Prelude conversion project - some questions

[rhills]

Any idea how much water is bad for a cell? I'm thinking of where mine are right now, where they are probably getting at thin film of condensation on these cold mornings.

Hmmm, might need to bring Brumby in by the fire on those cold winter nights

[jonescg]

Looks like the least-worst way to resurect this car and my dignity is to build a non-cooled battery using the 63 Ah LG pouch cells. I'l put them inside sealed polycarbonate enclosures and maybe pack them out with grease to ensure a long service life.

I can fit 8 stacks of 12 cells each inside the main battery pack, meaning NO stuffing around in the boot of the car. I'll have to mount the isolator somewhere convenient, perhaps under the bonnet but I will make the fuse accessible from underneath the car by cutting a small hatch in the base like an iMiEV. Slight hit on range, though. It will be 22 kWh instead of 24 kWh.

Best of all the price for these cells is very reasonable. About $5000 landed, and maybe another $2000 in enclosures and buslinks.

All coming out of my own pocket mind you

[smokinjoe]

Sorry to hear about this Chris

[T1 Terry]

Would using one of those "E" spec coolants get around the problem?https://multimedia.3m.com/mws/media/199 ... -fluid.pdf No water to condense and cause problems and the coolant turns into vapour if it gets above its boiling point condensing again once it cools.

T1 Terry

[jonescg]

It would have been much safer, yes. But it would have also continued to leak - I still have no definitive answer on the source of the leak, but I'm guessing it was the back of the cooling plate on module 3. At least it would still be working, albeit oily.

[jonescg]

Time to resurrect this thread of toil.

Francess' car has been in my shed while I wait for new cells to arrive. The LG cells I ordered back in June never arrived, and the vendor claims they were destroyed in customs. When asked why, they said they didn't like the brand name on them (?) but assured me they could find some more. They claimed to have found some more, but insisted they woul dneed to remove the logo so that they wouldn't be destroyed in customs again. My interpretation was - "We took your money but can't find the cells you wanted, so we found some who-flung-dung knock-offs with nothing written on them and we hope you will take them and shut up".

So I have asked for my money back. And went back to my usual cell supplier who has never let me down. I have decided to get 2300 Panasonic NCR19650BD cells and build a 24 kWh pack with cylindrical cells
https://evwest.com/support/panasonic-nc ... asheet.pdf

This is a mountain of work, but there are two upsides. The cells are 100% legit and not knock-offs. I've bought them before and they deliver; they will do a peak of 10 A, but I would generally consider them to be a 1C cell. Secondly, they mean I can fit the original 24 kWh capacity in the under-floor battery space with no need to run cables up into the boot. The overall weight will be slightly less than that of the original battery, and it's all in front of the rear axles.

I'll build the battery as four modules - each in a 24s23p configuration. They will be held together with a pair of polycarbonate capture plates use busplates much like a Tela Model S module, but I doubt I can get them wirebonded easily.

cylindrical template colour.jpg (1.1 MiB) Viewed 1769 times

cylindrical pack 3d.jpg (371.98 KiB) Viewed 1769 times

At least spotwelds work. The busplates will be made from laser cut aluminium which is then nickel plated. This means a small nickel tab can be spotwelded to the cell end and to the busplate. If the spotwelds don't stick to the nickel plated ally, they can always be soldered.

robot drool.jpg (497.83 KiB) Viewed 1769 times

When all is said and done, the modules will be dammed up and fully potted with epoxy encapsulant. Sure, this adds weight, but I want a robust, waterproof battery.

cylindrical pack dimensions.jpg (774.77 KiB) Viewed 1769 times

The Zeva BMS modules will fit on the back side where the fly-wires will lead around. I'll sit one on top of another on either side of the 100 mm SHS beam in the battery pack. Fix them down with a pair of clamps, and we're good to go. A massive job, but hopefully worth it.

[brendon_m]

Okay, I'll bite. What's the black cell about?

[brendon_m]

Wait, I've got it. Its a missing cell.

[jonescg]

Give that man a cell!
Yes, you have no idea how many iterations of 24s23p I did to make it work, and in the end, this was the least crazy cell arrangement. Even if it left a vacancy right in the middle.
Which is good, as this gives me a decent location to start filling the whole thing with encapsulating resin.

[4Springs]

Which is good, as this gives me a decent location to start filling the whole thing with encapsulating resin.

Will you encapsulate the connecting plates as well? Or leave those un-potted and open to a cooling breeze?
Are you talking plain old resin, or some kind of thermally conductive goo?

[jonescg]

I'll encapsulate the lot. I don't want any corrosion opportunities anywhere.

[antiscab]

Put some moisture absorbing material in there for good measure?

[francisco.shi]

What are you going to use as the encapsulant?

[jonescg]

I have a few options - EPL-9 polyurethane potting compound, or West Systems epoxy. Either would work I think.

[francisco.shi]

How heavy is it?
Is it like foam?

[jonescg]

No, EPL-9 is more like regular resin when it sets, but with a little bit of give.
http://specialty-products.com/spi_products/epl-9/
Density is about 1.1 g/ml

[T1 Terry]

Have you seen the episode on the Electric Samba where he links the cells using wire he tested to fail at what ever current to create a safety fuse just in case a cell went short circuit?
Have you thought about filling the whole battery with a non conductive cooling fluid so it can be cooled easily and be certain no water can get in?

T1 Terry

[rhills]

I have a few options - EPL-9 polyurethane potting compound, or West Systems epoxy. Either would work I think.

Chris, have you seen the Australian manufactured BoteCote Epoxy? Not sure how it compares price wise with West Systems, but I use it exclusively now, having found it easier to work with than West Systems. Being more flexible might help reduce the risk of cracking causing leakage too.

Cheers,

[jonescg]

Yeah I've looked at a few different epoxy systems. All would work, but I have no idea about the long term performance in a battery application. Main goal is to seal the battery from moisture ingress and help protect it from vibration and physical trauma. So something with a little give is ideal, but not too much. It also needs to flow into every nook and cranny really easily.
The polyurethane stuff I bought earlier was originally destined to protect the pouch cells, and apparently a harder resin us used on cylindrical cells. Which makes sense.

[jonescg]

Minor update - I found a laser cutter who will make all my bus plates from 5005 aluminium for a very reasonable price ~$1150. My usual place originally quoted me this much, and then said they'd accidentally underquoted me, and it would be more like $5500 :O So I went down the street to someone who would cut them up at a fair price. I suspect they had a better paying, angrier customer who wanted their work done yesterday, and they needed to bump me off the queue. Worked out OK in the end.

The capture plates are well o their way and are a good fit, although I heard that one of them is a bit tight. So long as I make that the first one I add cells to, it won't matter a lot. Then its a case of assembling the 24s23p modules very carefully, making sure the cells go into exactly the right spot in the right orientation, and dam it up securely for the resin treatment later on. Bus plates are screwed down and I'll need to weld each cell through the 10 mm hole to the bus plates. An automatic wire bonder would be ideal here, but I have a feeling I'll be spot welding all of them

Then finally I'll get the BMS wiring out the other end, connect them up, glue some FR4 sheets top, bottom and around the sides, and fill 'er up with resin. Waterproof, vibration proof, idiot proof

[jonescg]

I have 100 of these guys all cut up nice and pretty. I'll first take them to the nickel platers who will do an electroless method which I believe has a short zinc treatment in there.
It will add to corrosion resistance, but it also makes bonding to it a lot easier.

20201113_121052.jpg (160.89 KiB) Viewed 335 times

Only 100 of them. Hopefully not too expensive to treat.

The $3500 worth of polycarbonate machining will be done next week too, and the cells will be delivered late next week. So I should be able to make a decent start.

[jonescg]

Jigsaw puzzle time...

cut bars1.jpg (192.14 KiB) Viewed 313 times

cut bars2.jpg (223.28 KiB) Viewed 313 times

I'll need to touch them up with a deburring wheel before getting them plated I think. There's a few daggy bits.

[jonescg]

cylindrical cells arrived.jpg (123.04 KiB) Viewed 202 times

And the work is about to commence...

[jonescg]

I have three of the four modules capture plates made up, so I started to populate one of them. I can leave the cells as they are without needing to pull the whole lot apart this way.

Partially assembled module.jpg (201.12 KiB) Viewed 140 times

Each module will store a bit over 6 kWh and sit at 100 V fully charged. And should weigh no more than 30 kg each.

I want to bend the terminals down for safety and have a brass nut soldered to the back for ease of terminating. But sealing the whole lot up with resin will make this step a bit difficult. Will need to think about this a bit more before getting too carried away.

The internal resistance of these cells is about 60 mOhms, or in a 96s23p arrangement , about 0.25 ohms total DCIR. So at 36 kW, it will be generating 2.5 kW of heat! Hopefully it's only generating that much heat on the way up Greenmount hill. Genuine 3.2 Ah capacity though. Theoretically there's 26 kWh all up, but I'll say 24 kWh to be safe.

Last thing to do is deburr and nickel plate the busbars.

[jonescg]

tab.jpg (126.56 KiB) Viewed 134 times

The attachment strategy...