LF628 batteries just arrived, compress or nah?

[jedics]

I was just going to wire them up with a space between each to allow for expansion and not stress the terminals and call it a day but then I got on the internet.

Does the expansion and contraction of charge/discharge delaminate the layers and reduce cycle life or are the alloy boxes they are made with enough?

searched here for battery compression and found nothing surprisingly!

[Dacian Todea (electrodacus)]

Yes the expansion and contraction will significantly reduce the cycle life thus the need for compression to improve cycle life.

Check the spec of your cells and they will provide you with a range of compression force for optimal cycle life.

The cells will still expand and contract even with compression but much less and thus the significant cycle life difference.

Here is the spec for your particular cells https://paircycle-batteriespeicher.de/wp-content/uploads/EVE_MB56_628Ah_Datasheet.pdf

So they recommend a 2mm EVA foam between cells or silicone foam those will act like a spring.

Then while cells are relatively discharged 15 to 40% SOC you apply 3kN to 7kN of compression force.  This force will of course increase while cell is fully charged 100% SOC

With age this force will increase and mention 45kN at 70% SOH so by the end of life.

Max pressure on poles (Table 11) is 0.5kN (500N) so you need to use flexible bus bars else cells will be damaged.

[Dan Karau]

There is no need to over think this, I used ready rod with nylock bolts and 3/4" plywood and just ensured that everything was tight but nothing crazy.   Pack has been in use (fulltime RVing) for 4 years now and still balance to within 6-7mv.  Also need to consider that this pack has driven 1000s of miles across Canada/US in that time.  If you watch any of Andy's videos as he puts together his packs you will see just how little pressure is needed (0.5kN) is across all your bolts not for each bolt!   https://www.youtube.com/@OffGridGarageAustralia

There is no better resource than Andy when it comes to these batteries.

[jedics]

Thanks for simplifying the process, seems straight forward enough now. I will probably put it all together with the 2mm foam separators and hand tighten the nuts, then use a spanner until the foam begins to slightly deform to get about the right compression?

[jedics]

But he is the main person I found who says you dont need to compress the cell at all.

[jedics]

I have these already, anyone see a problem with me just curving the fine stranded wires like the connector at the back of this photo to become the "play" to take up the expansion?

[Dacian Todea (electrodacus)]

There is no problem to bend those as the one in the back.

[jedics]

thanks, your a credit to your species answering all our questions so quickly after 7 years :)

[jedics]

I will probably use 6mm aluminium plate to compress, I feel like I need to put the 8M rods on the top and bottom (Green) to reduce flexing in the centre rather than on the sides (Blue) or is the amount of pressure so low that it doesnt matter which I choose?

I cant put the two rods on top between the terminals as it will get in the way of my thick cable connectors.

[Dacian Todea (electrodacus)]

You can use my setup as inspiration

Your cells are twice as wide and I think 6mm plate will not be thick enough.

Here is my other battery made with A123 20Ah LFP pouch cells this is 9 years old and still works great. These use 10mm thick plates.

If you will be using PVA foam between cells then you do not need the springs.

Either use thicker plates or add L or C channels and just those 4 blue rods. Also leave space so you can isolate the rods in my case I used clear tubing so there is no chance of cell casing getting electrically connected.

[jedics]

Thanks for the pics, yes will use 10mm plates

[jedics]

So 1 to 3 nm of torque per nut, is doing them up as tight as I can with my fingers and another half turn of each nut going to be close enough?

[Dacian Todea (electrodacus)]

I do not think you can estimate the torque with this sort of method.

A cheap torque wrench seems to be 30 to 40CAD on Amazon just make sure it can start below at 2 or 3Nm set it at 5Nm and you should have a good connection maybe for the rest of the battery life.

Set it at 6Nm for those large cells for the smaller ones I used 5Nm.

[jedics]

Where did you get the 6nm number from, thought it was supposed to be 3nm spec to apply to each nut which did sound very low to me.

[Dacian Todea (electrodacus)]

It was 6Nm max for the M6 terminals on the 280Ah cells. I applied 5Nm to mine and never had to do anything since they where installed.

You have M8 terminals so larger but I think the 6Nm may have to do with the thin aluminium case and that is about the same for both type of cells.

See page 17  Max Torque on Poles   https://www.gobelpower.com/download/EVE-MB56-628Ah-LiFePO4-Cell-Specification-Datasheet.pdf

[jedics]

I am talking about the amount of compression for each 8M nut on the 4 threaded rods that will compress the entire 8 cells.

[Dacian Todea (electrodacus)]

:) That is a very different question.

It is much more complicated to calculate torque Nm to compression force since there are to many unknown.

They suggest 3000 to 7000N of compression force.

Say the pitch of your thread is 1.25mm standard for M8 but you may have a different pitch.

The equation for ideal case where there is no friction between the nut and threaded road.

F = (2*3.14 * T) / 0.00125m

So say you apply 5Nm

F = (2 * 3.14 * 5Nm) / 0.00125m = 25120N  so way to high in ideal conditions but in reality there is quite a lot of friction but highly variable depending on the road and nut surface and if there is any lubrication or not.

If you will have used some springs you could directly say what the force is based on amount of spring compression.

I will that if you add 2mm EVA foam between cells (also cells and plates). Have the cells at around 40% SOC and tighten the bolts with 3 to 5Nm as in real wold due to friction the compression force is just around 15% of the ideal value so 0.15 * 25120N = 3798N

[jedics]

Does anyone have links to right spring pressure?

[Dacian Todea (electrodacus)]

The manufacturers or sellers will have the spring specs.

Here is a link https://www.aliexpress.com/item/1005009115563879.html  you need to expand "View more" to see all spec

You will need either 9 or 10mm internal diameter (for 8mm threaded roads) and either green or brown.

So say you get 10mm inside diameter brown then fully compressed the force is 1568N per spring (you do not want to fully compress so say 1000N for each spring) Then 4 springs will provide 4000N if you compress them so length is reduced by around 12% (20% is max for brown).

[jedics]

My torque wrench arrived the day before I saw this and to use it I had to cut off the threaded rod to get the socket on so now they are to short to take the springs. I would have to buy 4 new rods and either get off the hose covers or buy some more. What is the consensus on just using nuts/spring washers and the wrench at 2nm of force for each nut with the eva foam?

 

[Dacian Todea (electrodacus)]

There is no need for springs in your case. What SOC are your cells currently ?

You may want to check some weeks in to the future when you know your battery SOC is 30 to 40% and check to see if you still have 2Nm then it should be fine.

Maybe replace the nuts with nylon insert lock nuts especially if this battery will be installed in a vibration environment.

[jedics]

All cells were delivered at 3.26v. My rear nuts are nylon and the front ones I use to tighten have spring washers to stop the nuts backing out.

Am I right in thinking that my method of holding the batteries to the floor needs to also allow for the movement of batteries breathing when charged/discharged? 

Eg. simply bolting the end plate to the floor after compression will stop them being able to expand when charged after.

[Dacian Todea (electrodacus)]

Cells will only expand on one direction the one you already clamped them. The top or bottom of the cells as well as the narrow side do not need compression.

If your floor is conductive you will want to add something non conductive to make sure the cells remain isolated.

[jedics]

I finally got the pack "fully charged" where the smbms stopped charging but that was at about 27.4v shouldnt it be getting up to at least 28v?

[Dacian Todea (electrodacus)]

That seems about right.

When cells get around 3.4V they need just a few seconds max a few minutes to get to full charge 3.55V default value. There will always be one cell to get there first so while most cells are super close (a fraction of a percent from fully charge) one cell will get there first.

So you have around 3.4V * 7 + 3.55 * 1 = 27.35V    

[Dave McCampbell]

Did you top balance your cells first?  Do you know what your cell internal resistances and max capacities are?  Are you using the SBMS0 to balance your cells?
  
To me, 150 mV seems like an unusually high delta between cells, even given the small amount of capacity lost by terminating the charge early.  As I recall in an earlier discussion, Dacian mentioned that the SBMS0 internal 200 mA balancer typically kept his cells at about a 100 mV delta.  I think that assumes that the cell internal resistances are close.  That right Dacian?

We top balanced our 12 EVE cells after receipt and checked the internal resistances and capacities to make sure they matched the very tight deltas on the EVE testing paperwork.  After a year in use on our boat overseas, we are keeping our cell deltas at max 5 mv, but using an external NEEY active balancer occasionally, rather than the SBMS0 passive balancer full time.  So two different ways to do cell balancing.

We are using a fiberglass compression box with 4 threaded rods to a plate on one end rather than just two wood end pieces.  It is held firmly in place under a seat so can't move even if we go inverted.  I think Dacian's aluminum end pieces are better than wood.  An aluminum box might be better on a boat.

[jedics]

I just installed them as is and let the sbms120 balance them, hopefully is will sort them out over time. From what Dacian says it should balance them eventually now its summer getting them to full every day? I dont have the tools to check internal resistance but they came all the way from Chain so what am I going to do if they are out by a lot? I wish I had as much attention to detail as you  Thanks for the help everyone.