Battery cooling plate

[CrAzY_DrIveR]

Hi everyone,

So i am now looking at battery cooling plates for my new conversion from old batteries to NMC.

After a lot of research i see many options:

1. No cooling/air cooling.

I have been told that driving normally and rarely using rapid charging i would not see a rise in temperature.
It is tempting but in case it does not work the battery case would have to be redone.

2. Glue 2/3 cooling plates. Weld?

In this forum i see two examples of plates glued with Loctite 5926 or some other kind of silicone.
2 plates would have a cnc channel in the middle and both plates would be joined, 3 plates could be laser cut and then glued together.
Leaks would have to be carefully tested.

innerplate.JPG (28.83 KiB) Viewed 10249 times

Welding the thin plates would bring a lot of distortion, holes could be filled in the middle to add strengh and avoid separation of the plates.

Friction Stir Welding (FSW) looks perfect but how hard is it to find who does it?



Laser welding does work as long as the penetration to the lower layer allows a good bonding.



3. Extrusions

peugeot 208

Extrusions are very interesting, unfortunately it's not easy to find the one you want but custom ones can be made.
I sent multiple emails to businesses around me asking if possible how much would it be to make the following design,
what do you think?

[arber333]

Hi everyone,

So i am now looking at battery cooling plates for my new conversion from old batteries to NMC.
....

I will be editing this as soon as i get back. Thanks

NMC should be pretty good technology for thermal stability. If you dont fast charge them they will be good up to 1C charge discharge.
I noticed charge is really a problem at higher than 1/2C, while discharge doesnt bother cell even up to 1C. I noticed that temp even drops a little after quick charge session. They should be good at simple air cooling. But i am not sure what to do for heating. You could use lizard heater foil or camper heating foil at 12V...

My idea was to have a sandwich with more surface and lower channels to have more sealing area. I would use black glass sealant glue you can buy in hardware store. Or alternatively engine gasket sealant as it will hold at 300degC!!!

Another option would be to make a plate in 2 pieces and cut coolant channel with additional sealant channel around it. There you introduce O-ring and compress the plate using small countersunk screws.
For fittings you could make them from teflon on a lathe with a wide fitting on one end to keep 2 screws and wide O-ring for sealing. Screws will keep the fitting to the base and oring will seal the darn thing NO WELDING that way. But that would require a bit thicker plate assembly...

[tom91]

https://www.diyelectriccar.com/threads/ ... st-1127063

Good topic where someone tried all sorts of approaches.

[jrbe]

I'd suggest copper or aluminum tube pressed into an aluminum plate with channels cut in it. Tom's link has this style.

Should consider corrosion though too if using copper. Clear anodizing the plate might work. Watch for vibration / bending where the tubes exit the plate.

Another option is a CNC'd plate with a bolt on cover. You could do all the machining in that one plate, tapping, threading for hose fittings, etc. then have the top plate laser cut or punched using flush head screws and sealed with silicone.

[muehlpower]

18 years ago I built an icy lettering for Audi. The cold was generated by a refrigerator compressor. I used an insulating rear panel, copper pipe and an aluminum front plate on which the letters were mounted to transfer the cold to the aluminum letters. The space around the copper pipes was filled with heat-conducting paste. It worked very well. I could imagine the same setup with battery modules instead of the letters and water in the copper tubes. It doesn't have to be done very precisely and is always tight.

[arber333]

https://www.diyelectriccar.com/threads/ ... st-1127063

Good topic where someone tried all sorts of approaches.

I actually made some cooling plates for chargers using a copper tube locked in aluminum. I had to mill the straight channels into a 12mm slab of alu. I made it 8mm deep and 10mm wide. I used 10mm tube for climate.
For DIYing It requires a lot of straight tubes, no curves... And when pushing copper tube in i had to tap aluminum on the sides to lock it in. Mostly it did the job as alu is such a good thermal conductor. However as car is vibration intensive enviroment i was afraid copper would rattle and jump out of the groove... Maybe 3 slab sandwich is not such a terrible idea after all...

[Daveturpin]

Following, this task is one of the next things I need to do on my project.

I have CATL prismatic cells, which are cooled conductively in the Mercedes EQ they came from with what, in my opinion, is a bit of overkill. Like it was said above they probably don't need MUCH cooling.

I built the battery boxes with an extra 1" of space under the cells with the modules bolted down:

Right now the plan is aluminum tubing bent into long runs and pressed into a conductive substrate. But until I get that done, I am pretty sure that just blowing air through the gap would be enough for testing?

[CrAzY_DrIveR]

direct contact with the floor of the car will cool a lot

[tom91]

Like it was said above they probably don't need MUCH cooling.

All depends on your application, main reason for cooling or heating is fast charging at low or high temps.

Also best get a BMS that is capable of limiting charge speeds based on temperature to manage this.

[CrAzY_DrIveR]

more than 1 month later, i have been drawing and asking for quotes in china.

The extrusion guys in poortugal or spain are all "censored" and don't need the work(must be a socialist thing...).

The first quote in china is 360 bucks + 200 shipping + 23% tax in customs would put me back more than 700 bucks.

I am lost and will give up on cooling real soon...

[tom91]

back more than 700 bucks

That is cheap for such a large custom one off piece from china. Most the money probably goes into setup costs.

You will not find a cheaper way to do such bespoke items unless you can do most the work yourself. Even the costs of all the aluminum required is very high.

[CrAzY_DrIveR]

i understand, got some quotes of 3500 and 1500 right now.
but it's a cnc plate that goes to an oven for brazing i don't know how much setup it needs.

[tom91]

but it's a cnc plate

Machine setup time, material setup and CAM. Plus Brazing analysis. All is time. Employees get paid. Its simple math.

Do more parts the fixed engineering cost gets split of multiple units.

Keep in mind if things are cheap they will be rushed and done incorrectly. Propper custom parts cost money.

[arber333]

Hm...
I am now working on new battery box.
My idea is to have a modular box with double bottom.
The concept is to sacrifice like 5mm of space for insulated layer on the bottom. I would use 5mm of self extinguishing material for floor boarding...
Then i would bend and weave 10mm climate copper tube (it is much softer than other tubing) on the floor of battery box so i could get coolant through it to heat/cool my alu box cells.
Well in between that tube i would put 8mm insulation strips to direct heat to alu plate above.
Now then i would put 6mm plate over the tubing to transfer heat or cold through it to the batteries.

I still need to think on how to fix the plate to the box bottom and battery modules to that. I guess 12mm alu blocks should be good for that?
I think for heating i would have to use higher temperature than just mild temp... directly from cabin heater...well see.

[Daveturpin]

I have completed my front battery heat exchanger. Again caveat is that my main concern is balancing temperatures. I am not DC fast charging, but regardless I've done thermo calculations in Autodesk and I should be able to dissipate 2000 Watts of heat with the following design, even with the not-ideal connection between the tubes and sheet.

Basic design is 1" foam with 1/2" aluminum tubing, 3" bends, 20ga aluminum sheet to spread the load and heat. The front HX uses 7 bends because it needs to be in "landscape" orientation. The bottom HXs will be a simpler 3-bend design because they are portrait orientation. Here is the tool I used to cut the grooves in the foam:

And cut out. It was a little rough at first until I got the hang of it. But this is just to be able to recess the tubing.

Hand bending of the tubing (annealed refrigeration tubing) with a spring tool to prevent kinking. Still, doing this by hand was a pain.

Once bent I used a plastic-to-metal epoxy to put the sheet on...

And placed the sandwich between pieces of steel and all the heavy things in my shop. The following day I took it out and the sheet is very tightly bonded:

And it is basically done just need to trim (or bend over the edges) and figure out how to terminate the pipes.

Theoretically you could get this down to 1/2" thick and cut the foam instead of grooving it. There is also a lot more that could be done with a mandrel tubing bender, brazing, etc. A double run design I ran in Autodesk with tighter bends appeared to be able to dissipate more than 6000 watts... Which I would think would be enough even for DC fast charging. (It would be 15% heat losses on a 200kW charge over 5 such HXs)