diy solar

diy solar

Best practice for 300kgf ‘fixture’ 280Ah cells

It would be really awesome if the thread called "best practice of clamping force" actually answered the question.

Remind me never to paint a shed with you all.....

"using 4 standard unlubricated 1/4” bolts, what level of torque (inch pounds) should be applied to get about 12psi of clamping force (and about 142% that much torque for 17psi, correct?)?"

I think I did answer it:


What I said was, "DIY!"
Which is what this forum is all about, after all.

I also presented an example of having calculated clamping force, and compared it to strength of a repair technique I had recommended.



"Give a man a fish ..."
 
Dzi,

I have been looking for a flexible battery bus bar solution and you have provided some great ideas.

The 71mm CALB bus bars made for the CALB CA180 cells look like the best solution to me and they appear to be able fit the EVE 280 cells (holes are a bit oversized however for M6). However, no one seems to sell the CALB bus bars separately from the CALB cells. Anyone know where you can get these?
I have noticed the same thing. The sellers I have seen do not sell them separately from the cells. A starting point for your search could be the list of authorized sellers on the CALB USA website (if you are located in the US)
 
Rigid busbars themselves represent a counteracting force against shrinkage / movement at the top surface of the cells (which is exactly the stress we are stressing out about), so if rigid busbars are re-seated when the cells are full, the full pack may retain ‘full’ dimensions at the top of the e pack and shrink primarily along the lower ends (like an accordion).

This is exactly what happened to me: https://diysolarforum.com/threads/c...lar-10-kw-inverter-rv-build.13786/post-195598
More to come in that thread ...
 
  • Like
Reactions: Dzl
Posted this in the wrong thread. Reposting this here where it belongs:

Got results from my research to report.

I found a stackup of McMaster Carr spring washers that will do what I think is needed.

My criteria is to limit force at 100% SOC to 15 PSI. I have read that these cells expand between 0.7mm to 0.8mm from 0% to 100% SOC. I am working with 4 cell stacks in each compression frame. This totals up to 3.5mm of travel so that became my goal.

I am using 5/16" x 18 hardware and 18 TPI = 0.058 "/turn:

0.058 "/turn x 25.4 mm/" x 2.5 turns = 3.683 mm of travel.

I ordered and tested a number of different spring washers from McMaster Carr, and one the worked for me is this one.


What I found is that a stack of 12 of these disc springs in series <><><><><><> resulted in a force of 100 kg when compressed with 2.5 turns of the nut. This is starting with the nut finger tight (close to 0 kg). I then turned the nut 2.5 turns and my load cell read ~ 100 kg (+/-5 kg).

I am going to declare victory and that is how I am going to build my compression cells. I am going to keep working to see if I can refine these results, but I think I am close enough to move to building compression cells. When I finish my compression cells I will test a set of 4 cells in compression with the load cell on one of the threaded rods and measure the peak force at 100% SOC. I will determine what the correct preload is then.

Here is a pic of my test setup.

PXL_20210120_022059386.jpg

And I finally found a use for that old welding coupon. Makes a nice base plate for a load cell.

F.Y.I. Avoid stainless steel hardware like the plague. I started testing with a stainless steel bolt and flange nut and after a half dozen cycles the nut galled on the bolt so bad I had to cut if off my fixture. I am really glad I have a band saw.


PXL_20210120_010350337.jpg

P.S. Forget using a torque wrench to set compression. I tried that with my load cell and saw tremendous non-repeatability (+/-50%) when trying to use a torque wrench to set pressure. Springs are very repeatable. I am getting +/-5% variation from one test cycle to the next.

Using the torque wrench with lubrication on the threads might be more consistent, but don't forget these cells are going to expand as you charge them. Even if you just run the nuts in finger tight at 0% SOC, I predict you will be applying way too much compression at 100% SOC unless you have springs to provide compliance.
 
Last edited:
Stainless is best when used with a light lubricant which mostly stops galling. Also using two different alloys of stainless (one for the nut the other for the bolt) reduces galling.
 
Update:

One of the reasons why I am particularly interested in using Belleville Disc Springs is it makes it easy to create a progressive spring (lighter compression in the first part the of the travel, heavier further in). Next step was to try to do so.

Same 12 larger springs combined with 6 smaller springs. These springs in particular.


Here is what I get now:

Starting from finger tight (2 kg)
1 turn = 25 kg
3.5 turns = 100 kg

PXL_20210120_180524934.jpg

What this means is I can set the 0% compression force independently of the 100% compression force by stacking the correct number of each spring type. These results are very repeatable. I do not have the final stack-up determined yet, but that can best be done using a stack of four cells and measuring compression force at 0% SOC and 100% SOC.

My target range of compression force is 50kg at 0% SOC and 95 kg at 100% SOC. That equates to a range of 8 to 15 PSI.
 
Last edited:
One of the reasons why I am particularly interested in using Belleville Disc Springs is it makes it easy to create a progressive spring (lighter compression in the first part the of the travel, heavier further in). So I did this.

Same 12 larger springs combined with 6 smaller springs. These in particular

Or different thickness Belleville washers. Or stack several in a row of same orientation.
 
Or different thickness Belleville washers. Or stack several in a row of same orientation.
Right. With these washers the problem is travel, not force. I am stacking them up in series to get sufficient travel to deal with the expected range of expansion and contraction of the cells while staying within my desired range of PSI. I have several other thickness disc springs if desired.

McMaster Carr has a number of different disc washers I can use with my 5/16" threaded rods. I ordered 4 likely candidates from this list.


I didn't order any of these, and looking at the results, I think these will be a better match to get the 50 kg force.

 
McMaster Carr has a number of different disc washers I can use with my 5/16" threaded rods. I ordered 4 likely candidates from this list.

Hmm. Maybe the washers would perform more predictably on a smooth shaft, not a threaded rod.
The stack probably tries to buckle and washers are jammed against the threads.
Better at opposite end, away from the nut? (something other than threaded rod the entire length.)
 
Hmm. Maybe the washers would perform more predictably on a smooth shaft, not a threaded rod.
The stack probably tries to buckle and washers are jammed against the threads.
Better at opposite end, away from the nut? (something other than threaded rod the entire length.)
The travel is so limited, I don't believe this will be a problem with the Belleville washers. I was testing on a threaded bolt and didn't see any jamming.
 
Posted this in the wrong thread. Reposting this here where it belongs:

Got results from my research to report.

I found a stackup of McMaster Carr spring washers that will do what I think is needed.

My criteria is to limit force at 100% SOC to 15 PSI. I have read that these cells expand between 0.7mm to 0.8mm from 0% to 100% SOC. I am working with 4 cell stacks in each compression frame. This totals up to 3.5mm of travel so that became my goal.

I am using 5/16" x 18 hardware and 18 TPI = 0.058 "/turn:

0.058 "/turn x 25.4 mm/" x 2.5 turns = 3.683 mm of travel.

I ordered and tested a number of different spring washers from McMaster Carr, and one the worked for me is this one.


What I found is that a stack of 12 of these disc springs in series <><><><><><> resulted in a force of 100 kg when compressed with 2.5 turns of the nut. This is starting with the nut finger tight (close to 0 kg). I then turned the nut 2.5 turns and my load cell read ~ 100 kg (+/-5 kg).

I am going to declare victory and that is how I am going to build my compression cells. I am going to keep working to see if I can refine these results, but I think I am close enough to move to building compression cells. When I finish my compression cells I will test a set of 4 cells in compression with the load cell on one of the threaded rods and measure the peak force at 100% SOC. I will determine what the correct preload is then.

Here is a pic of my test setup.

View attachment 33863

And I finally found a use for that old welding coupon. Makes a nice base plate for a load cell.

F.Y.I. Avoid stainless steel hardware like the plague. I started testing with a stainless steel bolt and flange nut and after a half dozen cycles the nut galled on the bolt so bad I had to cut if off my fixture. I am really glad I have a band saw.


View attachment 33865

P.S. Forget using a torque wrench to set compression. I tried that with my load cell and saw tremendous non-repeatability (+/-50%) when trying to use a torque wrench to set pressure. Springs are very repeatable. I am getting +/-5% variation from one test cycle to the next.

Using the torque wrench with lubrication on the threads might be more consistent, but don't forget these cells are going to expand as you charge them. Even if you just run the nuts in finger tight at 0% SOC, I predict you will be applying way too much compression at 100% SOC unless you have springs to provide compliance.
Great work!

I assume you’ve been tracking this thread and seen this post: https://diysolarforum.com/threads/c...w-inverter-rv-build.13786/page-12#post-195598

I’m a bit confused as to whether he’s testing a 6-cell stack or not, but the overall 0% to 100% spring range he is measuring averages 2.2mm (for a total of 6 cells would mean 0.37mm per cell).

Will be very interested to see the range data you measure once you get set up for it.

Also, that poster says he’s seen enough movement of the terminals to be concerned about stress with solid busbars - what busbar solution did you end up settling on for your build?
 
Great work!

I assume you’ve been tracking this thread and seen this post: https://diysolarforum.com/threads/c...w-inverter-rv-build.13786/page-12#post-195598

I’m a bit confused as to whether he’s testing a 6-cell stack or not, but the overall 0% to 100% spring range he is measuring averages 2.2mm (for a total of 6 cells would mean 0.37mm per cell).

Will be very interested to see the range data you measure once you get set up for it.

Also, that poster says he’s seen enough movement of the terminals to be concerned about stress with solid busbars - what busbar solution did you end up settling on for your build?
I am using this braided bus bar. They are only rated for 150A, but since I am using my cells to power a 2000W inverter I am OK with that. If I was using the cells in higher current application, I would double up on the straps. The price is low enough that would not be terrible if I had to.


These straps are 150mm long which means I can use them in a Z shaped orientation. That way cell movement puts almost no stress on the cell terminals.
 
Here is what my battery pack will look like including the compression frames. The two 4 cell sets are completely independent except for cabling between them.

BatteryPack.jpg

Each cell pack will mounted on a set of these rubber shock mounts to keep them from bouncing around inside the battery compartment.

 
Here is what my battery pack will look like including the compression frames. The two 4 cell sets are completely independent except for cabling between them.

View attachment 33928

Each cell pack will mounted on a set of these rubber shock mounts to keep them from bouncing around inside the battery compartment.

Probably the best design I’ve seen yet.

Did you have a system constraint to have the battery + and - terminals close to each other?

Bus bar resistance could probably have been more uniform if you had connected from one 4-cell set to the other the short way (though this would distance the + terminal from the - terminal...).
 
These straps are 150mm long which means I can use them in a Z shaped orientation. That way cell movement puts almost no stress on the cell terminals.
I like this idea. Do the straps have flat ends? The price looks good to me. I will also be using a 2000 watt inverter.
 
Probably the best design I’ve seen yet.

Did you have a system constraint to have the battery + and - terminals close to each other?

Bus bar resistance could probably have been more uniform if you had connected from one 4-cell set to the other the short way (though this would distance the + terminal from the - terminal...).
I like to maintain high current wires in close proximity (minimum loop area). What can I say, I'm an EE.

The resistance doesn't matter since they are in series. That black wire is 1/0 AWG copper wire and it will only be 1' long. I decided to make it longer so that it will be more flexible and easier to install. Short, really thick wires can act like solid bars of copper. Point of this is to avoid putting stress on the terminals after all.
 
Even if you just run the nuts in finger tight at 0% SOC, I predict you will be applying way too much compression at 100% SOC unless you have springs to provide compliance.

Do you think this is true even if you are using something with some give, like wood plates for the end caps? What are you using for endcaps?
 
Do you think this is true even if you are using something with some give, like wood plates for the end caps? What are you using for endcaps?
I am using 1/2" plywood with 1"x1" steel reinforcing tubes. Take a look at this post for details.

 
Updated to show my top balancing compression fixture. It is basically a dry run of the real compression fixtures. I will reuse the metal bits, but will use higher quality plywood for the end caps and add a bottom to the setup.

Dead End.

View attachment 34379

Rubber isolators installed. I have a slide these are going to mount to, so I can pull the batteries out the door of the van, making it easier to load and unload them.

View attachment 34380

Sprung end. I was near 50% SOC so I set the preload for 75 kg. Not cutting the rods until I make the two, 4 cell compression frames.

View attachment 34381
 
Back
Top