. Compered to say a steam engine. That is why it has not found it self any wide spread use yet
The basic Sterling engine is really quite simple consisting of an expansion cylinder and a compression cylinder. The main difficulty is that it requires very high temperature alloys in the cylinder to withstand the high operating temperature required for efficient operation. Here's someone who's developing practical Sterling engines including using one in a prototype hybrid car.
 
Top