Brass for spot-welding 18650 parallel cell groups?

[fechter]

OK, tried with the riba welder. No luck with the brass. It came close but didn't really stick. I could tell I was getting good weld current by the way the cables jump. My riba welder has beefed up source connections and freewheel diode. My supply battery is still a bit on the small side, a 35Ahr AGM battery, but it was fresh and fully charged.

I also tried welding some .15mm nickel to the brass. This worked pretty well on the highest setting. On the end cell groups where you want a hefty parallel connection, this may be a good alternative to soldering copper wire.

I'm starting to think if you have enough current to weld the brass properly, you will be risking blowing holes in the cells.

 

[litespeed]

Well I can tell you that happens pretty easy or at least in my case it did. That is one of my worst battery building nitemares is to blow a hole in a cell in the middle of the pack. The dead one I blew a hole in yesterday bubbled and oozed out for a couple hours. I set in on my concrete steps at my walk out surrounded by concrete block. Being a storage charged cell in the middle of a hot glued pack would/could be a hell of a story one day.

Did I mention I have nerves of steel and that I'm fearless?

Didn't think so, that was more than a few unexperienced years back.

Tom

 

[kdog]

Try copper electrodes, I had no success with tungsten but copper works fine. I used solid core copper wire.
I can't do thicker than .2mm un slotted.

 

[kdog]

ive got myself some guilding metal. its c21000
its 95% cu and 5% zn with an IACS of 55%.
i had to buy it in 1.6mm thick so ill cut some strips and get them rolled out hopefully. in the mean time i might just hammer some out.
so far ive been able to weld
pure copper .1mm
brass 60/40 .2mm
red brass 70/30 .2mm and .25mm (slotted only)
ill post a summary after ive done the guilding metal- which im secretly hoping will be the shizz.
Edit corrected thickness of red brass

 

[litespeed]

So good solid welds? I can get what looks like a weld but it's weak and kind of brittle at the weld site.

I'll try some copper when my shoulder is usable.

Tom

 

[kdog]

yeh the brass welded reasonably well- especially the .25 red brass (70/30) but only if it was slotted- i couldnt weld it un-slotted. bc it was a bit thicker, it was really strong, almost the same as .2 nickel.
if i could find a supplier red brass would be a good option.
i really like the pure copper except that it is so weak. its actually pretty easy to weld now that ive got the hang of it and it carries good amps, but it break soo easy im worried about building another pack with it. (i really should inspect the one i did a year ago) the welds get an annular fracture around them if you get it a bit hot. there are fancy current profiles that minimse this but thats wayy to hard for me.
im going to test this guilding metal soon, but the alloy i really want is c19700. it has better stress relaxation so i think it should help with the fracturing and its IACS is 80%, prob perfect for battery building.
anyone in/near rhode island want to help me out getting a sample? :wink:
Edit- correction of thickness of red brass

 

[spinningmagnets]

Do you have an address, email, or phone number for them?

 

[fechter]

I guess the fatigue strength of the welds is an important factor. Not so good if it breaks after being flexed a few times. Even if the cells are glued together, there will be some expansion from thermal cycling.

 

[kdog]

hey SM
theyre called millard wire
they say they are happy to accomadate small orders but didnt respond to my email. might have beenchucked in the too hard basket given im in aus
contact details:
MILLARD WIRE & SPECIALTY STRIP COMPANY
449 Warwick Industrial Drive
Warwick, RI 02886
TEL: 401-737-9330 • FAX: 401-737-9340
link to web site
http://www.millardwire.com/
theyve got a couple of alloys id like to try if you hear from them...

 

[kdog]

Well I hammered out some gilding metal to .15 and .2mm. Its not the best way to prepare a sample cause of the surface irregularities. So the results were a little inconsistent. But having said that I welded the .15 mm straight to the can with reasonable results. I say reasonable because whilst the first weld pair were great( nice and strong) the second weld was a bit hit and miss sometimes good sometimes didn't take.
The .2mm slotted was just ok straight to the can- wouldn't persue it, but if welded to nickel this resulted in a good weld.
So in summary so far:
.1mm pure copper: OK (main daw back is weakness)
.2 yellow brass : good
.2 yellow brass slotted: very good
.25 red brass:very very poor (no weld at scary power)
.25 red brass slotted: very good (very strong,but req a decent amount of power)
.15 gilding metal slotted: good
.2 gilding metal slotted: fair (hard to stick second weld)

All results better if welding to nickel instead of directly to can.
Slotting improved results even more so than in nickel..I guess with high conductivity alloys the trouble is actually getting the current to go down into the can to make the weld points, hence slotting making a bigger difference than in nickel. And I needed to keep the welds as far apart as poss to help stick the 2nd weld.
I've got some .15 yellow brass on order. I want to see if this will make good broad sheet peices to cover entire P groups- think Madin88 style.
( If anything it is way cheaper than nickel. )
I'm trying to find two separate outcomes .
One is a cheap nickel substitute- doesn't necessarily need to carry more current, but needs to weld well. I'm thinking broad .15yellow brass for this. Cheap, available every where,will prob weld fine.
The other, a high current option that can take around 20 amps from a cell(s) and not get hot.
Still working on this but .25 red brass or .2mm gilding metal may be contenders esp if welded over nickel
Hard to source though...

 

[kevbo]

Just a thought:

Instead of slotting a single strap, it might be easier to use two straps in parallel, welded at the same time, one electrode on each strap, forcing the cell to complete the circuit. You would need some means to maintain a small gap between them until they are welded down. Tape might be enough

 

[spinningmagnets]

The "two strips" idea is what I want to try for the parallel connections, after I get my hands on a spot-welder. I like the configuration of using the plastic cell holders, and they have an 8mm slot pattern across all of the cell tips for both parallel and series.

I'd set up a jig away from the cells, and spot-weld the copper series flat-bars onto the brass parallel strips. Then set it onto the cells in their plastic holders, and spot-weld the assembly onto the cell ends.

Each bus assembly would look like a ladder, with the steps being thick copper, and each side-post being two thin brass strips very close to each other. Edit to add drawing. The problem with spot-welding the parallel strips onto the cell-tips is that the heat might damage the cell if yo have to take it too high. Brass has "just enough resistance to spot-weld without damage, and more conductivity than nickel, so less heat wasted from current (plus available and cheap)

Copper is super hard to spot-weld onto the cell ends, but we only need the series connections to be high current, so...if we spot-weld the brass and copper together (using high heat) away from the cells, no risk of damage to the cells. I may have to make the copper strip-tips forked to get the welding probes onto the brass for the final welding to the positive ends of the cells (the broad negative ends need no special considerations), but that would be easy to figure out. I might even be able to weld the brass and copper together (in a jig to hold the precise shape of layout) with a caveman MOT welder.

 

[kdog]

Parallel strips works awesome for the first weld. Then it's still pretty good for the second. But... It soo fiddly and time consuming it's often not worth it. A longer (20mm) slot works almost as well without all the difficulties. having said that there would be times when it probl is worth it though.

 

[vex_zg]

bought some .2mm yellow brass from aliexpress.

Welds easier than copper, less energy needed. Little bit more energy needed comparing to nickel.

Mechanically - better than copper. More flexible, less prone to shearing when wiggled. Worse than nickel though. Difficult to quantify, but perhaps halfway between nickel and copper mechanically.

Electrically - I can only generate 5A of test current at the moment, and it doesn't get warm. I would need to get it longer to measure resistance.

so if nickel got difficult or expensive to get, I think this would be a viable alternative.

 

[vex_zg]

I'm a bit disappointed actually that nobody picked up on this topic, but I'll let it stay in the archives of the internet...

anyway, I did some electrical testing, and roughly, this brass has typical resistance some 30-50% lower than pure nickel (for same cross section of course).

 

[vex_zg]

I think if your goal is incremental improvement over existing spot welding, then brass is interesting. If your goal is to get to copper then it is not.

If there was a solution to make copper work, why would anyone pursue nickel or brass?

I can get copper to weld but the energy required for that weld is a bit high, in range of 200J making it a bit unsafe for cell IMO. Also the mechanical durability (elasticity and strength) are lower than nickel.

.2mm Brass welds similar as nickel, needs about 60-70J per weld. It's cheap and available.

I see brass not being competition to copper, but an alternative, easier to weld, and better mechanically.

For the record, I'm using Capacitive Discharge welder which I designed myself, 3F capacity, I'm working with weld energies instead of pulse lengths but pulse lengths at around 1-1.5ms end up releasing the energy around 60-80J @12V)

 

[jonescg]

The internal impedance of the cells is what generates heat under heavy loads - I^2R basically. So if your battery has an internal resistance which causes heat, the last thing you want to do is contribute to that loss by using lousy conductors. As has been covered already - Nickel is a lousy conductor, but for most purposes it's 'good enough'. If however you start pulling high amps the system heats up and the resistance increases. In addition to the electrical conductivity, the thermal conductivity will play a big role in the thermal properties of the system. Nickel is a lousy conductor of heat too, so any heat being generated at the conductor is going to be soaked up by the closest thing with enough of a temperature difference - the cell! So your cells will get extra hot if you skimp on the conductors.

So it seems that a copper contact would be good, but clearly quite difficult to do with a spot welder. Solder will make a good joint, but has issues with application too (easy to overheat the cell etc). A decent thickness of copper, like 1 mm or so, and pressure contact seems like a good option for power applications. I'm more convinced than ever that a decent conductor will resolve a lot of needless heating under heavy discharge.

 

[spinningmagnets]

I am revisiting brass as a bus material, because there have been some good results from spot-welding with the kWeld by ES member Ossivirt.

https://endless-sphere.com/forums/viewtopic.php?p=1442806#p1442949

50j was maybe little much for 0.10mm materials so I tried with 45j and it was little better in my opinion. 0.1mm copper was both cases very easy to remove. 0.10mm brass was a bit better and 0.20mm brass actually felt best because it welded with same current level as 0.10mm copper and material itself and welds felt like 3x stronger. Nickel welded nice at 20j and as you can see almost impossible to rip off.

I also tested 0.20mm copper but that needed 150j to make any kind of weld and at that energy level it almost burned through the olfa blade so that might not be good for negative end of 18650 cell.. And after just a few welds you can feel the energy in the warm cables..

This test was made for my reference to understand and choose right energy levels for different materials in the future but I thought someone could find it useful as well. Next I am planning to nickelplate some copper but since brass seemed to work so well I am not sure if electroplating is worth the effort

Having a slot between the electrodes would improve the welds, and reduce the amount of energy to accomplish a weld, so I am impressed that this is working on copper and brass even without a slot.

 

[Jmac1775]

Any updates on this? I am currently waiting on my spot welder to arrive. I have a few small packs i plan on building to learn with the ultimate goal of building a 300 amp capable battery for my stealth clone.

 

[spinningmagnets]

There has been a recent development of using a "sandwich" of nickel and copper. Copper and brass sheet are both cheaper and more readily available locally than ordering and waiting for pure nickel (8mm wide, 0.15mm thick, or 0.20mm).

For the SERIES connections (high amps), if you put a copper strip on the cell, and overlay a nickel strip, it welds much easier using the common welders (Riba, Malectrics, kWeld, etc). Since the copper strip carries the current, then the nickel strip does not need to be thick. 0.10mm-0.15mm nickel is fine for this.

The parallel connections can be pure nickel of any thickness, and even nickel-plated steel. Expected parallel currents under all conditions are very low.

"Spot Welding Copper Strips to 18650 Cells"
https://endless-sphere.com/forums/viewtopic.php?f=14&t=84680&start=200#p1470956

Nickel strips:
8mm X 0.15mm = 1.2mm squared = 16 ga round wire
8mm X 0.20mm = 1.6mm squared = 15 ga

Copper sheet:
0.15mm___6-mil__34 ga
0.20mm___8-mil__32 ga (recommended for initial experiments)
0.25mm__10-mil__30 ga

 

[Robert Larrison]

This is the missing link I was looking for trying to nickel plate some copper for pack building

A way to electroless plate copper without the $$$ price of Caswell/Alibaba for spotwelding

https://www.instructables.com/Nickel-Plating-Copper-Without-Electricity-or-Activ/

Do A Lot Of Safety Planning/Research On Nickel Chloride
I set up a 6" in line fan on a custom exhaust