Eve 280ah cell compression?

[alexander ross]

Hi, I am trying to understand the numbers in the datasheet. 

Do they mean 300kgf or 20kgf? I believe +/- refers to tolerance but i am not sure which number it refers too. Converted to psi, i get 284psi for 20kgf. Which sounds like a lot! 

Trying to work out what springs to buy in the uk or from aliexpress? 

Thank you for your time. 

[Dacian Todea (electrodacus)]

Yes that is correct and it is 300kg of force with some tolerance (even more than 20kg tolerance is not a problem as that tolerance is what they had for measuring those dimensions).

You can see how I compressed my cells here https://groups.google.com/g/electrodacus/c/yN3i-jHNEcQ/m/0gFujrU5AgAJ

Here is a link to the spring I used https://www.amazon.ca/gp/product/B07Z3D4J6L

[Dacian Todea (electrodacus)]

I think you made the conversion to psi wrong as those 300kg (3000N) are for the entire surface of the cell witch is way larger than 1 square inch. But get those springs I recommended and leave them just slightly uncompressed so I mean spring is not fully compressed as then you have no idea how much force you apply. 

204.6 * 173.7 mm = 35539mm^2 = 0.0355m^2  that is according to google 55 square inch

1kg is about 2.2lb so 300kg is 661lb and 661lb / 55 square inch = 12psi unless I made some conversion wrong.

[alexander ross]

Thank you. I see i forgot psi is imperial and should have converted metric weight and dimensions into imperial. Silly me! 

Thank you very much for the explanation! 

Now trying to find equivalent springs in the uk or aliexpress or Amazon US if need be for 8M threaded rods I have already cut and used. Hmm converting max newton or weight+dimension of spring to get a psi number… hmm hmm

Any keywords or suggestions for maths books, sections,webpages  i should ready is welcome! 

[alexander ross]

I found these springs that would fit a M8 rod but they are  stronger I believe.

sourcing map 3D Printer Die Spring, 5pcs 18mm OD 20mm Long Spiral Stamping Light Load Compression Mould Die Springs for 3D Printer Electric Part, Blue https://amzn.eu/d/d5tx7uh

[Dacian Todea (electrodacus)]

The springs are color coded and brown (the ones I used are the hardest to compress).

I used 4 of those springs each with a fully compressed force of 736N so about 75kg thus a total of 300kg for 4 of them.

The springs you looked at are only 422N so about 42kg thus you will need to use more than 4 maybe at least 6 but if you only need 4 then check for a different color something that is around 700 to 800N

[alexander ross]

Thank you! Was wonder what the colour referred too. Thought it was wrongly inner dimension.

Found what looks like the right brown size for M8 rod on aliexpress. Which also explains the colours in the description! 

1/2/3Pcs Yellow/Blue/Red/Green/Brown   Spiral Stamping Compression Mould Die Spring OD 18mm ID 9mm Length 20-150mm
https://a.aliexpress.com/_EQKSDFT

I am assuming same 20mm length as yours?

 I’ll check and learn more tomorrow, if bigger inner diameter means different length required but if the length can be confirmed for 18mm OD, that would be appreciated. 

Great gratitude to you.

[Dacian Todea (electrodacus)]

Since you intend to use a thicker spring you will not want brown as that will be to strong. But blue at just 422N is to weak so you want some other color

The length is irrelevant. The force needed to fully compress the spring is what is important so if you have a 20mm spring or two 20mm springs in series same force will be needed same with a single 40mm spring.

I found here that red is 637N so close but not quite there so likely you need green

I found for green also https://vi.aliexpress.com/item/1005003226761222.html  but this is already a bit to strong at 951N fully compressed and you want 700 to 800N so in the middle of reed and green.

The green 8mm inner diameter will be great but maybe to close to the limit and it will not fit or maybe the 10mm inner diameter red witch is 784N https://vi.aliexpress.com/item/1005003225916611.html

You will just make sure is not fully compressed so there is at least space to insert a piece of paper of two between the coils. A longer spring will apply the same force but you have more control when getting close to full compression.

[Charlie G]

Hey gentleman,

I've been Dacians build for some concepts, and while his setup is indoors in the northern environment, mine is in an insulated camper van further north than him being exposed to minus 45 below.. 

I have a small coolant tank as the heat exchanger underneath the 24 volt 900 amp hour pack.

Previously I had a 300 amp hour pack constructed of 150ah cells, but the only compression I had was screws holding pressure treated lumber together tightly, as well, no heat. That's $1500 pack only made it to Winters for a couple other catastrophic reasons also including myself.

Here's some suggestions.  

Springs.- sourcing springs from China seems pretty far away and I'm sure the selection is pretty limited.. should you need another one?, You only have to wait a month or two.. 

There's lots of spring manufacturers in the US and I can't remember the name but I think I deal with Jones spring. But whatever there's so many it doesn't matter, and they literally have a thousand different sizes on hand for like a dollar or two a piece. Need something custom, they can do it. 

Also for compression springs, best practice is not to exceed 65 to 70% of its compression. There's a little more to it than that, but Google's search that or ask the spring builder. 

I never research or even sought any clarity on Eve's Datasheet. I interpret the data sheet as if that load factor in KFG needs to be applied per a single cell. So when you have multiple cells, your amount of force is going to be significantly more then spec during expansion events of the cells. Again, back to the 70% and I arbitrarily guessed The 12 springs pressing the three cells could make up for that extra force. 

Here's a picture of my build. It's going off to be laser cut and CNC bent next week. 

[Flint Smith]

Off-Grid Garage (youtube) got his hands on a cycle-life vs. compression graph.  Yes, max was at 660 lbs, but it was a very gradual change after an initial rapid rise.  His interpretation was that compression helps a lot, but any compression will do.

Also, the compression must be applied to NEW batteries.   The cycle-life extension is only seen if the compression is applied during the first few cycles.

I bought these springs from Grainger.  I bought a bunch as they were in deep clearance.  I'll send you a set(4) for the cost of shipping (from California). You can see springs on the website, but they're mostly gone.  They won't ship fewer than their minimum even if they have some.   3/8 (10mm) shaft diameter, but you could maybe use a sleeve?

They're 1 1/4 inch long and apply 147 pounds if you compress them 30% (0.31 inches), which is their intermediate duty rating. You can go higher.  And, they have a beautiful blue powder coating!

Manufacturer's page(Imperial)

Manufacturer's page(Metric)

[Dacian Todea (electrodacus)]

Charlie,

The 300kgf is the same for 1 cell or multiple cells (like in your case 3 cells). Each of those 3 cells will experience the same force/pressure.

Any compression up to 300kgf will be better than no compression but is best to have close to the recommended compression.

Cells expand and contract with SOC but that is not much.  During charge and discharge there will be some gas formation but that gas will not separate the anode and cathode if the cells are compressed as per spec. So cells without compression will expand more and that separation is what contribute to increase degradation rate.

Off-grid Garage considers that the extra cost and trouble of compressing the cells is more than he wants to deal with and he will prefer to buy a new pack in a few years that deal with compression.

 

Regarding spec for those springs you got at Grainger (they are very confusing). Compared to the spec here for a similar item https://vi.aliexpress.com/item/1005003222854836.html

The "Spring Rate 387.7 lb/in" makes absolutely no sense.

Where did you got the 147lb at 30% compression ?  They are 1.25" long so how is 0.31" 30% ?

Say max compression length is the same as for the aliexpress spring so 40%

Then the most that 1.25" spring can compress is 0.5" (distance it travels so spring will be 0.75" long when fully compressed).

Then I can maybe interpret that 387.7 lb/in as 0.5 * 387.7 lb/in = 193.85 lbf for max compression and that will be about 88kg much more that the equivalent sized one on aliexpress so not sure that is correct.

Will love to know if anyone understand what this spring rate in lb/in actually means and how it is used.

[alexander ross]

Thanks muchly. I have ordered the red spring and will try to not overcompress them! 

https://vi.aliexpress.com/item/1005003225916611.html

[Flint Smith]

Yes, you're right.  Sorry.  I copied the wrong number.   I don't know how that happened as I was looking right at the bag.

Optimal life 25% (0.31in) 122lb
Long life 30% (0.38in) 147lb

Maximum operating deflection 40% (0.50in) 196lb  

  

Spring rate is used as a ratio as you say. 

660lb/4springs = 165lb/spring

165lb/388lbin⁻¹= 0.43in compression

1.25in full length - 0.43in compression = 0.82in compressed length

[Dacian Todea (electrodacus)]

It makes no sense to use such a unit 388lb/in as that will be different for same spring with different length.

So if you get a 2.5" long spring same as having two of those 1.25" springs in series then spec will no longer be 388lb/inch but half that.

Since the blue spring you got seems to be very similar to the ones on Aliexpress the variation in characteristics is very large making me doubt that spec is correct.

The Aliexpress one also compresses 40% and the one with 10mm internal diameter it has 54kgf that is way less than 196lbf (89kgf). The material used will need to be way more different as dimensions are similar in fact diameter interior and exterior is slightly larger for the one on aliexpress.

And then you have this https://www.grainger.com/product/RAYMOND-Die-Spring-Medium-Duty-44V063

Basically the same spring just 1" long instead of 1.25"

At full compression should be the same but 519.6 * 0.4" = 207.84lbf  vs 388 * 0.5" = 194lbf

Then this 2" long one that will deflect max 0.8"  https://www.grainger.com/product/RAYMOND-Die-Spring-Medium-Duty-44V067

227.8 * 0.8" = 182lbf already 10% error from the 1" long.

In any case it will seem more logical to use something like 196lbf at max compression rather that having to make calculations from the 388lb/in  

I made a search for similar size spring and the closes I found is this https://www.norelem.com/us/en/Products/Product-overview/Systems-and-components-for-machine-and-plant-construction/26000-Compression-springs-Elastomer-springs-Rubber-buffers-Shock-absorbers-Gas-springs/Springs/26001-Compression-springs-ISO-10243-moderate-load.html

The closest will be the 20mm external 10mm internal and that will be 820N but that compresses 38% or so not 40% so also thicker material on top of being a bit larger.

In any case it was just surprising to see lb/in in the spec. To me that value is more confusing than helpful.

[Don Fukushima]

Just curious about the topic of compression. Most if not all commercial battery packs using LiFePO4 cells have a simple case that houses the cells and I'm struggling remembering any that had anything more than some packing between the cells and the housing. No springs or other variable compression. Is this new or getting to that 80/20  truism about getting 80 percent of benefit for 20 percent of the cost or the lat 20% benefit costs another 80%?

[Dacian Todea (electrodacus)]

The springs are used in DIY and professional manufacturers of packs will use some specialty spongy material between cells and design the case so that is rigid enough to provide enough compression together with that thicker elastic material placed between cells. Most LiFePO4 12V/24V batteries are not properly done but pack used in EV's and other large scale applications are done correctly.

Here is an example of acrylic foam compression pads with quite a bit of details  https://www.iwatani.co.jp/eng/business/material/electronic-materials/products/pouch-type/

And here is another large manufacturer that has Silicone and polyurethane based compression pads https://www.tapesolutions.saint-gobain.com/na/markets/automotive/ev/compression-pads#

And here is a manufacturer that you may recognize offering a similar solution https://www.3m.com/3M/en_US/oem-tier-us/applications/propulsion/ev-battery/ear-cushioning-materials/

[Don Fukushima]

Agree that some form of pliable material is used between cell and case. Seems much easier and cleaner to do it this way. And pack is better enclosed so I guess the springs are a solution where there will be no enclosure.thanks

[Dacian Todea (electrodacus)]

On most of those 12V LiFePO4 there is no compression pads just some isolation pads and most of them are just tide up with some sort of adhesive tape that will not offer any compression. Also they use generic plastic cases that are not specifically designed for the cells they used. In case of commercial available EV's all of them have custom designed cases and proper compression pads unless they se cylindrical cells where compression is not needed.

So for DIY is not about the enclosure is about the fact that to my knowledge compression pads are not available in low volume for DIY use thus the use of spring is lower cost and more convenient. Springs are also used in labs for cell testing as they offer a more linear so more predictable compression than pads.

https://i.ytimg.com/vi/gIDtX0Yptws/maxresdefault.jpg example of what will look like a properly designed 12V battery but it is not as it only uses plastic spacers and no compression pads.

I doubt that there is any 12V or 24V LiFePO4 that is properly designed since those are mostly made with random cells that they can get the hands on not new cells from manufacturer and they need to be low cost to maximize profit thus investing in proper engineering the cells will be to expensive and they will not be able to compete. There are those that use cylindrical cells and there there could be other issues but the compression is not needed there.

[Don Fukushima]

That makes a lot of sense and good explanation of the difference between 100% engineering vs mid-level battery assemblers vs DIY. As always there is more to learn about all this stuff. I'll have to go back and read the net benefit of a properly designed battery pack vs cost/time/energy to do it. lol

[Flint Smith]

 Thank you Dacian for your support and guidance.  Having someone who knows what's up say "those numbers don't sound right" is immensely valuable. 

I compressed a stack of 4 springs by standing on them.  My method was error-prone and would easily under-report the value. The spec sheet (spring constant calculation) says they were exerting 127 pounds but I weigh 138.  Off, but not by a scary amount.

I see the advantage of just using the MAX loading, but that puts you in the conditions where springs break....or so the spec sheet implies.  In real knowledge, I have no idea about how springs decide to break.  The advantage of the spring constant is that it allows a builder to adjust the force in a predictable way.  As you say, it's less helpful in selecting the spring initially.

The specification sheet you linked above is very interesting and shows the same things you noted in the Raymond/Grainger springs.  Short springs are definitely stiffer than long (bulk) spring...by 22%.  My guess is that the cut ends are much stiffer, having been hardened by cutting/grinding so the spring acts ~2.5mm shorter at each end, 20% of a 25mm spring.  I hoped to figure out what is going on and started a spreadsheet to do some curve-fitting, but I gave up.

Speaking of learning things, I took apart my spring tester and galled the 3/8" stainless threaded rod!  Sure, stainless is soft, but come on.  138 pounds of tension and the nut locked to the shaft within 1/2 an inch.  I guess I'll switch to steel because I can't see myself using a Sawzall to get the cells out.  

Flint

On Wednesday, June 7, 2023 at 2:22:05 PM UTC-7 electr...@gmail.com wrote:

[Dacian Todea (electrodacus)]

You do not need to be that precise with compression force as it will work with a fairly large range as you can see in the graph below that is taken from the A123 System spec sheet.

The springs I used are around 300kgf (all 4 of them together 75kg each) that is around  12psi as calculated in one of the earlier comments to get an idea of where in the graph that is since the same 12psi is what EVE also recommends.

Now 12psi will be if I will have fully compressed the springs but that was not the case. Even If I only compress them to half since force is linear it means 6psi and that will still be a huge advantage compared to no compression and not far from ideal 12psi.

I did compressed mine more than half maybe around 8 or 9psi so is still close enough to ideal that it will make no difference. It is the same for my A123 System battery and that is already 5 or 6 years old and in excellent condition. For that cell there is proper spec as seen by the graph below and the difference is huge with 3000 cycles no compression vs 20000 cycles with ideal compression so even with 6psi you still have 15000 cycles that is 5x better than no compression at all.

 

[Don Fukushima]

Excellent reference and shows how dramatic a bit of pressure can triple the cycles. I'm convinced to ensure the new cells are enclosed such that there is 5-10 psi. It also shows how variable pressure via springs is an easy way to set and maintain constant pressure.