Best practice for 300kgf ‘fixture’ 280Ah cells

[Tradewinds63]

After reading the EVE test data amongst the several LF model batteries and comparing it to this chart...

It's clear that the EVE experimental test subject was the EVE LF 304. If you compare the surface areas of the various batteries, only one, the LF 304 comes to approximately 660 lbs (300Kg) total force upon its surface (55.86 in²) when using the optimal 12 PSI as the clamping force/axial force applied to the battery.

So, should someone use the 300kgf for the LF105 (surface area 40.5 in²) as an example. They would be applying 16.3 psi, which is still in the peak zone but does reduce it by a projected potential 5k cycles, from nearly 20k cycles to about 15k cycles, when compared to 12 psi. An LF 105 would produce best results not with 300 kgf but rather 220 kgf.

It's too bad EVE didn't express their unit as 2.874 N/cm² Y axial force, which is 12 psi Y axial force. Then, they'd have avoided all the confusion with respect to the applied force and been far more accurate.

 

[upnorthandpersonal]

The fact that they use an outdated, non standard unit like the kgf has always puzzled me...

 

[Tradewinds63]

The fact that they use an outdated, non standard unit like the kgf has always puzzled me...

I'm going to guess they contracted the test out to some old geezer in China who provided them with that? What's sad is that they just slapped it into the literature for all their LF line-up, altering only the surface applied to data and battery model info. Not realizing it became inaccurate information when they applied it to batteries other than the LF304, solely because the unit expressed wasn't a standardized unit but rather a custom unit for the LF304, only.

 

[HRTKD]

Since we are discussing an applied mass force upon a surface area of the battery, the second unit of measure is typically expressed in either mm², cm² or m², etc. If it represented mm², the cells would be crushed if the force applied was equal to 300 kgf/mm² and the cells would also be crushed by 300 kgf/cm². it would be like you standing on a tin can and it buckling under your body wiegth.

While I haven't done the math, wouldn't 300 kg/m² be the more likely measure?

 

[Tradewinds63]

While I haven't done the math, wouldn't 300 kg/m² be the more likely measure?

That's what I thought also as mm² and cm² would be insane amounts of pressure. But the literature expresses within a single snippet mention, the area in which the force is applied to, referencing the plane perpendicular to the Y axis and providing that areas dimensions. So it isn't applied to any standardized unit, making it non-transferable upon batteries with different surface areas. Yet, they do just that throughout the many manufacturer specification docs.

Even companies such as CATL references the same experimental results using the informal non-standardized 300kgf for their lifepo4 prismatic batteries axial compression force. Which if used will result in fairly different results among the various battery models, in some cases, failure. For example the LF32ah with effectivly a 148x91 mm surface area or 20.875 in² almost half the LF105 and a mere 37% the surface area of the LF304. Providing a total 300kgf axial clamping force to the 20.875 in² surface would result in 14.37 k/in² or 31.68 psi! That would eventually likely result in cell damage as they say 9 kN will cause internal failure. On the LF304 that 300kgf comes to about 3 kN on the planar surface which on the LF32 is 2.7 times the force of 3 kN, at nearly 9 kN or more presicely 8.1kN and the literature clearly states that 9kN force will cause battery failure. So, 8.1kN force through repetitive charging would very likely eventually cause internal failure also.

What I don't like, I had to read through the literature of several models to realize what they had done, copy, paste, edit the same experimental literature. Then it took a while to determine which model was most likely used in the referenced experimental setup (LF304). Between that and the other independent experimental test results it became clear what EVE and others have done with respect to that single experiment they did. They dropped the ball and didn't convert the measurment to a standardized unit.

Unfortunately, that's going to result in many improper applications of the "fixture." It's pretty clear they don't care. LOL, Otherwise they'd issue a standardized unit update to clear the air on the matter.

So, after an exhausting time looking though many documents to flesh out what is going on. I recant some parts of my initial and second post on this topic.
Keeping the batteries in the fixture is paramount BUT employing the correct 12psi is also paramount to achieving the projected maximum potential cycle life. Forget the damn 300kgf, it's only pertinent to one battery models planar surface.

Next is understanding the forces at work within the batteries and how their expansion can and will effect improperly designed "fixtures."
I took the time to run the structural force calculations for a battery bank I'm building now and the materials I'd have to use to make it fit in the cabinet it will be placed in. I picked up 8" steel C channel with 1/4" webbing and 2" flanges. Will use 1/4" 20 tpi steel allthread to provide the axial clamping force. Some seriously expensive material to clamp this group without any chance for expansion.

This bank will be 5 parallel 8 series batteries, 40 batteries in one "fixture" by two fixtures total for an 80 battery bank of LF105, 51.2v 26,880 wh.

5 ea LF105 planar surfaces parallel within a single fixture comes to 5 x 40.5 in² = 202.5 in² by 12 psi = 2430 lbs axial force total. Using 4ea 1/4" 20 tpi steel allthread to provide the axial force, that's 607.5 lbs each axial force using 30.375 in-lbs torque each or 2.5313 ft-lbs each.

I've got 80 batteries to drain down between 30-40 SOC, they came with 3.293v each +/- .001v. Pretty impressive.

 

[upnorthandpersonal]

most likely used in the referenced experimental setup (LF304)

the 300kgf was already present in the literature before the LF304 existed though - at least as early as the original LF280, but since the surface area of those two is pretty much identical, it's valid for both.

 

[HRTKD]

the 300kgf was already present in the literature before the LF304 existed though - at least as early as the original LF280, but since the surface area of those two is pretty much identical, it's valid for both.

That's what I was thinking too.

 

[S Davis]

That's what I was thinking too.

Yep that is what was being used for the LF280N, mine are just under 12 psi. at 640 lbs total across the face.

 

[fafrd]

I was responding to post 242 talking about crushing the cells.

My apologies - my post was directed to post 242 by Tradewinds62 - I was surprised any of us were taking that concern of crushing Eve cells in a 300kgf clamping fixture seriously (at least any of us that have actually built one ).

 

[fafrd]

That's what I thought also as mm² and cm² would be insane amounts of pressure. But the literature expresses within a single snippet mention, the area in which the force is applied to, referencing the plane perpendicular to the Y axis and providing that areas dimensions. So it isn't applied to any standardized unit, making it non-transferable upon batteries with different surface areas. Yet, they do just that throughout the many manufacturer specification docs.

Even companies such as CATL references the same experimental results using the informal non-standardized 300kgf for their lifepo4 prismatic batteries axial compression force. Which if used will result in fairly different results among the various battery models, in some cases, failure. For example the LF32ah with effectivly a 148x91 mm surface area or 20.875 in² almost half the LF105 and a mere 37% the surface area of the LF304. Providing a total 300kgf axial clamping force to the 20.875 in² surface would result in 14.37 k/in² or 31.68 psi! That would eventually likely result in cell damage as they say 9 kN will cause internal failure. On the LF304 that 300kgf comes to about 3 kN on the planar surface which on the LF32 is 2.7 times the force of 3 kN, at nearly 9 kN or more presicely 8.1kN and the literature clearly states that 9kN force will cause battery failure. So, 8.1kN force through repetitive charging would very likely eventually cause internal failure also.

What I don't like, I had to read through the literature of several models to realize what they had done, copy, paste, edit the same experimental literature. Then it took a while to determine which model was most likely used in the referenced experimental setup (LF304). Between that and the other independent experimental test results it became clear what EVE and others have done with respect to that single experiment they did. They dropped the ball and didn't convert the measurment to a standardized unit.

Unfortunately, that's going to result in many improper applications of the "fixture." It's pretty clear they don't care. LOL, Otherwise they'd issue a standardized unit update to clear the air on the matter.

So, after an exhausting time looking though many documents to flesh out what is going on. I recant some parts of my initial and second post on this topic.
Keeping the batteries in the fixture is paramount BUT employing the correct 12psi is also paramount to achieving the projected maximum potential cycle life. Forget the damn 300kgf, it's only pertinent to one battery models planar surface.

Next is understanding the forces at work within the batteries and how their expansion can and will effect improperly designed "fixtures."
I took the time to run the structural force calculations for a battery bank I'm building now and the materials I'd have to use to make it fit in the cabinet it will be placed in. I picked up 8" steel C channel with 1/4" webbing and 2" flanges. Will use 1/4" 20 tpi steel allthread to provide the axial clamping force. Some seriously expensive material to clamp this group without any chance for expansion.

This bank will be 5 parallel 8 series batteries, 40 batteries in one "fixture" by two fixtures total for an 80 battery bank of LF105, 51.2v 26,880 wh.

5 ea LF105 planar surfaces parallel within a single fixture comes to 5 x 40.5 in² = 202.5 in² by 12 psi = 2430 lbs axial force total. Using 4ea 1/4" 20 tpi steel allthread to provide the axial force, that's 607.5 lbs each axial force using 30.375 in-lbs torque each or 2.5313 ft-lbs each.

I've got 80 batteries to drain down between 30-40 SOC, they came with 3.293v each +/- .001v. Pretty impressive.

I appreciate your ‘ recant’ on portions of your earlier post.

For those of us with Eve 280Ah cells, if you read through this thread from the beginning, you’ll see that 12psi is what we’ve all bee aiming for (at 50% SOC - the thread also includes limits for minimum pressure @ ~10% SOC and maximum pressure @ ~99% SOC important to stay above/below (especially the staying under the max pressure) and recommended by Eve through back channels.

 

[Tradewinds63]

I've finally had the opportunity to begin the assembly of our 26,880 wh bank.
Eve LF105 80 batteries total. 5 parallel 16 series banks. 51.2 vdc nominal.
8 cells deep by 5 wide, per splint group.

Opted to clamp them at around 7.5psi for approximately 15k cycles. About 370lbs compression force per allthread from 30 in-lbs torque force, tested separately in a jig using a digital crane scale and digital torque wrench.

The splints are 8" x1/4 thick steel C channel with 1.5" flanges, and very heavy.
1/4" allthread with heatshrink insulators. 1/4" polycarbonate insulation sheet between C channel and batteries.
Also used plastic and wood composite decking boards planed down and epoxied to the C channel for the BMS and Bus mounting.

C bank is to the far right top and bottom, split.
Bank layouts are:
A-A B-B C
--------------|
D-D E-E C

The metal cabinet is equipped with thermal controlled fan for cooling with outdoor air and a smoke detector to kill the fan power should it sense smoke in the cabinet. This cabinet assembly is contained within a 20' shipping container.

Will share final pics once the BMS's and everything else is wired in.

 

[Skypower]

I've finally had the opportunity to begin the assembly of our 26,880 wh bank.
Eve LF105 80 batteries total. 5 parallel 16 series banks. 51.2 vdc nominal.
8 cells deep by 5 wide, per splint group.

Opted to clamp them at around 7.5psi for approximately 15k cycles. About 370lbs compression force per allthread from 30 in-lbs torque force, tested separately in a jig using a digital crane scale and digital torque wrench.

The splints are 8" x1/4 thick steel C channel with 1.5" flanges, and very heavy.
1/4" allthread with heatshrink insulators. 1/4" polycarbonate insulation sheet between C channel and batteries.
Also used plastic and wood composite decking boards planed down and epoxied to the C channel for the BMS and Bus mounting.

C bank is to the far right top and bottom, split.
Bank layouts are:
A-A B-B C
--------------|
D-D E-E C

The metal cabinet is equipped with thermal controlled fan for cooling with outdoor air and a smoke detector to kill the fan power should it sense smoke in the cabinet. This cabinet assembly is contained within a 20' shipping container.

Will share final pics once the BMS's and everything else is wired in.
View attachment 239506View attachment 239507View attachment 239508View attachment 239509

Looks good. Do you have insulators between cells?

 

[rhino]

Just a comment that the greatest force for expansion is going to be near middle of cells and not at top and bottom so I wonder if what you have there is going to bulge outwards because there is no support in the center.

 

[Skypower]

Just a comment that the greatest force for expansion is going to be near middle of cells and not at top and bottom so I wonder if what you have there is going to bulge outwards because there is no support in the center.

I was wondering the same thing, but zooming in, I could see the texture and feel certain that that’s a silver painted steel channel. There’s a pretty good span and it might bow out a very small amount, however it’s a heck of a lot stronger than any sever rack box. I’ve seen cell case wrap split so as long as there’s some Formica or fiberglass between all cell interactions I’m pretty sure that it’ll work fine.

 

[damo green]

That's what I thought also as mm² and cm² would be insane amounts of pressure. But the literature expresses within a single snippet mention, the area in which the force is applied to, referencing the plane perpendicular to the Y axis and providing that areas dimensions. So it isn't applied to any standardized unit, making it non-transferable upon batteries with different surface areas. Yet, they do just that throughout the many manufacturer specification docs.

Even companies such as CATL references the same experimental results using the informal non-standardized 300kgf for their lifepo4 prismatic batteries axial compression force. Which if used will result in fairly different results among the various battery models, in some cases, failure. For example the LF32ah with effectivly a 148x91 mm surface area or 20.875 in² almost half the LF105 and a mere 37% the surface area of the LF304. Providing a total 300kgf axial clamping force to the 20.875 in² surface would result in 14.37 k/in² or 31.68 psi! That would eventually likely result in cell damage as they say 9 kN will cause internal failure. On the LF304 that 300kgf comes to about 3 kN on the planar surface which on the LF32 is 2.7 times the force of 3 kN, at nearly 9 kN or more presicely 8.1kN and the literature clearly states that 9kN force will cause battery failure. So, 8.1kN force through repetitive charging would very likely eventually cause internal failure also.

What I don't like, I had to read through the literature of several models to realize what they had done, copy, paste, edit the same experimental literature. Then it took a while to determine which model was most likely used in the referenced experimental setup (LF304). Between that and the other independent experimental test results it became clear what EVE and others have done with respect to that single experiment they did. They dropped the ball and didn't convert the measurment to a standardized unit.

Unfortunately, that's going to result in many improper applications of the "fixture." It's pretty clear they don't care. LOL, Otherwise they'd issue a standardized unit update to clear the air on the matter.

So, after an exhausting time looking though many documents to flesh out what is going on. I recant some parts of my initial and second post on this topic.
Keeping the batteries in the fixture is paramount BUT employing the correct 12psi is also paramount to achieving the projected maximum potential cycle life. Forget the damn 300kgf, it's only pertinent to one battery models planar surface.

Next is understanding the forces at work within the batteries and how their expansion can and will effect improperly designed "fixtures."
I took the time to run the structural force calculations for a battery bank I'm building now and the materials I'd have to use to make it fit in the cabinet it will be placed in. I picked up 8" steel C channel with 1/4" webbing and 2" flanges. Will use 1/4" 20 tpi steel allthread to provide the axial clamping force. Some seriously expensive material to clamp this group without any chance for expansion.

This bank will be 5 parallel 8 series batteries, 40 batteries in one "fixture" by two fixtures total for an 80 battery bank of LF105, 51.2v 26,880 wh.

5 ea LF105 planar surfaces parallel within a single fixture comes to 5 x 40.5 in² = 202.5 in² by 12 psi = 2430 lbs axial force total. Using 4ea 1/4" 20 tpi steel allthread to provide the axial force, that's 607.5 lbs each axial force using 30.375 in-lbs torque each or 2.5313 ft-lbs each.

I've got 80 batteries to drain down between 30-40 SOC, they came with 3.293v each +/- .001v. Pretty impressive.

Thanks for updating your stance as you saw fit; ". I recant some parts of my initial and second post on this topic." ..I've been curious where your original thought that they only required clamping during the initial cycles to "de-gas" ... was it from a manufacturer or ...from where?

 

[Skypower]

Thanks for updating your stance as you saw fit; ". I recant some parts of my initial and second post on this topic." ..I've been curious where your original thought that they only required clamping during the initial cycles to "de-gas" ... was it from a manufacturer or ...from where?

Yeah, 300 number only applies to the 270 to 320 ah size cells. My personal opinion is that fixed- none spring loaded should be from firm contact to 8 psi and spring or foam loaded 10 to 12 psi. Cells will keep expanding after the initial charge. So un clamping would be counterproductive imho. On a hot day and full cycle they really want to expand. The end game here is that the layers of the “jelly roll” are kept from rubbing past each other as the cells cycle or straining.

 

[damo green]

Yeah, 300 number only applies to the 270 to 320 ah size cells. My personal opinion is that fixed- none spring loaded should be from firm contact to 8 psi and spring or foam loaded 10 to 12 psi. Cells will keep expanding after the initial charge. So un clamping would be counterproductive imho. On a hot day and full cycle they really want to expand. The end game here is that the layers of the “jelly roll” are kept from rubbing past each other as the cells cycle or straining.

My question was to and referencing Tradewinds63 original thought, not anything else. To clarify..NP you probably didn't realize what I was referencing.

 

[Brucey]

The metal cabinet is equipped with thermal controlled fan for cooling with outdoor air and a smoke detector to kill the fan power should it sense smoke in the cabinet. This cabinet assembly is contained within a 20' shipping container.

It may be better to have the smoke detector set to turn on an exhaust fan to vent out possible venting hydrogen.

 

[Tradewinds63]

Just a comment that the greatest force for expansion is going to be near middle of cells and not at top and bottom so I wonder if what you have there is going to bulge outwards because there is no support in the center.

The splints are structural grade 8 inch C channel steel 1/4" thick and 2.26"(corrected) flanged. There's zero chance that something that is countered by 12 psi force is going to cup the web of such a structural C channel beam. In other words, there's ample center support across the entire member. Each splint weight about 27 lbs each.

 

[Recovering EE]

It may be better to have the smoke detector set to turn on an exhaust fan to vent out possible venting hydrogen.

A BRUSHLESS exhaust fan!

And no spark-generating relay contacts nearby, either.

 

[Tradewinds63]

It may be better to have the smoke detector set to turn on an exhaust fan to vent out possible venting hydrogen.

This isn't an exhaust fan. It pushes cool air into the battery box for cooling purposes and that air vents out of the battery cabinet and into the shipping container. The reason for turning off the fan when smoke is detected is to keep from feeding a fire from the plastics of the battery, oxygen.

The cabinet is not in an occupied space.

 

[Tradewinds63]

Looks good. Do you have insulators between cells?

The insulation between the batteries is what comes standard, already on the battery.
The case isn't wired to either terminal but the case is polarized between itself and the cathode terminal with the cathode terminal and case demonstrating a diode effect of a potential of the cathode terminal acting as positive cell voltage to the case, if connected externally. That effect is just the limited potential current between the case and the most external negative plates.
In other words, if the cases contact one another, there should be no short effect as the current flow is direction restricted from the case to the negative terminal internally only if they are shorted extarnally only. It's not as if the negative terminal and case are one in the same, they share a polarized potential connection only.

 

[upnorthandpersonal]

It's not as if the negative terminal and case are one in the same

There are some rumors (I've not seen it myself) that certain cells have the cell positive directly tied to the case. I will believe it when I see it, but it's a story that has been doing the rounds... It's also the reason why many DIY battery boxes ship with epoxy sheet cell separators.

 

[Tradewinds63]

There are some rumors (I've not seen it myself) that certain cells have the cell positive directly tied to the case. I will believe it when I see it, but it's a story that has been doing the rounds... It's also the reason why many DIY battery boxes ship with epoxy sheet cell separators.

I tested my cells before settling on the construction method and after finding only a naturally occuring chemically produced polarity between the Cathode terminal and the case, I wasn't convinced additional insulators were needed. I was actually considering the removal of the blue insulator film that comes with them and tying all the cases together, in effect, puposely grounding them to one another.

 

[Tradewinds63]

With age for some comes a reduction in attention to detail. At 60, my vision isn't as good as it used to be and my hands aren't as steady as they once were. My patience is also diminishing, LOL.
So I made another mistake when wiring up these batteries to the BMS. After trying to make sense of the thumb sized wiring schematic of the BMS balancers and reading the instructions that more resembled an alternative version of L Carols Jabberwocky. I thought I had the wiring pattern down correctly. Unfortunately I originally wired the sensor wires in reverse order and burned out 3 units number #1 sensor.
Fortunately my background in electronics, etc made it possible for me to diagnose and repair the resulting burned out resistors after challenging my microscopic electronics rework abilities.

Here's the final mess with the doors off and not yet plugged in or tied to the main system.
Don't get me started about my opinion on 100 Balance. Wish I had exercised more patience and read others opinions on the best BMS out there. Will have to make due with these, for now.
I posted a video on Youtube for the sensor repair procedure. If anyone wants the link, let me know and I'll see if I can import it over to here.
That's the tip of a bic ballpoint pen pointing at the resistor. Tiny stuff these days.