[428RC]

Hi Remy,

on the Mach pack, the long bolt through the battery packs is a shoulder bolt, so bolt torque is moot in relation to the battery compression, the length of the shoulder vs the battery pack/foam spacer stack is the relevant dimension. I removed the foam and reassembled the stack battery to get the dimension, then extrapolated that to get the bolt length for an expanded pack. shown here

 

[remy_martian]

Thanks for the insight on how they do it.

 

[remy_martian]

This is what I was thinking of, which Simon initially posted: https://silex.co.uk/shop/sheeting/s...f-thermally-conductive-silicone-sponge-sheet/

Not sure what @D&VsEVJeep are planning to use, or if there's other thermal interface material we can use that you don't need to buy several tons a year of.

 

[D&VsEVJeep]

This is what I was thinking of, which Simon initially posted: https://silex.co.uk/shop/sheeting/s...f-thermally-conductive-silicone-sponge-sheet/

Not sure what @D&VsEVJeep are planning to use, or if there's other thermal interface material we can use that you don't need to buy several tons a year of.

We are going to just reuse the compression pads that are in the modules... They actually come out easily and are in good shape...

 

[remy_martian]

Yup. There are what you call thermal compression pads withon the modules, then there are the strips that interface the module to the chiller plate. The link I provided was for the latter....no need to change out the former....they have a tacky, reusable adhesive, whereas the module to chiller isn't easily removed/reused from the looks of it let alone available to people who buy loose cells or modules vs a pack.

 

[kennybobby]

Oh yeah that's the compression factor "activation" feature that i was thinking of, having brain fog as to the technical name for this property

 

[428RC]

in my disassembly of the Mach batteries. I noticed that the design seems to work to prevent the battery pouch surface or the heat sink that touches them from touching any other metal. this includes from heat sink to heat sink, from pouch/heatsink to the long axis shoulder bolt, the endplates to the heatsink, and from the heat sink to the chiller plate. the heatsink to chiller distance was rather large and was filled by the thermal goo. this was at least a 3mm gap. I assume that this is to prevent leakage or shorting from the pouch outer surface to any grounded metal on the vehicle? maybe an EE can confirm/explain this is what I am seeing? I am trying to maintain all of these gaps in my reconfigured battery.

 

[remy_martian]

The cell pairs ("modulette") that I have are fitted with the compression foam between the cell pairs in the modulette and the cell pouches are in direct contact with with the cooling plates of the modulette, but is there another set of pads or plastic film between modulettes or is it metal to metal?

I only have modulettes so am blind to how Ford & Chevrolet assembled their modules.

I checked @CyberBill's video and it's not readily evident:

 

[CyberBill]

Thanks for the tag, @remy_martian !

Funny enough, I've been watching along with the Electrified Veronika YouTube channel since they started, but didn't know about this thread.

Seems like they are on the cusp of their first drive, which I am very much looking forward to! I'm in a similar situation myself - just a few more hours of work (fingers crossed) and I should be able to drive forward a couple of feet and then back it up, right back to the lift.

The crazy thing is that back in June I was "so close to having first drive!" and was trying to push to be able to drive people around for the 4th of July. Now 6 months later I still feel months away from that goal.

Good luck D&V!!

 

[jonatalvarz]

Oh

 

[Functional Artist]

Hey guys (& gals)
Just watched your latest video (update)
...& it's looking great!

As for cooling plates, it seems like a single loop wouldn't be very thermally effecient.
(maybe not the proper terminology)

Some "food for thought" maybe think like a reverse radiator
...an "input" outlet, to a header "tank"
...thinner channels going across the coolong plate
...to the other header "tank"
...connected to the "output" outlet

Just thinking "out loud"
Possibly a flat aluminum bottom plate
...a aluminum mid plate with the channels (with inlet & outlet)
...& then, an aluminum upper plate (bolted & sealed or bonded together watertight)

The Chevy Volt battery modules use a "header, channels, header" concept similar to this

 

[remy_martian]

I think that's the kind of plate stack @CyberBill may be up to.

 

[D&VsEVJeep]

This was actually one of my first ideas... A "laminated" sheet metal cooling plate with the channels cut into the inner plates but I was concerned about the sealing of the plates...

 

[CyberBill]

Can confirm that this is what I'm doing in the EV DeLorean. I have similar concerns as Don & Veronica regarding the sealing, and I also am concerned that a little bit of pressure on the plates will spread them apart and allow the coolant to bridge. They were also not cheap - the inner channeled plates were ~$80 each from SendCutSend and I need 10, and then each side plate was another ~$15. In total it's roughly $1200, plus all the time and materials for the welding.

My plan is weld all along the outer edge, then do some spot welds (or plug welds) along the interior to prevent splitting.

I would prefer to be able to find a manufactured cooling plate like used in the Bolt and many others, where they use a stamped aluminum sheet with channels brazed to a flat aluminum sheet. I reached out to a few manufacturers without any luck. They want 1000+ plates before the prices make sense (because these are custom size, of course). And both the stamping and brazing are things you can't DIY.

A third option is to use a router to cut out a channel in a thicker sheet of aluminum, then attach a thin sheet to the top to seal it off. But again - you'll need to affix it well with either welding or some kind of gasket sealant and then fasteners.

 

[remy_martian]

Seems to pretty much cover it all.

The welding runs a danger of heat distortion, not to mention being labor intensive. It's also the approach I'm looking at but am very unhappy with it. Who wants to do, maybe, 50 feet of TIG welding for coolant plates...perfect, zero-leak, zero contamination, and oxide, welds? And you're always 3/4 of the welds behind you on an aluminum project when the welding goes out of control on you.

Here's a compendium of what I've found on the forum. Feel free to add others and other sources' approaches:

@snowdog used sealant (silicone?) and screws with milled plates - beautiful:

LG Chem battery dissection

I was looking at this: but going back and looking at it again, it seems to be the protruding end of a bolt, not a plug. Ah yes, it is a bolt with a nut on the end. Kind of a funny place for one.

www.diyelectriccar.com

@428RC considered a milled plate approach:

LG Chem battery dissection

I was looking at this: but going back and looking at it again, it seems to be the protruding end of a bolt, not a plug. Ah yes, it is a bolt with a nut on the end. Kind of a funny place for one.

www.diyelectriccar.com

I sort of liked the approach here by @tiger82, but it didn't seem robust enough to last a decade or two or three:

Modified Bolt Pack for Tesla Cobra EV Race Car

Our initial Kia Soul EV Pack performed flawlessly and was able to power our Tesla drive unit at up to 375HP at the rear wheels. However, 27kW-hr was not enough for our 20 minute races. We knew this when we put it together, but when we designed the car in 2015, there are no other packs that...

www.diyelectriccar.com

It got refined, still used Permatex Black, though it's intense for the number of plates needed:

Modified Bolt Pack for Tesla Cobra EV Race Car

Our initial Kia Soul EV Pack performed flawlessly and was able to power our Tesla drive unit at up to 375HP at the rear wheels. However, 27kW-hr was not enough for our 20 minute races. We knew this when we put it together, but when we designed the car in 2015, there are no other packs that...

www.diyelectriccar.com

 

[electro wrks]

One cooling plate design I came up with (but have not built yet) is to use inexpensive and readily available transmission oil coolers :3/8" Universal Aluminum Transmission Oil Cooler 15-1/2" x 10" x 3/4" | eBay

These come in various sizes and could be paralleled to cover an area, with care taken to match coolant flow rates for even cooling and to have sufficient combined tube cross sectional areas for the required flow rates. The fins would need to be packed with a heat transfer compound like this: Groen Heat Transfer Compound, 20Oz 161005 - Free Shipping + Geniune OEM | eBay

You might be able to sand/grind the fins down close to the tubes in these transmission oil coolers. This would reduce the required amount of the expensive and heavy heat transfer compound needed and improve the heat transfer, along with reducing the thickness. The whole mess would need to be sandwiched between two aluminum plates. It might have to be sealed to keep the compound in and uncontaminated.

 

[Functional Artist]

This was actually one of my first ideas... A "laminated" sheet metal cooling plate with the channels cut into the inner plates but I was concerned about the sealing of the plates...

I always try to come up with simple & inexpensive solutions
so,
Just thinking "out loud"

Would an externally, water-cooled (or heated) heat sink(s) be effective?

Like a copper (or aluminum) sheet installed under the existing OEM heatsink
...but, with some material, still "sticking out" on each side
...that, are wrapped around (probably) a copper tube (that the coolant flows thru)

Maybe, something like this:

The coolant could "loop" around a single module
...several modules could be "daisy chained" together
...or even cool each side independently

* Just an idea

 

[remy_martian]

In the "Modified Bolt Pack..." thread I had linked, they mention not being able to conduct heat away quickly enough from the cell to the plumbing, so conducting lengths of 3-4" may already be excessive (mind you, that's for 1mm Aluminum...).

Using a radiator is a good idea, but we're close to @CyberBill's $80 a plate. Thermal goo is not that conductive - it relies on being a very short thermal path and it is pretty expensive. Maybe bending the fins all in one direction so they are flat and overlap, then gooing those tiny gaps vs filling the space between fins?

 

[D&VsEVJeep]

I will try and use these existing plates...they have multiple paths for the cooling fluid inside to dissipate over the plate... Plus, I have the very nice mating plugs...

 

[Functional Artist]

In the "Modified Bolt Pack..." thread I had linked, they mention not being able to conduct heat away quickly enough from the cell to the plumbing, so conducting lengths of 3-4" may already be excessive (mind you, that's for 1mm Aluminum...).

Using a radiator is a good idea, but we're close to @CyberBill's $80 a plate. Thermal goo is not that conductive - it relies on being a very short thermal path and it is pretty expensive. Maybe bending the fins all in one direction so they are flat and overlap, then gooing those tiny gaps vs filling the space between fins?

IIRC they were testing for use in a race car (not your average DIY EV conversion)