[rfengineers]

I'm a little unclear on (among other things) battery capacity requirements. This uncertainty has arisen due to comments I keep seeing from people about LiFePO4 cells with a C3 rating being marginal for a street EV.

For example: Someone recently said that a Hi-Power 100A cell with a C3 rating was inadequate for a 500A Curtis controller and that you needed a C5 rating. My understanding is that the 500A Curtis controller is only rated for about 250A or so continuous (1-hour = infinity for this purpose) service. If the battery rating is C3 continuous and (hopefully) something greater for intermittent duty, why would you need a C5 rating?

The above comment was about a 144-Volt battery. 144-Volts at 500-Amps is 72kW (96hp). That seems like a heck of a lot of power to draw for any length of time. My current conversion project will use [email protected] for 18.6kWhr (about 11kWhr usable) of energy. A 72kW load would give me about 9-minutes of use (assuming the motor and controller could handle it)! This just can't be right (I hope)!

There seems to be a great deal of (well-intentioned but) wrong information out there. This really bothers me. OK, never mind the rant, lets get to my question.

Does anyone have any real-world measured numbers for typical compact car acceleration/cruise under various conditions? That would be battery pack Volts & Amps (everyone knows that motor Amps only indicate torque, right?).

And while we're at it, a battery C-rating is usually continuous, right?

Joe

 

[TX_Dj]

Joe,

I shot an email over to the NTEAA president about your questions. He has real-world experience with LiFePO4 cells in his Fiero. Hopefully he can lend a hand with some of these questions.

There are usually more than one C rating for batteries, at least that I've seen.

I.E. the batteries I have for my scooter are rated as follows:
Max Short-duration Discharge Current: 10C (10 Seconds)
Max Discharge Current: 3C (7 Minutes)

Anything below 3C is continuous in this case.

One thing to keep in mind, though, is that your current limit on the controller is in relation to motor amps, not battery amps. I.E. once the RPM comes up, so do the voltages, and current comes down. I don't think you'll ever see 500A at the batteries, but I could be wrong.

 

[rfengineers]

...your current limit on the controller is in relation to motor amps, not battery amps. I.E. once the RPM comes up, so do the voltages, and current comes down. I don't think you'll ever see 500A at the batteries, but I could be wrong.

DJ,

Thanks for helping out! I look forward to seeing that LiFePO4 data.

Yes, I have the motor-current/battery-current concept well under control. When trying to guestimate range and select batteries it's battery-current that is important. Motor current (usually limited by the controller) tells you how much torque the present load is demanding. Without knowing the PWM duty cycle (the ratio of pulse on-to-off time) motor-current tells you little about the demand on the battery.

I'm a crusty, old-school, electrical engineer with US Navy electronics technician experience, so I have the electrical part comfortably in hand. I have the assistance of my, "master mechanic and former professional race-car builder," brother-in-law in the automotive/metal fabricating area. We are seriously entertaining the notion of making a business of converting cars for other people. Now THAT is a frightening thought!

Joe

 

[jbrecher]

I have ChinaHypower 100 AH cells in my Fiero. I don't have a lot of data yet as I'm waiting for battery monitoring system. I do have regulators on each cell from Hot Juice Electric.
The C3 rating would not make much difference in regular driving. Those who think they need " continuous" longterm have never driven an EV. I have a 1221C which is 400 Amps. I may see 200 Amps for maybe 3 or 5 seconds after startout from a light. Then it drops back to something about matching the MPH. I have a Corbin Sparrow that has a 1000 Amp controller that sees 300 Amps for 2 to 4 seconds from a light.
I think the 100 AH cells are a bit small for my Fiero but wasn't going to pay for double the price 200 AH cells. The car has lost 1000 LBS with the change from lead to lithium, so I'm more than happy with the performance of the car. When I get more data on the daily use after I get straps working better and the monitoring installed I'll post about it.
So far I've driven 40 miles several times on a charge but don't know where the bottom of the charge is so am not pushing it. I start out with 117 volts (32 cells at 3.65 V) and see it drop to 107V or so after one throttle application. Then it stays there for 30 plus miles. I drove it yeterday in Dallas 45 MPH traffic and when I got home I had dropped to 104.7 V. 87 volts should be the bottom. I'm just afraid the bottom will come as quickly as the first throttle application.

 

[samborambo]

The first step to determining continuous power drain for your batteries is to calculate the forces associated with sustaining speed at a given speed. You can break down the forces into these categories:

1. Wind resistance - majority of losses at high speed (highway). Wind drag force (in Newtons) is proportional to the square of the speed. Other factors are the vehicle coefficient of drag, the vehicle cross sectional area (frontal area in metres squared) and the air density (kg/m3). There's a good Wikipedia article on this.

2. Rolling resistance - most prevalent at lower speed. Rolling resistance force is proportional to the vehicle weight (in Newtons) and the tyre rolling resistance coefficient.

Multiply the two above by the speed in metres per second to determine the power required. This means the wind drag is proportional to the CUBE of the speed.

3. Battery internal resistance. This is commonly shown as the Peukert effect associated with lead acid batteries. Power lost from internal resistance is proportional to the current squared.

4. Controller losses. This is a bit tricky to calculate unless you designed the controller yourself. It's made up of switching losses and on-state losses. Switching losses aren't worth guessing. On state losses for MOSFETs are an ohmic loss similar to the battery internal resistance.

Lucky I saved you the trouble of working this stuff out - well, the mechanical stuff anyway. I put together a spreadsheet and posted it yesterday:

http://www.diyelectriccar.com/forums/showthread.php/open-source-vehicle-sizing-calculator-20803.html

The worksheet you want is Component Selection. Set your parameters there and then check the power and energy consumption curves. There's also worksheets showing drag and tyre data.

Its not documented very well but I hope it helps. Ask for any explanations, etc.

 

[jbrecher]

I've been driving an EV on the road for 8 years. Yes, I know about aerodynamics as I build and fly "experimental aircraft. Yes, I know about LRR tires. Peukert theories are great. Get out and get an EV and start driving and see if any of your theories work.
My point is so many folks set at their desk and talk theory. Go out and try something!

 

[samborambo]

I've been driving an EV on the road for 8 years. Yes, I know about aerodynamics as I build and fly "experimental aircraft. Yes, I know about LRR tires. Peukert theories are great. Get out and get an EV and start driving and see if any of your theories work.
My point is so many folks set at their desk and talk theory. Go out and try something!

Real world data may not be as useful as you think. Unless you completely replicate a build with measured data you'll end up with different results and, more importantly, probably not know why you have a variation. Since performance results are dependent on most of the elements in an EV, changing one may have a great impact on overall performance.

I think it is important to get the calculation process down packed and prove it with measured data. That way any modifications in copying a build can have anticipated performance results.

There's a few books on EV conversions but I've yet to see any open documentation, like a manual, on EV conversions. I see there's some formula pages here but I feel we need something a bit more comprehensive.

 

[jbrecher]

I have members of my chapter of the EAA in North Texas who work the numbers and compute to data and tell all what we should do. When it boils down they look at the data we "doers" give them and make it so.
Most of the engineering and data is done years ago. What I'm doing is building your data on lithium cells so you can compute some more numbers for those thinking of lithium. If I don't drive and make real data, you have no figures to put down.
Push away from the computer and GO BUILD SOMETHING! Then tell me your data on things like the "Peukert" effect, LRR tires, bellypans for cars to increase aerodynamics, this size and material for bussbars and cables, the regulator needed to give long life to the lithium cell, the charger charging curve needed, and on and on.
Do you see that until I give the "calculators" data, they can do nothing "real".

 

[rfengineers]

Hey guys!!!

This Phenomenological vs. Theoretical debate has been going on in academia for generations. Can't we either "agree to disagree" or just use one to verify/support the other??

Joe

 

[todayican]

Speaking as one of the guys who just goes out and builds something, THEN try's to figure out why it didn't work ;-)

I can certainly appreciate both sides of the argument. The guys who have the calculators ARE doing something in my opinion. its just a different something.

When it comes to Chinese lithium batteries, I am interested to know how they react to taking the full rated (2c?) charge, or even being pushed a bit?

Thanks

 

[samborambo]

Speaking as one of the guys who just goes out and builds something, THEN try's to figure out why it didn't work ;-)

I can certainly appreciate both sides of the argument. The guys who have the calculators ARE doing something in my opinion. its just a different something.

When it comes to Chinese lithium batteries, I am interested to know how they react to taking the full rated (2c?) charge, or even being pushed a bit?

Thanks

I've read on the ZEVA forums that the Thundersky LFP cells droop from 3.2V/cell to around 3.0V/cell at 3C discharge. That's 741 micro ohms for a 90Ah cell! Still anecdotal evidence though. TS haven't published any internal resistance data which is disappointing.

 

[madderscience]

OK, real world data for a 1985 Toyota MR2 EV that weighs 3400lbs curb weight and has a traction battery of 21 T-105s, a curtis 1231C and a 9" ADC motor:

55mph steady state, level cruise: 90 battery amps (3rd gear)
40mph steady state cruise: 50 amps. (2nd or 3rd)
25mph steady state cruise: 25 amps (2nd gear)

10% hill climb at 30mph: 300 amps.
Hard acceleration: brief spike to 400+ amps and then 300 or so.

So a single string of 100AH, 2C battery would probably work for the car especially since the lighter weight should help with acceleration and hill climbing peak amps, but I agree with the NTEAA' presidents thoughts that the batteries might be a little bit small. (Even though I am just calculating)

If I could afford it, I'd use 150AH, 3C thunder sky batteries in my car if going to LiFePO4. 3C 450 amps is close enough to 500 that the batteries should not be overly stressed even under worst case conditions. At 2C and 200A for the highpower batteries a long but moderate hill or aggressive driving could take them over their rated limit pretty easily. Maybe not a problem if you live in a flat area and have a light foot.

As for build it vs. calculate it, Build whatever you want of course, but stuff has specifications on it for a reason, and it makes sense to take those specs into consideration when building something, or durability/performance/etc might be compromised. I would not want to risk the cost without planning it out. I tried to be pretty rigorous about this (with the notable exception of GVW, though I compensated for it) and I am happy I did.

 

[madderscience]

I've read on the ZEVA forums that the Thundersky LFP cells droop from 3.2V/cell to around 3.0V/cell at 3C discharge. That's 741 micro ohms for a 90Ah cell! Still anecdotal evidence though. TS haven't published any internal resistance data which is disappointing.

Ah! So (warning, wielding calculator) 0.2 ohms (internal voltage drop) * 270 amps == 54 watts of power being dissipated inside the battery as heat.

That's about like having a 60W incandescent light bulb inside a plastic case the size of a typical paperback novel. Plastic is a poor conductor of heat, so I imagine that 54 watts could heat up the battery pretty well and rather quickly. Put another way, if you had a ~150V traction battery (50 Cells) that is 2700 watts of heat being put out by your battery pack. Better have some cooling if that is going to be sustained for any length of time.

 

[sailfish11]

I'm sure he wouldn't mind me posting a link to his page (after all, he put it on the web for all of us to see) -- Chris Jones has a fantastic website for his '66 Mustang conversion, using Valence batteries. TONS of real-world data:
http://www.geocities.com/[email protected]/EV/1.htm

 

[rfengineers]

You guys are the best! This is exactly the kind of data I have been looking for.

Joe