diy solar

diy solar

Source of Belleville washers for 300kgf fixture?

PSI is not a meaningful spring rate. Springs are rated using force per distance (lb/in or kg/mm) and free length.

how-to-measure-free-length-of-spring.jpgfree-length-and-deflection-of-spring.jpg

Here is a design process I came up with that should work. I am not an ME so if someone wants to double check my work, please do so.

You have to know 3 things going in:
  1. The compliance distance the spring has to work over. This is how much a cell is going to expand and contract as you go from 0% to 100% state of charge over the operating temperature range multiplied by how many cells you are compressing in a stack.
  2. The minimum and maximum force at each end of the compliance range.
  3. Minimum Spring ID (diameter of the threaded rod plus some clearance value)

Here is what we think we know:
  1. The optimal compressive force on the cell is 12 PSI with an acceptable range of 8 PSI to 16 PSI. For my design I am going reduce the acceptable pressure range to 10 to 14 PSI.
  2. Cell Dimensions (measured from my cells) = 173mm x 200mm = 0.0346m^2
  3. Compliance (how much the cells expand from 0% to 100% SOC) = 1mm per cell x 4 cells = 4mm total
  4. The individual spring force is the total force divided by the number of springs (4)
First convert PSI to kg force applied to each spring:
  • 14 PSI = 9742kg/m^2 * 0.0346m^2 / 4 = 85.1kg
  • 12 PSI = 8436kg/m^2 * 0.0346m^2 / 4 = 73.0kg
  • 10 PSI = 7030kg/m^2 * 0.0336m^2 / 4 = 60.8kg
Once you know this we calculate the spring rate = (max force - min force) / compliance:
  • Spring Rate = (85.1kg - 60.8kg) / 4mm = 6kg/mm
Spring loaded length (length where force = 73kg) = 73kg / 6kg/mm = 12.2mm

Spring free length (length where force = 0kg) = 24.4mm

I am using 5/16" threaded rod so the spring ID should be > 8.5mm

Looking at the Lee Spring on-line catalog, one spring that meets these requirement is LC 049E 09 S
 

Attachments

  • LC 049E 09 S.pdf
    528.3 KB · Views: 10
Last edited:
In use, adjust the spring preload to 50% (12 mm). It should then be applying ~ 70 kg force.
 
Turns out I don't know my butt from a hole in the ground. Forget that spring. IT WILL NOT WORK!!!

I need to be taking into account the solid length (fully compressed).

I will be back after I edumacate myself.

I bet all the ME's where just laughing at my hubris.
 
F.Y.I.

I found a source for 120mm x 16mm, M6 braided ground straps rated at 150A. These would work for diagonal bus bars (eliminating the risk of stress on the cell terminals).


For my application (2000W Inverter), 150A is plenty. If you need the full C rating of the cells, you could double them up.
I am wondering if you received the ground straps and if you did what do you think of them?
 
As you guys are finding out. ... spring rates and clamping force for threads (torque vs clamping force) is a non-trivial calculation.
 
As you guys are finding out. ... spring rates and clamping force for threads (torque vs clamping force) is a non-trivial calculation.
Indeed,

I feel much more confident with what I am doing now, since I have a load cell and can actually measure the clamping forces directly. Belleville Disc Springs are great. First pick one with the correct working load then just stack up multiple discs in series until you get the desired amount of travel distance. The disks I picked have a working load of ~ 187 lb which is almost exactly what is wanted. The only remaining question is how much travel is enough. It is looking like 1 to 2 mm of travel is sufficient, so my plan is to use 10 disc washers in series. Charge the cells to 100% SOC. Then adjust the spring preload to the max compression I ever want to the cells to see: 195 lb per threaded rod, or 780 lb total which is very close to 14 PSI and call it a day. As the cells discharge, the springs will relax and the compression forces will drop to a minimum at 0% SOC of 110 lb per threaded rod, or 440 lb total which is very close to 8 PSI. This range encompasses the sweet spot of the life/compression chart that EVE has published.
 
Not here yet. Grrr.
Thanks. If you don't mind please post when you receive them. I am considering buying the 100mm ones which should provide enough slack between the cells. I know it's not as good as your design but I already have the short cable to connect the cells together.
 
Thanks. If you don't mind please post when you receive them. I am considering buying the 100mm ones which should provide enough slack between the cells. I know it's not as good as your design but I already have the short cable to connect the cells together.
Will do.

The braided straps I had previously purchased will work, you just want to double them up to get the ampacity high enough. If you are expecting to power a larger inverter, perhaps triple them.

These guys custom built me straps to the length I specified. I did this before I decided to go with the diagonal straps.


The cost was very reasonable.
 
This is what they look like. I am using the ones I ordered to connect to the most negative cell to the BMS and the most positive cell to the circuit breaker. I also think I will use them to connect the two packs together. Doubled up, of course.

left_75_0152_straightup.jpgright_75_0152_straightup.jpg
 
Thanks. My inverter is rated at 2000kw and 24 volts. So I think the 150 amp busbars Icotek will be good enough. I oversized the inverter anyways and probably will never use the full 2000. I am concerned about the possibility of the ends of the busbars buckling when mounted. I read it happened to someone else and I do recall they bought the braided busbars from somewhere else, not Icotek. As usual I don't recall where it was posted. :unsure:
 
Thanks. My inverter is rated at 2000kw and 24 volts. So I think the 150 amp busbars Icotek will be good enough. I oversized the inverter anyways and probably will never use the full 2000. I am concerned about the possibility of the ends of the busbars buckling when mounted. I read it happened to someone else and I do recall they bought the braided busbars from somewhere else, not Icotek. As usual I don't recall where it was posted. :unsure:
Mine is the same. A Multiplus 24/2000-50. I am planning on 150A peak load current, so all of my wiring, bus bars etc. are sized for that.

Check out my latest circuit diagram. I simplified the battery design a bit. I like this Inverter capacitor pre-charge concept better than previous ones.

I am accounting for the fact I will be using the braided straps to connect the battery packs to the BMS and circuit breaker. I also moved the battery voltage monitor to cell side of the BMS so it will show the actual cell voltage. The BMV-712 will show the battery pack output voltage.
 

Attachments

  • VanPower15.pdf
    504.4 KB · Views: 12
Last edited:
Indeed,

I feel much more confident with what I am doing now, since I have a load cell and can actually measure the clamping forces directly. Belleville Disc Springs are great. First pick one with the correct working load then just stack up multiple discs in series until you get the desired amount of travel distance. The disks I picked have a working load of ~ 187 lb which is almost exactly what is wanted. The only remaining question is how much travel is enough. It is looking like 1 to 2 mm of travel is sufficient, so my plan is to use 10 disc washers in series. Charge the cells to 100% SOC. Then adjust the spring preload to the max compression I ever want to the cells to see: 195 lb per threaded rod, or 780 lb total which is very close to 14 PSI and call it a day. As the cells discharge, the springs will relax and the compression forces will drop to a minimum at 0% SOC of 110 lb per threaded rod, or 440 lb total which is very close to 8 PSI. This range encompasses the sweet spot of the life/compression chart that EVE has published.
HaldorEE -- did this end up working out for you? I have went through the calculations and what you say makes sense. You are assuming hooke's law with the belleville disc springs for when the battery drains ... hopefully that is good enough. I'd like to know what you found, were you able to take any measurements to confirm or deny how well it turned out? Thanks in advance ...

If this works, I like it because it doesn't require nearly as much space consumption as compression springs, and isn't as involved as using expansion springs with turnbuckles (and I believe this would be expensive buying all the hardware for expansion).
 
Indeed,

I feel much more confident with what I am doing now, since I have a load cell and can actually measure the clamping forces directly. Belleville Disc Springs are great. First pick one with the correct working load then just stack up multiple discs in series until you get the desired amount of travel distance. The disks I picked have a working load of ~ 187 lb which is almost exactly what is wanted. The only remaining question is how much travel is enough. It is looking like 1 to 2 mm of travel is sufficient, so my plan is to use 10 disc washers in series. Charge the cells to 100% SOC. Then adjust the spring preload to the max compression I ever want to the cells to see: 195 lb per threaded rod, or 780 lb total which is very close to 14 PSI and call it a day. As the cells discharge, the springs will relax and the compression forces will drop to a minimum at 0% SOC of 110 lb per threaded rod, or 440 lb total which is very close to 8 PSI. This range encompasses the sweet spot of the life/compression chart that EVE has published.
HaldorEE I'd also love to see what the end result ended up looking like and which washers from McMaster you ended up using?

I couldn't seem to find the 5/16 187lb ones mentioned but I did find some rated at 144 lb working load. Also what did you end up using between cells? I saw some people in another thread I think using neoprene from McMaster.
 
Back
Top