Cross-checking my ideas for my Battery v2 after Battery v1 Nighmare

[silverstone]

So @upnorthandpersonal I tried to emulate your design, just to cross-check some dimensions.

I could barely made it below 600mm width and I'm not sure if in real life I can get it working.



Legend:
- Thick Beige is 45x100 mm wood for framing
- Medium Red is "Lister" 22mm x 100mm
- Thin Orange is Plywood 12mm thick
- Purple is compression steel block
- Green / Dark Blue is each 12mm plywood, used for mounting plate & "door"


There seems to be quite some overhead in terms of unused space.

Should I:
- Use 38 x 57 mm wood instead of 45 x 100 mm ?
- Use thinner "Lister" (non-treated I can only find 19mm thick and it's more expensive, otherwise 22mm)
- Use plywood all the way ?

 

[silverstone]

I think I'll go get the wood and threaded rods tomorrow.

Maybe I can avoid the INNER wood layer so I can stick the treated rods and compression steel bracket into where the big wood framing is. Potentially saving up to 80-90 mm in width.

The i just do framing and outer wood layers .

 

[Skypower]

I went with two side by side modules so I can pick them up. They are on a coaster board with corner brackets to keep them located and nested. They have a centralized compression spring. The most important thing I learned was that the pusher board and end plates needed to be at least two 3/4” plywood sheets glued together so the wood doesn’t deform/bow in time( If you plan on any forces). Header paint may give you a bit more fire protection on wood parts. Four pieces of all-thread would be easier though. My method needs a good table saw for good glue up.

 

[silverstone]

I went with two side by side modules so I can pick them up. They are on a coaster board with corner brackets to keep them located and nested. They have a centralized compression spring. The most important thing I learned was that the pusher board and end plates needed to be at least two 3/4” plywood sheets glued together so the wood doesn’t deform/bow in time( If you plan on any forces). Header paint may give you a bit more fire protection on wood parts. Four pieces of all-thread would be easier though. My method needs a good table saw for good glue up. View attachment 191410View attachment 191411View attachment 191412
View attachment 191413View attachment 191414

Looks very neat, congrats .

The issue with the threaded rods is that the requirement in terms of width increases (you have your cell then ~ 15-20mm + threaded rod 12mm + a bit of clearance on each side). Plus more if you use a wood framing scheleton like @upnorthandpersonal did and I was also planning. But the compression can be very good (although "by feel" and not as precise as with springs).

Battery v1 I also had an unintended "feature". I didn't want the terminals on Cell 01 and Cell 16 to be too close so when you torque and don't pay enough attention you inadvertently short your battery ... Didn't have that problem because I had the second series of cells in the same "order" (polarity) as the first one. So cell 08 and Cell09 needed a Cable Jumper.

I was also thinking of avoiding that and using a high "wall" like you did. Most likely polycarbonate or something similar, since it can be very thin then (2-3mm will be plenty).

I have a table saw but not very good at using it. It's almost inaccessible anyway, so I usually just end up using the circular saw .

Not sure what you meant by "Header paint" for fire protection. Are you sure it's non-electrically conductive ?

 

[Skypower]

Yeah, wood is going to be longer(mine) or wider(all-thread). That’s the nice thing about a metal box, some have a smaller footprint. My crossover buss/jumper (two ea. 2awg) work out to have the exact same resistance (internal resistance meter) as the flexible buss bars so keep that in mind when doing this or you could have some balance issues. Your cable /setup may vary. I probed the paint and it has zero conductivity. I did a test piece before painting it. Great minds

 

[silverstone]

Yeah, wood is going to be longer(mine) or wider(all-thread). That’s the nice thing about a metal box, some have a smaller footprint. My crossover buss/jumper (two ea. 2awg) work out to have the exact same resistance (internal resistance meter) as the flexible buss bars so keep that in mind when doing this or you could have some balance issues. Your cable /setup may vary. I probed the paint and it has zero conductivity. I did a test piece before painting it. Great minds

I did not manage smaller footprint with metal box. Quite the opposite. Fair enough, might also be because it's a standard 1.2mm (or 1.6mm) thick panel. So it will never be able to take the pressure of the cells when they are expanding. Thus it needed to be reinforced anyway.

I'm not sure in my case what's the higher priority ... Treatment to fireproof wood or waterproof it . Both would be ideal.

Battery will be sitting on 200mm wheels and 50mm extra wood below I guess from the floor. Hopefully the garage will never be flooded ...

 

[Skypower]

I did not manage smaller footprint with metal box. Quite the opposite. Fair enough, might also be because it's a standard 1.2mm (or 1.6mm) thick panel. So it will never be able to take the pressure of the cells when they are expanding. Thus it needed to be reinforced anyway.

I'm not sure in my case what's the higher priority ... Treatment to fireproof wood or waterproof it . Both would be ideal.

Battery will be sitting on 200mm wheels and 50mm extra wood below I guess from the floor. Hopefully the garage will never be flooded ...

Remember any type of wood mold or fire treated/treatment greatly affects the glue’s adhesion, so if you are planning on any structural gluing, treat it after assembly.
These little wheels from Amazon greatly exceeded my expectations for smoothness and ease of direction change:

 

[silverstone]

Remember any type of wood mold or fire treated/treatment greatly affects the glue’s adhesion, so if you are planning on any structural gluing, treat it after assembly.
These little wheels from Amazon greatly exceeded my expectations for smoothness and ease of direction change:View attachment 191421

Definitively in another league ... I get one, you get 4 for the same price, more or less




Reason for dimensioning was clearance w.r.t. floor, NOT weight .


But does the fire retardant treatment actually works or it's like putting a small patch on top of a huge bleeding ?

 

[Skypower]

One thing I’ve noticed about plywood is, the prettier the grain, usually the weaker it is. Expensive grain nice grain wood is usually applied in a thin outer veneer. That’s terrible for shear strength. You want to select a plywood that has a thicker outer ply, well glued to the others and the inner ply’s aren’t made of super soft wood(finger nails easily mark).

 

[silverstone]

One thing I’ve noticed about plywood is, the prettier the grain, usually the weaker it is. Expensive grain nice grain wood is usually applied in a thin outer veneer. That’s terrible for shear strength. You want to select a plywood that has a thicker outer ply, well glued to the others and the inner ply’s aren’t made of super soft wood(finger nails easily mark).

That's why I'm considering using some 22mm x 100mm solid woods instead actually. A bit thicker, but stronger and same price more or less.

But fireproofing wood around here doesn't seem so easy as in the US ...

 

[Skypower]

Definitively in another league ... I get one, you get 4 for the same price, more or less

View attachment 191426


Reason for dimensioning was clearance w.r.t. floor, NOT weight .


But does the fire retardant treatment actually works or it's like putting a small patch on top of a huge bleeding ?

Looks like you got what you paid for though. Mine is low so it rolls under the others on the workbench lower shelf.
A good fire retardant keeps the wood from burning when a flame is removed. Doesn’t keep it from charring. The plywood shelf above the rolling battery is treated.

 

[silverstone]

Looks like you got what you paid for though. Mine is low so it rolls under the others on the workbench lower shelf.
A good fire retardant keeps the wood from burning when a flame is removed. Doesn’t keep it from charring. The plywood shelf above the rolling battery is treated.

Again, I'm not so sure it's so easy to get those paints/coatings/primers etc here in Europe.
Any thoughts @upnorthandpersonal ?

But why aren't you rather focusing on prevention / mitigation ?
My (overkill) plan is:
- JK BMS Battery Monitoring
- Trip Circuit Breaker using Shunt Trip Unit, triggered by
a. Smoke Alarm
b. Temperature Monitoring Relay or Arduino/ESP32 + DS18B20 + Solid State Relay
c. Wiring Loop in Battery Compartment through some Temperature Switches (maybe 55°C)

In theory Thermal Runaway shouldn't be an issue with LiFePo4.
You remove the current and it should stop. And I'm guessing if you catch it early enough, there shouldn't be enough heat to self-combust the wood after removal of the flame/source.

EDIT: Or are you afraid that, under normal operating conditions with the battery say at +40°C in a hot climate, the wood will eventually dry out completely and start to combust ?

I once found a couple of reports about long-term and short-term temperature withstand of wood ....




So 256°F is 124°C ...

While another reference I had found back then




So not sure. The first reference points to sustained long-term 124°C without issues, the second one even short-term 77°C shows issues .

 

[silverstone]

Is there something that could be put on top of fibergips to avoid having all the dust floating around ? Something like a polyurethane coating or maybe just regular water paint ?

Fibergibs can be (quite) fire retardant while still allowing to install screws and support (some) weight, unlike regular gibs.

Just thinking if it's better to install on the inner or outer side of the battery ...

 

[LydMekk]

You don't think you're overdoing this a tiny, tiny bit? :D

 

[upnorthandpersonal]

Again, I'm not so sure it's so easy to get those paints/coatings/primers etc here in Europe.
Any thoughts @upnorthandpersonal ?

I don't bother with any of that. Simple fixture for the cells, insulation around if needed, and start using the battery.

 

[silverstone]

You don't think you're overdoing this a tiny, tiny bit? :D

Not as much as battery v1 hopefully .

Steel enclosure would be nice but it's absolute pain, cost and time consuming to build.

 

[silverstone]

I don't bother with any of that. Simple fixture for the cells, insulation around if needed, and start using the battery.

Just got the wood now. Originally planned to do something this weekend, but some mouses had some other plans for me. Damn them .

As I said, some fibergips plate might be helpful to help w.r.t. fire behaviour. Not sure if I have the space though.

But then again, we are talking essentially about a thermal runaway of sorts, which is NOT stopped:
- When the BMS trips (overtemperature)
- When the Smoke Detector trips the shunt trip of the Breaker (smoke detected)
- When the Temperature Switch / Thermostats trip the shunt trip unit of the Breaker (ambient air overtemperature)

Not saying it couldn't happen, but from what I read here if you remove the SOURCE of the heat/fire (e.g. loose connection stops being a problem when current drops to zero) everything stops.

Only exception would be an internal short circuit of one of the cells. That you cannot do anything about except possibly hope that it doesn't propagate. Or maybe trigger an emergency DISCHARGE so you force most the energy out before it propagates through some DEDICATED resistor ??? Otherwise being the cell with the lowest voltage it will actually tend to sink in all the energy of the other batteries ...

 

[upnorthandpersonal]

thermal runaway

And that thermal runaway doesn't really exist with LFP, off-gassing does not typically result in a fire: is not the same as Cobalt based runaway (i.e. it does not amplify itself), and needs an external ignition source when off-gassing.

 

[silverstone]

And that thermal runaway doesn't really exist with LFP, off-gassing does not typically result in a fire: is not the same as Cobalt based runaway (i.e. it does not amplify itself), and needs an external ignition source when off-gassing.

Yeah, I saw Will's and your posts in that other topic from a few years ago .

There seem to be still some cases of mysterious fires inside the forum, mainly:
- Loose connection (not a problem if you disconnect using a Breaker)
- Internal short / failure (a problem)

Maybe it's not a "real" thermal runaway, but if one cell cooks, nearby ones can also get damaged.

And potentially the wood near the cell could catch fire.

 

[silverstone]

Here they shorted a 160Ah LiFePo4 battery and it reached almost 200°C

(PDF) Temperature, Overcharge and Short-Circuit Studies of Batteries used in Electric Vehicles

PDF | The paper presents the results of temperature and short-circuit research of battery types most commonly used in electric vehicles. Basing on... | Find, read and cite all the research you need on ResearchGate

www.researchgate.net

How would scale up with regards to capacity ?

Small Brainstorming ....

Lower internal resistance, higher peak short-circuit current, more thermal mass with a 280Ah/304Ah battery compared to 160Ah.

But at first sight, in relative terms, if both batteries are meant for 0.5C normal use, then the ratio of internal resistance is 1/2.
On the other side, the ration of short circuit current will be 2/1. Assuming the internal resistance is sized according to the voltage drop. Or the power losses per unit of volume/mass, in that case also the resistance should have a factor of 2/1.

The power losses will be ~ R x I^2 so they should be twice. But we also have twice the thermal mass most likely ...

The discharge time will be the same (assuming ~ 10 C for short circuit current peak/max), so the energy will also be ~ R x I^2.

So the LF304 should reach 200°C ?

Please correct if something is entirely wrong ...