Compress or not, flexible busbar or not

[Cheap 4-life]

Yes but that rate you are taking about would have to match the expansion rate of the cells to keep from over compressing at 100% SOC. I am top balancing a new set of four cells.

Installed compression fixture set to 640lbs which is just below the 12 psi recommended by EVE, I am parallel top balancing with a 10amp power supply. They have expanded 1/16” 18 hours Into a 72 hour top balance with the recommended compression, there is no way threaded rod is going to stretch enough to keep the pack from being over compressed.

With springs I can set the pressure and based on the compressed length tell what pressure the pack is under, that is what die springs are used for. If over compressing wasn’t an issue they wouldn’t have included it in their testing and given a value not to exceed.

Everything you said above made perfect sense to me, however,,, it’s starting to seem like keeping the psi at only 12 with springs is not what should be done. It lets the cells expand to much.. 12psi should be set at low SOC and then as the cells try to expand more pressure is to be applied somehow to prevent expansion from occurring at all. Although some very minimal expansion is unpreventable. Expansion can be stopped in a rigid fixture without springs and using some poron foam to lessen the possibility of over compression at high SOC. Unless a spring can be used that gives exactly 12 psi and as the cells try to expand the spring applies more pressure, but I don’t see how that’s possible. The springs could only apply more pressure if expansion was occurring and from what I’m gathering we don’t want expansion to happen. We want to prevent it as much as possible..
So @justgary will be pleased,, I’m starting to agree with him..

 

[Cheap 4-life]

More thoughts on compression.

I've seen EVE's videos on youtube where they put pressure on a stack of batteries and then laserweld it into a fixture, and it seems like a common method on this forum, minus the welding.

But I can't square this with the battery spec. sheet. I'll use 1kgf = 10N for ease of conversion. Numbers for LF105 since that's what I intend to use for my UPS battery:


View attachment 113047

So we can see that they should have 300-500kgf of compression, shouldn't peak over 700kgf and at 900kgf it will cause damages, while 1500kgf will cause leakage.

If we then look at the expansion force the cells themselves generate:

View attachment 113048

At Beginning Of Life they'll generate less expansion force than the recommended compression, but at End Of Life they can generate 3000kgf. This obviously exceeds the force that'll crush the other cells. With multiple cells stacked in a static fixture this sounds like a countdown to cascading failure.

If I instead use 6x 32 mm/16mm hole/inner diameter ISO-10243 medium load springs that are 102mm long I get 32 N/mm/spring. Compressing these 15.625 mm give 15.625 mm * 6 springs * 32 N/mm = 3000 N ≃ 300kgf. If a cell nearing end of life decides to start bloating, it can violently bloat 10mm and I'd still only put 25.625 mm * 6 springs * 32 N/mm = 4920 N ≃ 492 kgf on the other cells, i.e. still remaining within the recommended compression force parameters.

Ignoring having to use flexible cell interconnections and other considerations that are besides the point, why wouldn't this be the preferred solution?

Comment #341 explains my newest viewpoint on springs, and I was all for springs

 

[thulle]

Unless a spring can be used that gives exactly 12 psi and as the cells try to expand the spring applies more pressure, but I don’t see how that’s possible.

But that's exactly how springs work, the force gets higher as the springs get squeezed. You just select springs that add the appropriate amount of N / mm of deflection. In my example I picked one to allow 10mm while not adding more than 2000N (200kgf) to the original 300kgf.

edit: This has also been explained to you earlier in the thread.

 

[thulle]

Again, the new EVE datasheet implies that this [overcompression, my note] is no longer a concern. You don't seem to agree, but that's fine with me.


Please point to the part of the datasheet that says the cells will be "overcompressed."

The "overcompression" is what I mentioned in my post earlier.

Maybe it's this comment I should address:

Compression is now apparently clarified in the datasheets, so it may be best to not cling to the information in older datasheets. They now mention a compression force range, which (for 230 Ah cells anyway) happens to equate to about 3kN (11.7 PSI) to 5kN (19.6 PSI). The implication is that this compression is the normal range of a cell taken from 30% SOC at 11.7 PSI to 100% SOC in a hard fixture (no springs or foam). Under normal conditions, the force should not exceed 7kN (27.5 PSI). It can exceed that if internal damage or cell leakage occurs.

In other words, when they said 11.7 to 18 PSI before, it seems they didn't mean don't let your compression exceed 18 PSI, they meant that if you clamp the thing rigidly the internal force will not exceed 18 PSI. This is what the new datasheet seems to say (well, now 19.6 PSI).

My bolding. The phrasing in the datasheet is "It can be seen from the above table, that the compression force of the cell exceed 9kN, otherwise the cell may be damaged.

I'm not interpreting this the same as you are. I read it as compression force -> result. The 6.2 test is specified as they're compressing the cell at 0.02 mm/s, slowly increasing the applied force, and at 9kN it results in cell damage.

 

[fafrd]

More thoughts on compression.

I've seen EVE's videos on youtube where they put pressure on a stack of batteries and then laserweld it into a fixture, and it seems like a common method on this forum, minus the welding.

But I can't square this with the battery spec. sheet. I'll use 1kgf = 10N for ease of conversion. Numbers for LF105 since that's what I intend to use for my UPS battery:


View attachment 113047

So we can see that they should have 300-500kgf of compression, shouldn't peak over 700kgf and at 900kgf it will cause damages, while 1500kgf will cause leakage.

If we then look at the expansion force the cells themselves generate:

View attachment 113048

At Beginning Of Life they'll generate less expansion force than the recommended compression, but at End Of Life they can generate 3000kgf. This obviously exceeds the force that'll crush the other cells. With multiple cells stacked in a static fixture this sounds like a countdown to cascading failure.

If I instead use 6x 32 mm/16mm hole/inner diameter ISO-10243 medium load springs that are 102mm long I get 32 N/mm/spring. Compressing these 15.625 mm give 15.625 mm * 6 springs * 32 N/mm = 3000 N ≃ 300kgf. If a cell nearing end of life decides to start bloating, it can violently bloat 10mm and I'd still only put 25.625 mm * 6 springs * 32 N/mm = 4920 N ≃ 492 kgf on the other cells, i.e. still remaining within the recommended compression force parameters.

Ignoring having to use flexible cell interconnections and other considerations that are besides the point, why wouldn't this be the preferred solution?

I’ve not seen that data for cell expansion force at beginning of life and end of life before - does the latest datasheet for the 280Ah cells (or whatever 300+Ah rating they are using for similar-sized cells now) contain similar cell expansion force data?

I built a 300kgf clamping fixture for my 16 280Ah cells and this cell expansion force specification is making me glad I did.

As my cells age, if they begin to exert more expansion force, my fixture will adapt and allow my cells to expand however much they need to without ever applying much over 300kgf.

Your question about how a rigid fixture can work in the face of increasing cell expansion force boils down to the cycles-versus-expansion-force data that we don’t have.

In an EV, for example, batteries are you ally replaced at 80% of original capacity. If by that point, cell expansion force for an WV battery housed in a rigid structure is less than 700kgf, there should be no issue.

I also suspect that EV batteries housed in rigid structures are assembled with cells that are full or at least 50% full. This would mean that when fully drained to 10%,20% or whatever lower limit has been set, there is likely to be space between the middle of the sidewalks of cells and they are being held imposition by their own rigid tops and bottoms.

The point being that I can easily see how a rigid structure housing cells that are starting off 50-80% charged may work without issue for an EV battery only being cycled to 80% of original capacity.

But for any application where cells will be cycled all the way to whatever the datasheet defines as EOL, I’d sure rather have a clamping fixture than rely on a rigid housing.

But of course, if cell expansion force only increases to ~700kgf once cell capacity has declined to ~50% of initial capacity, probably not a big deal for those of us DIYers using relatively new cells…

There is a whole aftermarket where used cells only delivering 70% or 80% or original datasheet capacity get resold as smaller-capacity cells or smaller-capacity batteries (sometimes more than once).

So DIYng (or buying cheap batteries) using cheap recycled cells housed in a rigid structure sounds as though it could be disasterous…

It’s a pity your datasheet did not also disclose the number of cycles to a cell expansion force of 7kN…

 

[thulle]

I’ve not seen that data for cell expansion force at beginning of life and end of life before - does the latest datasheet for the 280Ah cells (or whatever 300+Ah rating they are using for similar-sized cells now) contain similar cell expansion force data?

Interesting point. I don't have the LF280 datasheet, but in the newest LF280K sheet, version A 2022.04.20, that Amy Wan posted in August it says the following:


and then on compression force it says:



So, these have up to 50 kN expansion force at EOL, and compression force cannot exceed 50 kN. I'd read this as you just barely should be able to stack the cells without causing internal defects in other cells from one going into EOL expansion.

But note that this is how I read the data, and I'm in this thread to try to clarify my reading, and this is my first endeavor to build a LFP battery.

Your question about how a rigid fixture can work in the face of increasing cell expansion force boils down to the cycles-versus-expansion-force data that we don’t have.

Yes, I haven't seen any data like that either. Nor how much one could expect the cell to expand with a given force.
My plan is to automatically monitor the spring deflection, but that's more of another layer of safety by checking if anything starts to deviate. And since my battery is for a UPS it won't be cycled enough to give any data in the nearest decades.

In an EV, for example, batteries are you ally replaced at 80% of original capacity. If by that point, cell expansion force for an WV battery housed in a rigid structure is less than 700kgf, there should be no issue.

I have no clue about what additional monitoring exists in cars using rigid fixtures. They might just disable any use of the battery long before reaching EVE EOL. Thus I feel that I can't just adapt that practice and point to others doing it.

 

[Cheap 4-life]

But that's exactly how springs work, the force gets higher as the springs get squeezed. You just select springs that add the appropriate amount of N / mm of deflection. In my example I picked one to allow 10mm while not adding more than 2000N (200kgf) to the original 300kgf.

edit: This has also been explained to you earlier in the thread.

The point is that expansion should be prevented not allowed to happen with springs.... The only way the springs can apply more pressure is if the cells are expanding.. as I just explained

 

[thulle]

The point is that expansion should be prevented not allowed to happen with springs.... The only way they springs can apply more pressure is if the cells are expanding.. as I just explained

And as I said in my first post on the matter, the pressure the cells I'd use would exert on another near EOL in a static fixture seem to be enough for them to destroy each other according to the data sheet. We got an unknown here, as fafrd put it, we don't know the increase in expanding force as the cells age. In the case of LF280K it doesn't seem to matter, but in the case of LF105 it does.

 

[fafrd]

Interesting point. I don't have the LF280 datasheet, but in the newest LF280K sheet, version A 2022.04.20, that Amy Wan posted in August it says the following:

View attachment 113061
and then on compression force it says:

View attachment 113062

Interesting. I saw the 280K datasheet when it first came out (and added the data about increased lifetime with a 300kgf fixture) but there was no data on expansion force in that early version.

So, these have up to 50 kN expansion force at EOL, and compression force cannot exceed 50 kN. I'd read this as you just barely should be able to stack the cells without causing internal defects in other cells from one going into EOL expansion.

Again, knowing what the expansion force at HOL (half of life) sure would be helpful, but my general sense is that it’s going to be practically impossible to design a rigid fixture that remains rigid under 5000kgf(!).

So I take this new datasheet to mean that enclosing these cells in a rigid metal box such as that used for an EV battery will be fine.

But note that this is how I read the data, and I'm in this thread to try to clarify my reading, and this is my first endeavor to build a LFP battery.

If I had seen this datasheet before building my 300kgf clamping fixture, I would not have bothered.

On the other hand, if you are not using a ready-made box for the purpose and are already going the route of rigid end plates secured by threaded rod, the additional cost and effort to add calibrated springs to apply 300kgf is modest.

I don’t see anything in this latest datasheet specification to suggest that lifetime increases if compression force of greater than 300kgf is applied (though the risk of causing damage by applying even 3000kgf would appear to be low).

Yes, I haven't seen any data like that either. Nor how much one could expect the cell to expand with a given force.

Unconstrained 280Ah cells expand ~1mm in the center of the cell (from a starting point that might be sunken with respect to the rigid top and bottom caps).

My approach was to ‘calibrate’ my 300kgf clamping fixture when my cells were at their lowest usable SOC. Calibrated springs allow your to translate force to a number of turns of a nut on a threaded rod, so X turns on each nut from first compression translates to 300kgf.

My springs were chosen to maintain close to 300kgf over a compression range of 1mm-per-cell (16mm for the full pack) but my first charge to maximum SOC expanded by a fraction of that (possibly 2-3mm, possibly nothing at all).

If I’d seen this data and wanted to go to the trouble of building a clamping fixture, I would have had no concerns using a shorter calibrated spring (though the only advantage would have been slightly shorter overall dimension, since spring cost would have been about the same…).

My plan is to automatically monitor the spring deflection, but that's more of another layer of safety by checking if anything starts to deviate. And since my battery is for a UPS it won't be cycled enough to give any data in the nearest decades.

Been there, thought about doing that, decided it wasn’t worth the trouble and wouldn’t even consider it after one year of experience with my battery now.

Getting SCC and parameters set up to correctly charge and discharge cells within target voltage ange is a much bigger concern to me than anything to do with expansion or clamping force at this stage.

I got pretty lucid with cell matching but have a ‘runner’ (cell with lower capacity than the others) so making sure that cell is the one limiting capacity (meaning both the first to reach lowest SOC as well as the first to reach maximum SOC is my greatest concern.

The BNS is there as a fall-back safety device but when the BMS cut-off kicks in, it causes a fault to the overall system requiring a manual reset.

So my biggest concern now is keeping an eye on that weakest cell to assure I’m staying one step ahead of it (reducing usable capacity faster than mismatch increases).

I have no clue about what additional monitoring exists in cars using rigid fixtures. They might just disable any use of the battery long before reaching EVE EOL. Thus I feel that I can't just adapt that practice and point to others doing it.

I doubt any EV battery out there will not deform under 5000 kg of force. So first, batteries are unlikely to be used for anywhere close to the service life needed to generate 5000kgf of expansion force, and second, even if they did, the metal box is unlikely not to deform a bit and hence apply less than 5000kgf.

In addition, even if they do get exposed to 5000kgf, worst-case would be formation of some ‘internal defects.’

‘Leakage’ does not occur until over 100,000 kgf (before which I suspect the battery box would burst).

 

[fafrd]

The point is that expansion should be prevented not allowed to happen with springs.... The only way they springs can apply more pressure is if the cells are expanding.. as I just explained

I don’t know where you think you are seeing that.

The point is that cells allowed to expand under a clamping force of 300kgf will have a higher cycle life than cells allowed to expand without any pressure.

Beyond that, the old thought was that too high if a clamping force could also degrade cycle life but thus new data indicates that applied clamping force can be as high as 5000kgf before you have anything to worry about.

With a 300kgf clamping fixture, the cells basically don’t move at all after ‘settling in’ over the first few charge/discharge cycles (the rigid top and bottom surfaces limit compression and hold the cells in place, even after the internal surfaces of cell walks have relaxed to the point there is little or no pressure and they may not even be touching).

Us ‘old timers’ who struggled with Gen 1 300kgf clamping specs had to be concerned about applied clamping force being too high.

With this new data, you Gen 2 DIYers motivated to make a calibrated 300kgf clamping fixture only need to worry about applying 300kgf when your cells are discharged ri their minimum usable SOC - how high the clamping force may increase above that as the cells charge up is no longer a practical concern…

(Thanks EVE)

 

[Cheap 4-life]

And as I said in my first post on the matter, the pressure the cells I'd use would exert on another near EOL in a static fixture seem to be enough for them to destroy each other according to the data sheet. We got an unknown here, as fafrd put it, we don't know the increase in expanding force as the cells age. In the case of LF280K it doesn't seem to matter, but in the case of LF105 it does.

That is the only thing 2 things that still worries me about a rigid fixture. If that’s to much compression at high SOC and if the rigid fixture would cause to much compression of the cells when they age.
However it does seem like expansion should be prevented. With the cells moving so much from expansion contraction the busbars can become loose as several users are starting to realize, which causes the wrong cells to be balanced and inaccurate bms reading etc.. also I’m thinking expansion should be prevented to stop possible damage to the cells that happens from expansion.. so keeping the cells from expanding but not applying to much pressure at high SOC is what should be done. Springs allow to much expansion but the rigid fixture poron foam should all but stop expansion and prevent to much pressure at high SOC. Then people could just make sure to replace their batteries when they get to old or are over used so that extra aging expansion force doesn’t matter.

 

[Cheap 4-life]

The point is that cells allowed to expand under a clamping force of 300kgf will have a higher cycle life than cells allowed to expand without any pressure.

Correct compression needs to happen.. however with springs to much expansion is still allowed to happen. The 300kgf is compression at low SOC. As the cells go up in SOC the springs apply more pressure as the cells expand but the cells are still expanding because not enough pressure is appplied by the springs to stop the expansion.. @cinergi for example compressed his 16 cells with 660lbs of force and they expanded A lot. So much that his busbars were loosening causing all kinds of issues..

 

[Cheap 4-life]

how high the clamping force may increase above that as the cells charge up is no longer a practical concern…

I’m not onboard with this yet.. please explain how you are. Again they simply showed how they tested a cell. That doesn’t mean that optimal for prolonged health of our cells. Unless I’m missing something

 

[Cheap 4-life]

Again, knowing what the expansion force at HOL (half of life) sure would be helpful, but my general sense is that it’s going to be practically impossible to design a rigid fixture that remains rigid under 5000kgf(!).

My thinking is that the expansion force does not rise by much till near EOL, not a constant steady increase over its life. Maybe a slight barely noticeable increase thru it’s life and then more of a jump near the end..Or at least that’s what’s would make sense to me

 

[fafrd]

That is the only thing 2 things that still worries me about a rigid fixture. If that’s to much compression at high SOC and if the rigid fixture would cause to much compression of the cells when they age.
However it does seem like expansion should be prevented. With the cells moving so much from expansion contraction the busbars can become loose as several users are starting to realize, which causes the wrong cells to be balanced and inaccurate bms reading etc.. also I’m thinking expansion should be prevented to stop possible damage to the cells that happens from expansion.. so keeping the cells from expanding but not applying to much pressure at high SOC is what should be done. Springs allow to much expansion but the rigid fixture poron foam should all but stop expansion and prevent to much pressure at high SOC. Then people could just make sure to replace their batteries when they get to old or are over used so that extra aging expansion force doesn’t matter.

I understand the concern about stress on the busbars from cell expansion. I went to the trouble to manufacture 23 2/0 battery cables using welders wire I was so concerned about that.

Knowing what I know now (which is that there is NO expansion after cells have settled in under a 300kgf fixture), I would probably just stick to rigid busbars that I’d loosen and retighten at 50% SOC, highest SOC, 50% SOC and lowest SOC over the initial 2 or 3 charge - discharge cycles.

IfI had a suitable rigid container, I’d use that (padded as needed to press the cells together as tight as possible when discharged to lowest SOC) and if I was going the route of rigid endcaps secured using threaded rod, I’d either loosen and retighten to some reference torque over the same 2 or 3 discharge cycles or I’d throw on some calibrated springs and apply 300kgf after each of this 2-3 discharge cycles and call it a day.

At most, the issue is assuring that the full pack settles into position under 300kgf or more of force over the first 2-3 charge / discharge cycles without applying any excessive stress to the terminals (which may be connected to rigid busbars secured before the cell pack had fully settled in).

 

[Cheap 4-life]

Yes common sense to me previously said why not compress with springs instead of rigid, it can’t hurt and would avoid over compression at high SOC and rigid could most likely not avoid that.. now it’s seeming like it’s best to all but completely stop expansion with a rigid structure and the cells will not be over compressed at high SOC as per the data sheet if that’s actually what it says. I’d throw in some poron foam to further prevent over compression at high SOCZ

 

[Cheap 4-life]

Knowing what I know now (which is that there is NO expansion after cells have settled in under a 300kgf fixture), I would probably just stick to rigid busbars that I’d loosen and retighten at 50% SOC, highest SOC, 50% SOC and lowest SOC over the initial 2 or 3 charge - discharge cycles.

I’m not so sure this is always the case with all cells.. @cinergi 16cell packs and many others continue to expand through SOC using springs, not by crazy amounts but still do.. more cells the more expansion length of the pack under the same pressure spring,. Unless compression is stopped by higher pressure from stronger springs.. longer springs were used for larger packs to accommodate for the extra expansion

 

[fafrd]

Correct compression needs to happen.. however with springs to much expansion is still allowed to happen. The 300kgf is compression at low SOC. As the cells go up in SOC the springs apply more pressure as the cells expand but the cells are still expanding because not enough pressure is appplied by the springs to stop the expansion.. @cinergi for example compressed his 16 cells with 660lbs of force and they expanded A lot. So much that his busbars were loosening causing all kinds of issues..

I’ll have to go back and check up on Cinergi - he was the one who originally told me he got no compression after the first few cycles (which is what I’ve experienced).

If he has had to double his clamping force to prevent excessive expansion and stress in his rigid busbars as his cells have aged, I may need to keep an eye on my cells behavior.

On the one hand, because I’ve gone with flexible busbars, I’m certain I’ll have nothing to worry about in any case.

And on the other hand, should I ever see increased expansion or should new data emerge suggesting cycle life will greatly improve under higher clamping force of 600kgf (or whatever), it just means I’ll have to invest in new springs…

 

[Cheap 4-life]

Knowing what I know now (which is that there is NO expansion after cells have settled in under a 300kgf fixture), I would probably just stick to rigid busbars that I’d loosen and retighten at 50% SOC, highest SOC, 50% SOC and lowest SOC over the initial 2 or 3 charge - discharge cycles.

But as the cells age the expanding pressure is most likely slightly more.. that could continually loosen solid busbars. But as I said most likely the expansion pressure doesn’t go up much till near EOL

 

[Cheap 4-life]

If he has had to double his clamping force to prevent excessive expansion and stress in his rigid busbars as his cells have aged, I may need to keep an eye on my cells behavior.

So much so that he now uses the same flexible busbars I just purchased to help prevent bms issues from loosening busbars