Battery for race car

5999 Views 21 Replies 11 Participants Last post by Yabert, May 11, 2018

SpeedRacer93

Discussion Starter · Apr 22, 2018

I working with a small group to build an EV land speed streamline race car. We have done a couple landspeed cars and a couple EV conversions and own factory EVs but this our first race EV.

Like most landspeed cars this is a DIY type build so budget is in play.

We are looking for different battery options.

Some of the specs are 400+ volts, 800 total AMP draw for 90 seconds.

Looking at BMW i3, Nissan Leaf, Volt, ampahaulic sleeper cells, A123 20 Ah and other pouch type cells.

We have ruled out 18650 due to the number of cells needed and RC Lipo due to the AMP load.

RC Lipo look good on paper but when you start looking at AMP loads the wiring from the pre built packs will not hold up. How can they rate a pack for 300+ Amp continuous when the connectors and wires only support spikes of 50-100 amps.

We are limited to 1000 Kg total weight so weight comes into play but we not worried so much about cycle life.

Anyone have any good ideas or suggestions?

1 - 20 of 22 Posts

brian_

· #2 · Apr 22, 2018

Some thoughts:

The voltage is just above that typical of a production EV, which doesn't narrow the choice at all but opens up lots of options.

The current seems too high for a Leaf pack, but no problem for Tesla (Model S or X, as Model 3 wouldn't be readily available yet).

Yes, anything using 18650 or other small cylindrical cells will have an insane number of cells to manually assemble, but this doesn't matter if you use complete production EV modules... and it would be substantial convenience to use ready-built modules, of any cell format.

The Ampahaulic Sleeper cells are typically used by drag racers - the 90 seconds needed may be too long for these to be optimal, although you could use their "continuous" rating, so run 2 or 3 or even more in parallel... it's still a featherweight pack compared to production EV cells near their production ratings.

Kevin Sharpe

· #3 · Apr 22, 2018

SpeedRacer93 said:
Anyone have any good ideas or suggestions?

Contact John Metric (here)

Karter2

· #4 · Apr 22, 2018

Any production EV pack (or even reconfigured cells), will blow your weight limit .
Ampaholic or good 100C, RC lipo is about your only option unless you have access to non commercial SAFT or A123's F1 pack cells.
As Brian said , using multiple parallel RC packs (5x4Ah ?) and possibly reterminating the cells to buss bars, is your closest DIY chance using commercial products.
But even a minimal capacity pack (8-10 KWh) will still be 100kg of cells

Hollie Maea

· #5 · Apr 22, 2018

A few things to consider:

1. Although they are cheap, I don't think you will be able to get satisfactory performance from any salvage packs. They do not use power optimized cells, and are designed to be discharged in an hour or more, not a minute and a half. Tesla packs can output a ton of current, but they do that by having a massive pack that would blow your weight budget.

2. You have to consider both power and energy, especially when looking at ultra high C rate cells. John Metric's cells are unsurpassed for ten second drag races; his best ones can do peaks of 200C. But for a ninty second race, you are only discharging at a rate of no more than 40C or you run out of energy.

To be more explicit, you need 20Ah to sustain 800A for a minute and a half. John's cells are only five Ah, and you wouldn't be able to get quite that much out of them in race conditions, so you would need five in parallel.

That isn't to say they wouldn't work; they would work great, and voltage sag would be limited since you wouldn't be running them anywhere near their power limit. But you are paying for a feature you aren't using (ability to drain in twenty seconds) so it's not super cheap. If you went with with a 96S5P pack, at 25 cents a pop you are looking at $12,000 in cell cost alone.

See less See more

Duncan

· #6 · Apr 22, 2018

Hi
I would suggest Chevy Volt cells
I'm severely abusing mine and drawing 1200 amps motor current

My peak battery current is over 800 amps - my car is about 800Kg all in but it's road legal and actually much heavier than it could be

I'm using 14kWh from a volt pack - the batteries weigh less than 140kg - I'm running from 300v to 340v - you would need another couple of modules to get up to 400v

The Volt has a fairly small pack - 16kWh as it is a hybrid so it works that pack a bit harder - and the Chevy engineers appear to have compensated by making it quite robust

And they are cheap!

brian_

· #7 · Apr 23, 2018

So, I missed that 1000 kg is the mass target for the entire vehicle. Yeah, at 400V the stock modules of any production EV will be too large

You might look at modules from plug-in hybrids, since they are smaller and yet some run high voltages; the lower total energy storage doesn't matter, so the issue will be whether they can stand the very high discharge for the short period. Otherwise, to use production EV cells you need to rewire them from the stock module configuration to put fewer in parallel.

SpeedRacer93

Discussion Starter · #8 · Apr 23, 2018

The ampahaulic/John Metric cells are on the list and in the running. Looks like we would need a 110s4p pack of those cells to meet the needs. Cell weight only would be 127 lbs but as stated the $12k price but that is the almost the price of a small block long block less blower/turbos.

Looks like the Volt Gen 1 or Gen 2 battery packs may be able to fit the bill if they can be pushed to 16 C for the 90 seconds. Chevy rates them at 7.8 C. Reports online show they can be pushed to 20 C but length of load test is kind of unknown.

Understand that the 400 V and 800 AMP is the max side. With most landspeed cars getting the power down is the hardest part. We plan on doing AWD/4WD setup with dual motors. Nissan Leaf, BMW i3, Tesla small drive motors are on the list using aftermarket/DIY controllers. Based on the aero numbers, we need 260kw total to reach our goal. So 130k each motor is very do able.

I know some are pushing a single Leaf motor to these levels but using dual motors simplifies the drive train. We looking at toothed/timing/blower belt drive or spur gear box. (no diff needed) We can "gear" the motor the keep it in the high torque area of the motor. Think 1.6:1 final ratio instead of the normal 8:1 EV ratio.

We could do two battery pack (2 x 400V x 400 AMP packs) but weight becomes an issue. Also a high parallel count of a single pack will help with the amp draw and voltage sag.

See less See more

Kevin Sharpe

· #9 · Apr 23, 2018

SpeedRacer93 said:
Reports online show they can be pushed to 20 C but length of load test is kind of unknown.

Have you considered buying part of a pack and testing some modules? They're often available at very low cost

Karter2

· #10 · Apr 24, 2018 (Edited)

SpeedRacer93 said:
The ampahaulic/John Metric cells are on the list and in the running. Looks like we would need a 110s4p pack of those cells to meet the needs. Cell weight only would be 127 lbs but as stated the $12k price but that is the almost the price of a small block long block less blower/turbos.

Looks like the Volt Gen 1 or Gen 2 battery packs may be able to fit the bill if they can be pushed to 16 C for the 90 seconds. Chevy rates them at 7.8 C. Reports online show they can be pushed to 20 C but length of load test is kind of unknown. ........

It sounds like you certainly know what you are doing , but to be clear.
Ampaholic cells are expensive because you are buying the best , high discharge , (200C) chemistry.... Which you wont be needing...but they will certainly do the job.
The cheaper (??) Volt pack may work for you if it can stand the higher output ( needs to be tested ) and would certainly require cooling systems. However that all comes at a weight penalty (200+kg) and a size/ shape factor to work with.
There are other high "C" rate, RC lipo options that could also work , and cost less than either of the above and weigh in at around 50kg, (and could easily be split into 2 separate 400v packs)
It depends what your prime critera are, Performance, Cost, Weight, size, life cycles, reliability, availability, etc etc
Most EV builders would not consider using RC lipo due to its low life cycle, and high maintenance /monitoring requirements(for safety). But for competition/ race use, it is difficult to ignor its advantages.
If you have the budget , go with Ampaholic, if not, consider other RC lopo options.

See less See more

tiger82

· #11 · Apr 25, 2018

We are running a Kia Soul EV pack in our Tesla Cobra EV race car. The 27kwh pack can sustain 800 amps for the period you are looking at. With the factory clamshell removed and our tube frame support (but without cladding) the pack weighs in at about 400 pounds. Take a look at some of our earlier posts or Youtube build video for some battery pack shots.

John Metric

· #12 · Apr 26, 2018

2

Our cells are for real high power to weight ratio OR low heat build up. "Long course" land speed racing is a lot like hill climbing... long sustained high power.
We would be competitive if weight or volume was THE issue. As true long course racing is not really that weight sensitive, it will be a close call. At 1.5 mile and shorter it truly is a drag race, ask Frank "200mph club" John.

I would suggest 4P108S about 128lbs about 1.5 cubic feet.

Remember to account for motor plus controller efficiency. I hear some folks quote me the motor output power of say a YASA P400 at 160kW which you will note it only makes at about 1/2 peak rpm and the motor efficiency is about 84% at that point. counting controller and cabling, contactors, fuses, efficiencies can hit 80% even on a AC motor system. So the 160kW becomes 200kW at the battery. That's why I always quote "batteryhorsepower".
Also another common mistake is to use the motor peak kilowatts and assume it will make it at peak rpm.

Another common mistake is not comparing sag voltage
A 3.2V 5C cell pulling 20C might sag to 2.0V
Whereas one of my cells pulling 20C would be nearly 3.8V
I admit I don't test to 90 seconds a lot because we were melting Derek's load bank. But we could test them for performance if you need. Here are a few test curves out to near 100seconds. That's high current from second zero BTW.

Lastly, as far as safety, a 1000lb tesla battery inside an armored case impervious to road debris takes about two fire trucks of water to "put out" and burns all night long without it. A 100lb plastic cover pack on the other hand while a little more dramatic, can be cooled directly if built properly with 10 gallons of water (I know).

You guys going to have a transmission? I would be happy to sharpen the pencil.

We are thinking about trying to be first to 300mph on the Texas Mile LSR course. I have a plan.

See less See more

SpeedRacer93

Discussion Starter · #13 · Apr 28, 2018

John Metric said:
Our cells are for real high power to weight ratio OR low heat build up. "Long course" land speed racing is a lot like hill climbing... long sustained high power.

We looking at doing "long course" land speed racing.

John Metric said:
I would suggest 4P108S about 128lbs about 1.5 cubic feet.

This matches up exactly with my spreadsheet. I think I have 4P110S to make them the same voltage as other cells in the spreadsheet. This is a good sign my spreadsheet is looking good.

John Metric said:
You guys going to have a transmission? I would be happy to sharpen the pencil.

We are thinking about trying to be first to 300mph on the Texas Mile LSR course. I have a plan.

Based on 2 mile, 3 miles, 4 miles and 5 miles calculations we will be able to meet our goals running a direct drive with a 1.5 to 1.6 final "rear end" ratios.

Give the Taxas Mile a go!!!

ishiwgao

· #14 · Apr 28, 2018

John Metric said:
Our cells are for real high power to weight ratio OR low heat build up. "Long course" land speed racing is a lot like hill climbing... long sustained high power.
We would be competitive if weight or volume was THE issue.

Remember to account for motor plus controller efficiency. I hear some folks quote me the motor output power of say a YASA P400 at 160kW which you will note it only makes at about 1/2 peak rpm and the motor efficiency is about 84% at that point. counting controller and cabling, contactors, fuses, efficiencies can hit 80% even on a AC motor system. So the 160kW becomes 200kW at the battery. That's why I always quote "batteryhorsepower".
Also another common mistake is to use the motor peak kilowatts and assume it will make it at peak rpm.

Another common mistake is not comparing sag voltage
A 3.2V 5C cell pulling 20C might sag to 2.0V
Whereas one of my cells pulling 20C would be nearly 3.8V
I admit I don't test to 90 seconds a lot because we were melting Derek's load bank. But we could test them for performance if you need.

thanks john for the informative post. very useful to keep for future reference!

SpeedRacer93

Discussion Starter · #15 · May 5, 2018

Does anyone have details on the Hitachi pouch cells used in the 2016+ Chevy Malibu ECO?

Looks like the cells are 5.2 aH, 3.7 Volts, 240 Grams and 5000 W Output Density (W/kg). Looks like that would be about 1200 Watts per cell. Which would be 324.4 Amps so about 62C Max (10 seconds)?????? (I have been in the sun all day so my math may be off)

Hitachi Automotive Systems cells spec:

Item Specification
Size (mm) 120×80×12
Weight (kg) 0.24
Average Voltage (V) 3.7
Capacity (Ah) 5.2
Output Density (W/kg) 5,000
Energy Density (Wh/kg) 80

1.5kwh, 115 volt, 32 cells in total pack

https://insideevs.com/hitachi-delivers-5000-wkg-prismatic-lithium-ion-cells-chevrolet-malibu-hybrid/

http://www.hitachi.com/New/cnews/month/2015/05/150519a.pdf

I know the old round cells before 2016 does not give the old round cells good marks.

https://www.energy.gov/sites/prod/files/2014/02/f7/battery_malibu_3800_0.pdf

See less See more

jr dragster

· #16 · May 5, 2018 (Edited)

We used a123 pouch cells a123 32157 ? cells volt batteries
Then we bit the bullet and got johns cells they are hands down the best
for racing light and powerful.
Dont do like i did and spend money on cells that will not work for racing
All batteries will work just some not for long ,or as good.
If i added all the money and time on all the other batteries i bought it was
less than i paid john for good ones.
RACING QUESTION = HOW FAST DO YOU WANT TO GO ??
ANSWER = HOW MUCH MONEY YOU GOT !!!!!!
Just my 2 cents

SpeedRacer93

Discussion Starter · #17 · May 6, 2018

jr dragster said:
We used a123 pouch cells a123 32157 ? cells volt batteries
Then we bit the bullet and got johns cells they are hands down the best
for racing light and powerful.
Dont do like i did and spend money on cells that will not work for racing
All batteries will work just some not for long ,or as good.
If i added all the money and time on all the other batteries i bought it was
less than i paid john for good ones.
RACING QUESTION = HOW FAST DO YOU WANT TO GO ??
ANSWER = HOW MUCH MONEY YOU GOT !!!!!!
Just my 2 cents

John's cells are at the top of my list but other team members want to looking at other options so we put together a list of hybrid/EV cars so I have been pulling specs on their cells and Hitachi ones jumps out.

brian_

· #18 · May 6, 2018 (Edited)

SpeedRacer93 said:
Does anyone have details on the Hitachi pouch cells used in the 2016+ Chevy Malibu ECO?
...
Hitachi Automotive Systems cells spec:

Item Specification
Size (mm) 120×80×12
Weight (kg) 0.24
Average Voltage (V) 3.7
Capacity (Ah) 5.2
Output Density (W/kg) 5,000
Energy Density (Wh/kg) 80

1.5kwh, 115 volt, 32 cells in total pack

The count of 32 cells is from the 2013 Malibu with a BAS mild hybrid system and the cylindrical cells, not the 2016 Malibu with the Voltec system and these prismatic cells. The 2016 Malibu with 1.5 kWh capacity of 19 Wh cells must have about 78 cells; from a photo it appears to have eight groups of 10 cells (so 80 total). They appear to be all in series, which would imply 296 volts (nominal), which makes sense since it needs to work with the same Voltec hybrid system as the Chevrolet Volt.

Non-plug-in hybrids need high power density, not high energy density, so this type of vehicle does look like a reasonable potential cell source.

110 in series by 4 in parallel of these cells (for nominal 407 V and nominally 8.4 kWh) would be 106 kg and 51 litres of volume... plus wiring and packaging. 800 amps total or 200 amps per cell would be 38C. That doesn't sound bad, but it's five and half Malibu packs to find, tear down, and reassemble.

brian_

· #19 · May 6, 2018

John Metric said:
Our cells ...
I would suggest 4P108S about 128lbs about 1.5 cubic feet.

That's 58 kg and 43 litres.

brian_ said:
110 in series by 4 in parallel of these cells (for nominal 407 V and nominally 8.4 kWh) would be 106 kg and 51 litres of volume... plus wiring and packaging. 800 amps total or 200 amps per cell would be 38C. That doesn't sound bad, but it's five and half Malibu packs to find, tear down, and reassemble... and I have only considered nominal voltage, not what it sags to in use as John explained.

These two options are essentially the same configuration, mass, and volume.

SpeedRacer93

Discussion Starter · #20 · May 7, 2018

brian_ said:
110 in series by 4 in parallel of these cells (for nominal 407 V and nominally 8.4 kWh) would be 106 kg and 51 litres of volume... plus wiring and packaging. 800 amps total or 200 amps per cell would be 38C. That doesn't sound bad, but it's five and half Malibu packs to find, tear down, and reassemble.

Since the Malibu Hybrid is produced by GM there are number of them available. I found 2 at local salvage yards so going to give them a call tomorrow. Based on searches the 2016+ Malibu hybrid packs are going for $500 to $1000.

We are looking at a high load battery test setup right now.

1 - 20 of 22 Posts

This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.