EVE-280 cells should these be clamped tight or spaced for expansion?

[danphillips]

Is anyone interested in aluminum supports as in the attached picture. I have access to a fabricator that can cut custom sizes of 1/4 inch aluminum sheet. I haven't checked on price yet but if there is enough interest let me know the dimensions and I'll get some pricing.

 

[HRTKD]

Dan, nice setup. I like the plexiglass on top, that's a nice touch.

My location is going to be quite cramped, so the method I use to secure the battery cells is going to be very customized.

 

[mrdavvv]

Yes on the big side in 4 columns of 8 cells high. They fit nicely in a 30 x 30 metal wall cabinet. They will be clamped. Even though I will have a BMS, I like having access to the terminals for ease of manual testing.

Nice hope you post some photos, for easy access i put a conection block to the BMS to use my custom cables, but with the collateral benefit of being able to easily check voltages (Putting the probes on top of the internal block screws). Also i can change the BMS fast and without messing with the battery terminals too much.




BTW if you havent seen this:

 

[zorlig]

Nice hope you post some photos, for easy access i put a conection block to the BMS to use my custom cables, but with the collateral benefit of being able to easily check voltages (Putting the probes on top of the internal block screws). Also i can change the BMS fast and without messing with the battery terminals too much.

View attachment 16160


BTW if you havent seen this:

Thanks for that. I have the same balancer and want to do the same thing, pulling the data from it down to a computer or raspberry pi

 

[solardad]

@mrdavvv nice setup for the BMS in leveraging the terminal block. I have something similar with my current Leaf pack. The one thing to highlight for others thinking about something similar is make sure your lead wires are the same length. I have noticed some differences with smaller gauge wire and voltage drop - in short you want any drop to be the same so the BMS does not falsely read an incorrect cell value.

 

[mrdavvv]

@mrdavvv nice setup for the BMS in leveraging the terminal block. I have something similar with my current Leaf pack. The one thing to highlight for others thinking about something similar is make sure your lead wires are the same length. I have noticed some differences with smaller gauge wire and voltage drop - in short you want any drop to be the same so the BMS does not falsely read an incorrect cell value.

Yeah!, make sure all lengths are equal just in case..., also check with a good multimeter the end voltages (Terminal block) comparing with a measurement at the battery connector, to see if they are the same.. As you mention, the cables sometimes are bad.. i have even found some that dont have continuity (Probably because someone stretch them too much). Also check that voltages are good and in the correct order (3.2 - 6.4 - 9.6V.....etc)

Finally i recommend putting some crimp connectors or ferrules instead of bare cable, to ensure a good conection. Best case scenario a positive cable presents a false contact and the BMS shutoff, but ive read that if the reference point fails (Negative cable), the BMS can die (At least in the Daly's).

So need to be careful in a few words

 

[Freep]

Your head would probably spin if I posted a picture of my dual Trojan 6v batteries held down on one side with velcro and a cheap strap on the other side. My trailer goes down some fairly rough forest service and BLM roads without any problems.

I ended up using velcro at the top and bottom of each cell to keep them from slipping and to create a bit of a spacer so the cells can expand if need be. Then I used some of the battery packing styrofoam to protect the cells from the hose clamps I used to hold the pack together. Once the hose clamps were tightened, I slipped a little more styrofoam in between the cells at the corners and trimmed the excess.

I really like the custom aluminum supports @danphillips posted but I was in a time crunch and I too have a very unforgiving space.

After doing a bunch of rewiring in the battery box* I installed the battery and then used more leftover packing styrofoam to keep the battery from moving around.


*which amounted to pulling all the cable from the camper to the back of the battery box and then connecting them via wire or shunt to posts in the front of the box in order to deal with the mess o wires I had.

 

[Dzl]

I'll decide after (if) I get any more info from Eve on their "fixture" and any recommendation on applying the 300kgf force they mention in their specs. For now, still waiting on my cells...

Not sure who you're talking to here but it seems like a nice civil discussion to me. If you have any manufacturer information from EVE on recommendations for compression could you post it?

From EVE, there is just a vague mention of a fixture and 300kgf force applied. Lacking any further information, I plan to follow the example of marinehowto.com. If I learn more from EVE that it is vital to have exactly 300kgf force applied, I'll use compression springs. But I've never seen anyone do that yet, and I'm not sure it's worth it over just having rigid plates.

@Airtime I'm wondering if you ever heard back from Amy/Xuba or EVE regarding clarification on compression specs.

 

[Michael B Caro]

Is anyone interested in aluminum supports as in the attached picture. I have access to a fabricator that can cut custom sizes of 1/4 inch aluminum sheet. I haven't checked on price yet but if there is enough interest let me know the dimensions and I'll get some pricing.

Did you ever make any progress on looking into having these fabricated?

 

[Ampster]

@Airtime I'm wondering if you ever heard back from Amy/Xuba or EVE regarding clarification on compression specs.

I remember a thread somewhere that discussed it. I clamped mine because I did not want to stress those little 6mm shallow connections if mine did swell. I used 1/4 inch thread rod with aluminum plates and tightened the nuts with maybe 10 ft lbs of torque. I have had two packs puff on me over the years and the stress on the terminals can be significant, especially if it is a long pack of 16 cells or more.

Here is a picture of my 2P16S pack which is in a 30 x 30 metal cabinet. You can't see the allthread but it is between columns 1&2 and columns 3&4. There is 1/8 inch aluminum plate top and bottom reinforced by angle metal

 

[Airtime]

@Airtime I'm wondering if you ever heard back from Amy/Xuba or EVE regarding clarification on compression specs.

Sorry just saw this post. No I never got any more specific info.

I clamped mine because I did not want to stress those little 6mm shallow connections if mine did swell. I used 1/4 inch thread rod with aluminum plates and tightened the nuts with maybe 10 ft lbs of torque. I have had two packs puff on me over the years and the stress on the terminals can be significant, especially if it is a long pack of 16 cells or more.

I clamped mine as well with 1/4" aluminum plate and threaded rod. I only torqued to 5 ft-lbs when at 30% SOC. With 8 threaded rods spaced around my pack it seemed like enough compression to tighten it up.

Edit: Second guessing myself, I believe I actually torqued less, I'll have to check my notes when next at my shop. I used my 1/4" drive in-lb torque wrench and I'm pretty sure it was significantly less than 60 in-lb (5 ft-lb), maybe 20 in-lb. It's not a cylinder head and shouldn't take very much, I was trying to do a quantified "lightly clamped" at low SOC.

 

[Airtime]

Sorry just saw this post. No I never got any more specific info.


I clamped mine as well with 1/4" aluminum plate and threaded rod. I only torqued to 5 ft-lbs when at 30% SOC. With 8 threaded rods spaced around my pack it seemed like enough compression to tighten it up.

Edit: Second guessing myself, I believe I actually torqued less, I'll have to check my notes when next at my shop. I used my 1/4" drive in-lb torque wrench and I'm pretty sure it was significantly less than 60 in-lb (5 ft-lb), maybe 20 in-lb. It's not a cylinder head and shouldn't take very much, I was trying to do a quantified "lightly clamped" at low SOC.

OK so I checked the torque. My pack is currently at 40% SOC. The torque was about 10 in-lb and I remembered I actually had a reason for trying that, aside from the fact that the pack seems tight with no room for bulging but no over-compression either.

I'm not an ME but I found this site that calculates axial compression force from torque. I think that is only accurate if tightening against a rigid non-deforming surface where the bolt stretches under load--as it should for a properly torqued bolt in say an engine block. But I think the cells will flex or crush under load long before the bolts do any stretching, so I think the actual force applied will be much less than this predicts. Nevertheless, it is a datapoint and will provide a high side estimate of force. Maybe an ME on the site can offer some insight.

I entered 5/16" bolt, 10 in-lb torque, left the coefficient of friction at default 0.2, and it spit out 160lbs axial clamp force. I have 8 threaded rods compressing the plates (2 on top, bottom, each side) so a total of 1280 lbs force applied to my 8 cells. They are configured in two parallel (mechanically) stacks of 4 cells in series (mechanically). If I'm thinking about this right, that force would split across the two stacks of 4 in parallel so 640 lbs applied to each set of 4. The 4 in series (mechanically) would each be under the same 640 lb force.

Since the 300kgf force specified by Eve is 660 lbs, I at least can be reassured that I am not over compressing. But conversely, I don't think I'd want to go too much higher on torque and resulting compression than the 10 in-lbs I have applied.

 

[Ampster]

Thanks for the reinforcement on the torque issue. My BMS also has 4 temperature sensors that I installed at random spots on the pack. We have been having days in the 90s F and I can say that my pack rarely goes above 85 F. I am only using abour 25% of the capacity and most of that drain is at night. I am in the wine country of California and we get cool evenings in the 50s so I am comfortable with the pack dynamics so far.

 

[madmax]

the torque is applied on a plate. if the plate has deflection, then the pressure isn't exerted evenly along the entire pack.

like trying two bricks or trying to clamp your peanut butter sandwich.

 

[BiduleOhm]

I'm not an ME but I found this site that calculates axial compression force from torque. I think that is only accurate if tightening against a rigid non-deforming surface where the bolt stretches under load--as it should for a properly torqued bolt in say an engine block. But I think the cells will flex or crush under load long before the bolts do any stretching, so I think the actual force applied will be much less than this predicts. Nevertheless, it is a datapoint and will provide a high side estimate of force. Maybe an ME on the site can offer some insight.

Even if the cells would compress by say half the axial force will always be the same for the same applied torque (all other things being equal, like for example threads being non-lubed). That's because you measure torque, not a number or turns or another thing.

I entered 5/16" bolt, 10 in-lb torque, left the coefficient of friction at default 0.2, and it spit out 160lbs axial clamp force. I have 8 threaded rods compressing the plates (2 on top, bottom, each side) so a total of 1280 lbs force applied to my 8 cells. They are configured in two parallel (mechanically) stacks of 4 cells in series (mechanically). If I'm thinking about this right, that force would split across the two stacks of 4 in parallel so 640 lbs applied to each set of 4. The 4 in series (mechanically) would each be under the same 640 lb force.

That's correct

Since the 300kgf force specified by Eve is 660 lbs, I at least can be reassured that I am not over compressing. But conversely, I don't think I'd want to go too much higher on torque and resulting compression than the 10 in-lbs I have applied.

Yep, that's perfect

the torque is applied on a plate. if the plate has deflection, then the pressure isn't exerted evenly along the entire pack.

like trying two bricks or trying to clamp your peanut butter sandwich.

Yes, the best is to use C channels in addition to the plate for example, that will increase the rigidity a lot.

 

[Airtime]

the torque is applied on a plate. if the plate has deflection, then the pressure isn't exerted evenly along the entire pack.

like trying two bricks or trying to clamp your peanut butter sandwich.

The plates are 1/4" 6061-T6 aluminum, they are fairly stiff. There are 8 threaded rods--2 on top, 2 bottom, 2 on each side. I torqued the 8 threaded rods as if it was a cylinder head, alternating sides in small steps until all were at 10 in-lb. I suppose there is still some deflection in the plate, but if so I would think that would reduce compression on the cells. I'd rather under-compress a bit than over-compress and crush them.

 

[Airtime]

Even if the cells would compress by say half the axial force will always be the same for the same applied torque (all other things being equal, like for example threads being non-lubed). That's because you measure torque, not a number or turns or another thing.
...
Yes, the best is to use C channels in addition to the plate for example, that will increase the rigidity a lot.

Yes I woke up thinking about that. I realized that if torque is required it's because something must be resisting, and the opposing force is of course the compression of the cells. So I'm glad I stuck with 10 in-lbs! Thanks for confirming. And maybe I will add a couple C-channels, should be easy to do.

 

[nosys70]

the main concern with swollen cells is not the fact that they swell, this seems to be a natural behavior.
the problem is if you connect the cells with rigid busbar, you will get huge force applied to the terminal, because
the top of cell has fixed distance with busbar, the middle of cell will swell and the bottom of cell will move a lot.
so either you minimize this by spacing cells (and loosing a bit on lifecycle), or you use "soft" busbar, like
thick cables, or making sure bottom of cell is as tight as the top of the cell..
by design prismatic aluminium seems to be a lot more sensitive than thick plastic cells.

 

[miky]

I just received these cells from Shenzhen Basen on alibaba, sold me as 2020 production brand new. Can someone help me with these codes for manufacturing dates?

 

[Gazoo]

August 12 2018? The code meanings are listed in the specifications. I am thinking they are new old stock.