How tight to compress a LFP package?

[Dirk01]

I’m in the status to design & construct some LFP packages. Now I’m wondering about the compression I’ve to consider (compression is strongly recommended due to lifetime of the package).
In a seperate forum I read not to exceed 12 psi @ 50% SoC.

Off topic: Well, psi is not a SI-standard. As a German engineer I prefer N/mm² (N/sqr mm). Would be nice to get recommendations acc. to SI. Makes it easier to compare them.

The next thing I’m wondering about is, where these pressure applies to. Or, to be more clear, which area do I’ve to consider (remembering the "Nürnberger Halbkugeln" event)? So, is it only (one side) of the end of the package? Or was that recommendation related only to a 4 piece package (resulting in a 12 VDC system) and it doubles for a 24 VDC system and quadruples for a 48 VDC system? (I can’t find it anymore but the value of 12 psi was „burned in". (Just aside: As a backup I’m considering a (very little) 12 VDC system as well. TBH, up to now, I haven’t find what for. Hence, we’ll see what will happen…).

Any recommendations are welcome.

Cheers
Dirk

[T1 Terry]

!2 psi? where on the cell and how do you measure it? Sounds like someone plucking a figure out of the air and using that to make themselves sound like an authority.
This is a 200Ah @ 12v pack, 2P4S configuration. Basically, the end plates are 3mm aluminium, the straps are stainless and the strap retaining screws are 4mm. The thread istapped into the aluminium plate and the pack compressed roughly centre to each strap using a bottle jack lying on its side pressed between two right angle steel frames bolted to the steel topped welding bench. The 1/2'(12mm) plate for the right and bracket needs to be braced with another piece welded with the flat 90* to the "L" shaped bracket to brace it and it requires 12mm bolts x 4 to bolt it to the steel bench top, it snapped 8mm bolts off at the head after only a few compressions ..... I'll let you figure out just how much pressure is applied.
It is a bit of a trial and error thing, I use 19mm thick steel plates, that I also use in the hydraulic press ... they were there and did the job so ....) to sptead the load from the round head of the bottle jack piston, but care still needs to be taken to avoid pressing the shape of the steel block into the aluminium plate ... it just doesn't look good on the finished product. I'll leave you to figure out the formula for determining how long the straps are to get the required length for the folded piece each end, too short and the strain when the pressure is released tears the piece of aluminium out out the strap simply folds that piece of the plate reducing the crush pressure .... too long and it not only looks crap, it's near impossible to get the screw in around the "L" brackets and jack assembly ....... Once you figure it out, each different size cell requires different strap placement to spread the load and the 200Ah and bigger cells need the top and bottom of the aluminium rolled over to strengthen the compression face ... otherwise it will bulge in the centre defeating the whole purpose of the compression ......

Far more complicated than just dropping a few cells in a box, but that is what is required to maintain the capacity yr on yr. As soon as the cells bulge, the capacity drops and it is a spiral to the bottom from there, so worth the effort if you plan to offer a long warranty and extened warranty attached to a pack service plan. This system, along with a lot of other things included, has delivered us over 10 yrs pack life still holding the original pack capacity and still out there functioning like the day we installed them ...... not many littium battery installers can claim a record even close to that .....

T1 Terry

[Spot]

I would be contacting the battery cell vendor and brushing up on coefficients of thermal expansion, considering the coldest and hottest conditions the batteries and bands and box will experience. Macht's gut!

[Adelie]

Originally Posted by Dirk01

I’m in the status to design & construct some LFP packages. Now I’m wondering about the compression I’ve to consider (compression is strongly recommended due to lifetime of the package).
In a seperate forum I read not to exceed 12 psi @ 50% SoC.

Off topic: Well, psi is not a SI-standard. As a German engineer I prefer N/mm² (N/sqr mm). Would be nice to get recommendations acc. to SI. Makes it easier to compare them.

The next thing I’m wondering about is, where these pressure applies to. Or, to be more clear, which area do I’ve to consider (remembering the "Nürnberger Halbkugeln" event)? So, is it only (one side) of the end of the package? Or was that recommendation related only to a 4 piece package (resulting in a 12 VDC system) and it doubles for a 24 VDC system and quadruples for a 48 VDC system? (I can’t find it anymore but the value of 12 psi was „burned in". (Just aside: As a backup I’m considering a (very little) 12 VDC system as well. TBH, up to now, I haven’t find what for. Hence, we’ll see what will happen…).

Any recommendations are welcome.

Cheers
Dirk

Off the top of my head 12psi is about 80% of atmospheric pressure, you figure it out from there.
OR
Here's a converter for you. Have fun. https://www.convertunits.com/from/psi/to/N/mm+squared

A lot of the US folks are annoyed with SI and why can't all the non-US folks include SAE values as well as SI. Until the US grows up and joins the rest of the world everybody will just have to lump it, that mean both sides will have to put up with the other..

As an engineer in the US I learned both systems and am comfortable both ways. I periodically use all SI in a post just to troll the US folks

[Adelie]

The advantages of compressing the layers is why a lot of cells are rolled. If it were me I would find some sort of multi-channel aluminum extrusion to place between prismatic cells to aid in cooling while at the same time allowing compression forces.

https://www.orangealuminum.com/inter...xoCycwQAvD_BwE

[Adelie]

So using the dimensions of the cells, you figure out what the total load would be for a given pressure. Straight forward.

How do you measure it? The only way I can think of is to test the threaded rods you want to use in a load cell and measure torque. Where do you get access to a load cell?

I can do it because I work in materials testing, but I don't see any of you wanting to pay to have the testing done.

[s/v Jedi]

For my Winston cells I took the width of the cells from the specifications and provided that exact space, rather than compressing the cells.

Physically I used 1/4” thick aluminium end plates with 1/2” square solid stock spacers in between, fastened with machine screws through the plates, threaded into the spacers.

Attached photo shows this detail during assembly.

[T1 Terry]

Originally Posted by s/v Jedi

For my Winston cells I took the width of the cells from the specifications and provided that exact space, rather than compressing the cells.

Physically I used 1/4” thick aluminium end plates with 1/2” square solid stock spacers in between, fastened with machine screws through the plates, threaded into the spacers.

Attached photo shows this detail during assembly.

I assume the cells aren't lying down like that but rather sitting upright or on their edge ..... the electrolyte settles to the lower plates when laid down the way they are in the photo and capacity is gradually lost as the upper plates dry out.

T1 Terry

[Cpt Mark]

I wondered about this too and ran across this guy who did a very nice job of explaining why he is not got to compress his battery. Info from battery manufacturers.

https://www.youtube.com/watch?v=FdDy8b5x7Y8

In short, the mfr has special equipment to exactly compress the cells which in Eve's case, are slightly concave. There is no reasonable way a layperson can accurately do this to the engineered specs. I think you engineers will like the bit about exactly how much pressure they used and what the name of the measurement is. Hint. it's not psi, newtons etc.

[GordMay]

Originally Posted by Adelie

Off the top of my head 12psi is about 80% of atmospheric pressure ...

Indeed 12 psi = 0.8165 of Atm - a very slight vacuum.

12 psi ≈ 0.0827 N/mm²
1 standard Atm ≈ 14.6959 psi [or 0.101325 N/mm²]

[s/v Jedi]

Originally Posted by T1 Terry

I assume the cells aren't lying down like that but rather sitting upright or on their edge ..... the electrolyte settles to the lower plates when laid down the way they are in the photo and capacity is gradually lost as the upper plates dry out.

T1 Terry

They are not lying down in the picture… the forum shows the whole picture rotated.

[Turtle Blues]

It is probably necessary to focus the discussion to the specific battery design and manufacturer.
The Eve battery appears to be designed in a concave casing that will absorb some expansion while the Winston appears to have a more rigid casing structure.
I would interpret the Eve graphs shown in the video to be mechanical pressure applied at atmospheric pressure unless the engineers are working in a vacuum chamber. Not sure I would be comfortable applying that amount of force to any type of stacked multiple array.
It would be interesting to see each cycled against a load cell with thermal readings.
Jedi,
Quick scan of Winston's website literature I didn't discover anything with comparable concerns. Is there something on expansion?

[s/v Jedi]

Originally Posted by Turtle Blues

It is probably necessary to focus the discussion to the specific battery design and manufacturer.
The Eve battery appears to be designed in a concave casing that will absorb some expansion while the Winston appears to have a more rigid casing structure.
I would interpret the Eve graphs shown in the video to be mechanical pressure applied at atmospheric pressure unless the engineers are working in a vacuum chamber. Not sure I would be comfortable applying that amount of force to any type of stacked multiple array.
It would be interesting to see each cycled against a load cell with thermal readings.
Jedi,
Quick scan of Winston's website literature I didn't discover anything with comparable concerns. Is there something on expansion?

It is not a manufacturer specific issue, but a general lifepo4 prismatic cell issue. The Winston cells can bulge as well. When they don’t bulge (yet) you don’t have to be compressed… just don’t give them the room to bulge.

[Jenta]

After quite a bit of YouTube hours, much reading of literature and forums and discussion with a couple others that were working on similar projects to mine, I've settled on... no compression.

I'm building a 24v pack with the EVE style 280Ah cells and to get them all to fit I'm going 8s4p. I was going to go 4p8s but my space wont allow it. Each 8s bank has its own BMS but they are only 30amp DALY ones with BT/CAN and I am not running my loads through them but instead using that connection to trigger a Bluesea Remote Battery Switch for each bank. The Quattro is monitoring the current and voltage state of the whole pack.

Anyway - back to the topic. I've chosen no compression other than using some 20mm T-Slot aluminum bar to hold the pack in place.
https://www.framingtech.com/products...mm-5mm-t-slot/
Ironically, of everything that I am doing installing this pack myself - the batteries are such a simple and relatively inexpensive part of it that IF they die in 5 years instead of 10 - I'm okay with that. Maybe by that time I would be anxious to try what is available at that time anyway.

I wanted to do Winston 1000Ah cells but I could not fit them. If I had, I would certainly have done something similar to Jedi (though I doubt I could get it as nice looking as his turned out - nice job!) to protect that investment. As it is, I've only got $3500 into my 28kw pack ($110 per including customs/freight) and I made my own busbars out of 3/16 copper bar.
I know it is a lot of money and it is to me as well - but... that's only like $500 in non-boat money.

[Steve_C]

Originally Posted by Jenta

After quite a bit of YouTube hours, much reading of literature and forums and discussion with a couple others that were working on similar projects to mine, I've settled on... no compression.

I'm building a 24v pack with the EVE style 280Ah cells and to get them all to fit I'm going 8s4p. I was going to go 4p8s but my space wont allow it. Each 8s bank has its own BMS but they are only 30amp DALY ones with BT/CAN and I am not running my loads through them but instead using that connection to trigger a Bluesea Remote Battery Switch for each bank. The Quattro is monitoring the current and voltage state of the whole pack.

Anyway - back to the topic. I've chosen no compression other than using some 20mm T-Slot aluminum bar to hold the pack in place.
https://www.framingtech.com/products...mm-5mm-t-slot/
Ironically, of everything that I am doing installing this pack myself - the batteries are such a simple and relatively inexpensive part of it that IF they die in 5 years instead of 10 - I'm okay with that. Maybe by that time I would be anxious to try what is available at that time anyway.

I wanted to do Winston 1000Ah cells but I could not fit them. If I had, I would certainly have done something similar to Jedi (though I doubt I could get it as nice looking as his turned out - nice job!) to protect that investment. As it is, I've only got $3500 into my 28kw pack ($110 per including customs/freight) and I made my own busbars out of 3/16 copper bar.
I know it is a lot of money and it is to me as well - but... that's only like $500 in non-boat money.

Hi Jenta, where did you find cells for that price? Massively cheaper than Winstons. I am getting ready to build a new bank for my ordered Lagoon.

Thanks!