Which glue to join cells ?

[qwerkus]

Hello,

I'm working on a new pack where there is just no room for one of those nice abs cell holders. So I have to glue the cells together. Did some test with cheap hot glue, but it's frightening how easy you can take the cells apart again. No way this is going to survive motor vibration. I read somewhere em3ev uses "special heavy duty hot glue" - anyone knows if this is just marketing bs ? I'm considering epoxy, but even if I join the cells with a strong bond, the weakest link is just going to move to the super thin pvc cover, isn't it ?
Any advice would be much appreciated.

Best regards,

qwerkus

 

[flippy]

use these cell holders:

https://www.aliexpress.com/item/1-Pair-10x-Cell-Plastic-Battery-Holder-18650-Battery-Spacer-Holder-Cylindrical-Cell-Bracket-for-Battery/32889204292.html

and heavy duty hot glue. search for the hot glue used to pull dents from cars.

 

[john61ct]

Do you want to ever get them apart again?

There are hundreds of specialist adhesives that would do.

Selecting the ideal top dozen or so would be a challenge.

Flexibility for shock absorption and vibration resistance.

Thermal conductivity combined with high electrickal insulation (dielectric strength) would be critical.

And yes, nothing you'd find at the local hobby shop

A call into 3M tech support would be productive.

Be prepared to spend a lot per gram.

And you'd still want a protective enclosure.

 

[john61ct]

No "hot glue" would be appropriate, nothing water-based. Thermoset rather than thermo-plastic.

I's start by looking at epoxy-based potting compounds, maybe conformal coating.

Oh and fire suppressive, not too toxic when in flames is another desirable characteristic.

 

[Hillhater]

Unless you are going to strip all the shrinkwrap insulation off all the cells (risky for a tightly configured pack ). Then those wraps will always be your "weak link" and as long as you find a glue that bonds well to the wrap, there is no point in sweating over chasing down some high tech adhesive

 

[john61ct]

The cell walls are negative right to the gasket at the + end.

Full potting is what's called for, so thermal conductivity will be very critical.

In aluminum maybe

 

[Hillhater]

Full potting is what's called for, so thermal conductivity will be very critical.

Why ? Who called fot that ?

 

[flippy]

The cell walls are negative right to the gasket at the + end.
Full potting is what's called for, so thermal conductivity will be very critical.
In aluminum maybe

potting cells is bastically the the worst thing you can do.

 

[redilast]

When I have to glue cells together I use Quick Grip. It works great, highly recommended! Been using it for years. Part number 873388

 

[john61ct]

Obviously depends on whether or not there is anything else holding together and protecting the cells.

If not, are you saying just rely on gluing the plastic skins together at the interior connecting edges and done?

 

[amberwolf]

Since the shrinkwrap is the weakest link, especially as the cells heat-cycle it and age it, making it more and more brittle over time, the glue can be just about anything that's nonconductive, and that won't come undone or melt as it gets hot inside the pack (dependent on the temperature the pack reaches inside, which will be greater as it ages, as the cell Ri goes up over time).

Some glues have issues, like common hot-melt glue that can come loose as it ages and embrittles. (I don't know if all hot-melt glues do this, but at least some of them do; I've disassembled lots of things over the years that had hot-glued parts that weren't connected anymore).


If there's no outer cell holder (other than say, duct tape) to secure the cells in their relative positions, then the only thing that does it is the glue and the cells' shrinkwrap, which are both slightly flexible. THis can put stress on teh spotwelds, wiggling them a bit at every vibration and bump. Sometimes they eventually break, disconnecting cells from the rest of the group (reducing that group's capacity and current capability, causing imbalance between groups).

This is a (probably common) failure mode seen in common "duct tape" packs (which may use heatshrink over the pack instead of duct tape, but doesn't change the problem much, if any).

 

[markz]

The cell walls are negative right to the gasket at the + end.
Full potting is what's called for, so thermal conductivity will be very critical.
In aluminum maybe

potting cells is bastically the the worst thing you can do.

https://www.ebikes.ca/product-info/ligo-batteries.html

After much design consideration we decided to implement this as a flat 10s x1p 18650 cell layout, with every battery having its own independent BMS circuit and each pack being potted to protect against the elements.

Testing
https://www.youtube.com/watch?v=oBlac5gWLfI

 

[flippy]

yeah, i know. and as soon as a single issue pops up you can trow away the whole pack and buy a new one. solid business model, ensures repeating sales...

 

[qwerkus]

Sounds like there is no easy solution to this problem. Stripping the cell protective shrinkwrap is possible, together with using a strong metal binder. That would solve the problem for one P group, but not between the P groups. Here I'd still need an insulant glued on each side to a different P group. Maybe just simple fishpaper ? On the positive side, the Battery case itself is really strong. 8mm wall heavy duty PE pipe with screwed covers. Also the cells are a super tight fit, so they won't move around inside the box. The only thing the glue has to do, is ensure no weld spot comes loose.

 

[john61ct]

Stripping the cell protective shrinkwrap is possible

In which case the cell walls as part of the negative common side of the circuit, is much more likely to get shorted to the positive side when a mishap occurs.

> together with using a strong metal binder.
what is the meaning of binder here?


> On the positive side, the Battery case itself is really strong. 8mm wall heavy duty PE pipe with screwed covers.
Link to the product you're referring to please?

> The only thing the glue has to do, is ensure no weld spot comes loose.

Glue cannot do that, unless you're talking potting.

How much shock / vibration you talking about?

There are conductive tapes and epoxies maybe can use rather than welding.

Flexibility in helping to absorb shock / vibration is the key, but can surround the batts or even their case, rather than in between each one.

 

[larsb]

There are conductive tapes and epoxies maybe can use rather than welding.

If you find a 20A capable tape or glue then please let me know
Resistance in the order of mA needed..

 

[qwerkus]

> together with using a strong metal binder.
>>what is the meaning of binder here?
https://www.jbweld.com/collections/metal

> On the positive side, the Battery case itself is really strong. 8mm wall heavy duty PE pipe with screwed covers.
>>Link to the product you're referring to please?
https://www.international.geberit.com/products/piping-systems-drainage/geberit-pe-discharge-pipes/

> The only thing the glue has to do, is ensure no weld spot comes loose.
>>Glue cannot do that, unless you're talking potting.
You are correct. The actual job of the glue is: bind cells together so they do not move around; an play between cells endengeres the welding joint over time.

>How much shock / vibration you talking about?
Not sure yet. Probably bafang bbs02 or HD.

 

[john61ct]

There are conductive tapes and epoxies maybe can use rather than welding.

If you find a 20A capable tape or glue then please let me know

I'm thinking, if high enough capacity wiring is in solid direct contact with the battery terminals, the role of the outer adhesive layer is simply to prevent lateral slippage.

 

[john61ct]

Sounds like there is no easy solution to this problem. Stripping the cell protective shrinkwrap is possible, together with using a strong metal binder.

JB Weld is not conductive, so that's good.

It is just another adhesive though, not specifically intended for a particular level of dielectric strength.

It also does not have specific thermal conduction specifications.

And nearly every high strength industrial adhesive bonds well to metal surfaces with proper prep.

IOW JB Weld brings nothing special to this use case.

 

[john61ct]

PE piping as battery case. Seems unlikely to allow rapid heat transfer outward?

And looking to use a glue, between the cells and the interior pipe walls?

Having a hard time visualizing this.

What diameter pipe, what length, how many cells per bundle, maybe a picture?

 

[john61ct]

>How much shock / vibration you talking about?
Not sure yet. Probably bafang bbs02 or HD.

I meant, is this for a boat, a bike, off-road racing on corrugated dirt or what?

 

[larsb]

I think you're just complicating things. Used a good quality heat glue on my battery packs, 4years later it's still soft and in place - batteries are worn out..

But i live in a not so thermally challenging part of the world , could be different in Nevada desert.

 

[john61ct]

Batts getting hot are also an issue depending on the design.

130-140°F is not unusual at high C rates.

 

[larsb]

130-140c not unusual, got to be kidding me
That's killing the batteries, pack should not be used that high. Normal stop is at 80 deg c for a lithium battery.

 

[john61ct]

I agree not good for longevity and BMS should have temp-based cutouts set lower.

But planning the infrastructure with headroom is a good idea, and fact is many do get that hot on the inside positives.