I think the flexing issue has been "answered" but I can tell you that my 0.6" plywood flexed in the middle with my 4-rod construction before I added the square tubes. Imagine hanging the board 4" off the end of the table and then placing 300 pounds of pressure at the end - would it flex at all? even 1mm of flex throws all the calculations off.
Pack / Cell compression Optimized By Using Springs.
- Thread starter Bob B
- Start date
I think the flexing issue has been "answered" but I can tell you that my 0.6" plywood flexed in the middle with my 4-rod construction before I added the square tubes. Imagine hanging the board 4" off the end of the table and then placing 300 pounds of pressure at the end - would it flex at all? even 1mm of flex throws all the calculations off.
Well this is embarrassing. I do have some formal ME education. Tunnel vision is a real thing! Standby for updated calcs.
I don’t know what you see as ‘crude’ about the measurements we are making for this.
The 5/16” rod has 18 threads per inch (1.41mm/turn) and it pretty easy to control turns to within +/-1/16th turn = 1/8th turn = 0.18mm.
The spring I’m considering has a rate of 389 lbs/inch or 15.3lbs/mm.
So with a precision of 1/8th turn or 0.18mm of precision, pre-load can be controlled to within 2.76lbs or 0.05psi.
I spoke to the sales rep and confirmed that this is exactly the correct manner to set preload (he even called me the ‘perfect customer
).And I also learned something very interesting from him: if you preload one of their springs to solid (just once) it will not fully spring-back (so that is definitely something you want to avoid and it makes me wonder about Lee Springs in that regard).
Any preload exceeding their Max compression risks the spring not springing back to it’s full initial Free Length, so at least with those springs, you want to be conservative and calibrate Free Load from Free Length upward...
The Max compression results in lower cycle life of the Spring returning to its full initial Free Length, but since we are talking about 100s of thousands if not a million cycles, this is a non-issue for our application.
Hedges
I See Electromagnetic Fields!
- Joined
- Mar 28, 2020
- Messages
- 20,789
I tended to do that to professors (detached from real world) and Stanford PhD's.
A guy could calculate the curl, determine derivative, generate contour plots ...
but he didn't understand how heat flows, wouldn't understand that when tide is flowing into the bay there is a depth difference at the end (and depth difference tells you water is flowing), if he was a surveyor could generate a topo map but wouldn't be able to predict where runoff would flow. He is the author of a book and an instructor on the topic.
I bought his book (just so I could mock him with his own words), and it dove quickly into integral calculus of 3-dimiensional problems. But it also cited the errors he was making.
Wouldn't accept when measurement results didn't match his simulations ...
I've had to learn, too. Passed myself off as an RF and analog expert, but learned from others along the way. It was only from on-the-job feedback and reading app notes that I learned the pitfalls of using op-amps as comparators, had to re-engineer some of my circuits.
Im not doing this with you or going to derail the thread again. If you cant figure out what was meant by what was said, Im at loss.I don’t know what you see as ‘crude’ about the measurements we are making for this.
The 5/16” rod has 18 threads per inch (1.41mm/turn) and it pretty easy to control turns to within +/-1/16th turn = 1/8th turn = 0.18mm.
The spring I’m considering has a rate of 389 lbs/inch or 15.3lbs/mm.
So with a precision of 1/8th turn or 0.18mm of precision, pre-load can be controlled to within 2.76lbs or 0.05psi.
I spoke to the sales rep and confirmed that this is exactly the correct manner to set preload (he even called me the ‘perfect customer
And I also learned something very interesting from him: if you preload one of their springs to solid (just once) it will not fully spring-back (so that is definitely something you want to avoid and it makes me wonder about Lee Springs in that regard).
Any preload exceeding their Max compression risks the spring not springing back to it’s full initial Free Length, so at least with those springs, you want to be conservative and calibrate Free Load from Free Length upward...
The Max compression results in lower cycle life of the Spring returning to its full initial Free Length, but since we are talking about 100s of thousands if not a million cycles, this is a non-issue for our application.
Pretty sure we are talking about different flex.
I understand that without a reinforcing beam such as square tubing, wood in general and plywood in particular will flex, primarily side-to-side (the center will push out more than the sides where the springs are positioned).
What I am talking about is the difference between a single central beam across the ends halfway up versus the classic 2-beam design crossing the ends at 1/3 and 2/3 or 1/4 and 3/4 (or whatever),
Especially since if I go with wooden ends I’ll use 2x8 with vertical grain, I’m not too concerned about any flex in the vertical direction (and I’m not concerned about flex in the horizontal direction with 1-1/8th square tubing, whether single-through-middle or double-high-and-low).
What I am possibly worried about with a single support beam across the middle of the ends is that the ends may be able to pivot/rotate (not flex) so that the bottom of the cells can continue to expand more than the top.
Your dual-beam fixture eventually settled into uniform 4mm of expansion along the entire height of the cell.
My unfixtured cells expanded more at the bottom than the top (possibly because of the solid busbars) and I’m concerned that if I only go with a single pair of threaded rods and a single 1-1/8th square-tubing beam across the middle of either end, that non-uniform ‘pyramid’ expansion may continue...
Cutting the hardware cost in half is appealing, but not if it is going to introduce a new problem...
Of course, since I have two 8S batteries and only need one for now, the option I’ll probably choose is to purchase all of the hardware for 2 single-beam designs, build three first of those, and if I’m unhappy with how it performs, move the existing beam and add a second.
There is also one other Forum Member who has already built a single-beam fixture, so I may just hold off a bit longer before finalizing my plan to await his first test results...
What is the catalog item number of the spring you are considering? Please.I don’t know what you see as ‘crude’ about the measurements we are making for this.
The 5/16” rod has 18 threads per inch (1.41mm/turn) and it pretty easy to control turns to within +/-1/16th turn = 1/8th turn = 0.18mm.
The spring I’m considering has a rate of 389 lbs/inch or 15.3lbs/mm.
So with a precision of 1/8th turn or 0.18mm of precision, pre-load can be controlled to within 2.76lbs or 0.05psi.
I spoke to the sales rep and confirmed that this is exactly the correct manner to set preload (he even called me the ‘perfect customer
And I also learned something very interesting from him: if you preload one of their springs to solid (just once) it will not fully spring-back (so that is definitely something you want to avoid and it makes me wonder about Lee Springs in that regard).
Any preload exceeding their Max compression risks the spring not springing back to it’s full initial Free Length, so at least with those springs, you want to be conservative and calibrate Free Load from Free Length upward...
The Max compression results in lower cycle life of the Spring returning to its full initial Free Length, but since we are talking about 100s of thousands if not a million cycles, this is a non-issue for our application.
I understood what you meant -- I wanted you to know that the plywood flexed, so I would be concerned about the pyramid affect in your single beam design, for sure.
The 8" spring I was musing about that fafard thinks is a waste is this one: https://www.leespring.com/compression-springs-hefty?search=LHL750B15
I have ordered these ones:
- For the outside 4: https://www.leespring.com/compression-springs-hefty?search=LHL1000B09
- For the inside 6: https://www.leespring.com/compression-springs-hefty?search=LHL1000C09
Well, I am not going to go 8 but I am going to go longer. Makes sense.
If I came off as harsh, my apologies. I was considering that exact same spring for those exact same reason (space to burn, so why not maintain as close to 12.1psi as possible?).
The two things that have caused me to reconsider are:
1/ 36” length of my 5/16” threaded rod - I’ve got an 8S pack of 280Ah cells which adds up to just shy of 24”. Add 1-1/8” per side for square tubing, 1-1/2” per side for ends made out of 2x8, and at least 1/2” for nuts and washers on both ends, and suddenly I’m down to ~6-1/4” and no longer have room for an 8” spring unless I move up to 48” threaded rod...
2/ I was originally designing for my measured (unconstrained) expansion, which was over 8mm for eight cells. Now that others have reported that 1/2mm per-cell expansion is all you’ll have to deal with once the cells have settled down through s few cycles in the fixture, it changes the math significantly. If I can majntain anywhere between 10.1psi and 14.1psi (with headroom for an additional 2-4mm of unexpected expansion), I’ll be pretty pleased...
Cool, thanks. That’s probably going to nudge me towards the classic 2-beam design like you used...
9-1205-21 is the Spring I was looking at for a single-beam (single pair of rods), dual spring (spring in either end of each rod) design. [ie: 4 of these springs total]
I’ll need to find other springs if I want to go with a classic dual-beam design (whether single-spring or dual-spring)...
Oberon
New Member
- Joined
- Jan 27, 2021
- Messages
- 32
I'm planning to compress my single-row 8s battery with springs, however my guess is that the expansion is not equal across a cell's width - surely it expands more in the middle than at the edges? So I'm considering adding a resilient material between each cell, in addition to the springs providing clamping force, to allow a bit of flexibility for some cell bulge. Something like a 1/4" rubber mat. Thoughts?
Hedges
I See Electromagnetic Fields!
- Joined
- Mar 28, 2020
- Messages
- 20,789
Electrical insulation sounds good, since we hear metal case has electrical contact.
People have had their cells maintain spacing where secured, bulge where not. Seems like something extrudes, like biting into a pb&j sandwich.
Given that they're pouches of rolled layers, I'd expect it to try to expand everywhere. But edges are constrained by the can.
I would try to spring-load it between flat surfaces, but use compliant busbars.
People have had their cells maintain spacing where secured, bulge where not. Seems like something extrudes, like biting into a pb&j sandwich.
Given that they're pouches of rolled layers, I'd expect it to try to expand everywhere. But edges are constrained by the can.
I would try to spring-load it between flat surfaces, but use compliant busbars.
I'm finding that my cells remain perfectly flat if they're under compression, so I wouldn't worry too much about it - as long as the end plates are rigid enough to spread out the pressure provided by the springs, all the cells inside will be fine. The only reason I used 1/16" neoprene is because I'm in a mobile application and I really want to prevent movement and/or abrasion.
Oberon
New Member
- Joined
- Jan 27, 2021
- Messages
- 32
Good to hear that the cells stay flat. I'm also building for a mobile application, so those are good points about preventing abrasion and electrical contact. I'll look into adding something, but I won't worry about it having much resilience.
Now on to spring force calculations!
Now on to spring force calculations!
I like the look of this. Its closest to my thoughts at this stage.. My complication is I'm doing a 24 cell , 24V pack as 4 rows of 6 cells. Since each 'triplet' needs busbars, the complication of allowing movement means more complex connections of the 3. As designed below with 20mm x 1mm busbars tripled between all sets of 3 (plus one doubled across the back on all 6 terminals). I'm now leaning to a compressible layer between all batteries of a neoprene material, compressed to about 2mm between cells, under a fixed bolt compression. I have no confidence the resilience of the neoprene will remain constant under 2000+ cycles , and as to how to calculate approx pressure for another 1mm compression would be complex.
I'm in awe of the great work all you guys are doing in this thread, attacking a thorny issue that seems to be ignored by most commercial pack battery makers.
Attachments
Electro Dan-O
Current Situation
Here you go guys! I found out the hard way that 20 pictures is the max for one post.... It's a start and hopefully some ideas to build on using springs!
Fantastic, now that I see how you DIY your flexible busbars, I might revisit my decision to not go springs, for fear of movement of cells on fixed simple rigid 3 post busbar. I could make a 3 cell busbar in one with 2 flexi bits between allowing for movement. How many amps is that braided stuff rated for ? I need 200 Amps to be safe on my design ?
Similar threads
