Spot Welding Copper Strips to 18650 Battery Cells

[spinningmagnets]

[I've been investigating brass as a paralleling strip, but some of the brass info I found was also alongside copper spot-welding, so...I'm cross-posting this here too]

Page 298, "Copper and copper alloys" By Davis

Resistance welding of specific alloy groups
Copper and copper alloys having electrical conductivity higher than about 30% IACS (see table 18) are the least well-suited for resistance-spot, projection, or seam welding, mainly because of severe electrode pick-up. Thin copper stock can be welded using electrodes faced with RWMA class-13 (tungsten), or class-14 (molybdenum), but surface appearance is poor, and frequent electrode maintenance is required. A tinned coating on wire or sheet is helpful in welding copper.

Low and high-zinc brasses
The low-zinc brasses are difficult to weld, although easier than copper, and are subject to electrode pickup. Welds made in these brasses may lack strength, principally because of comparatively high electrical conductivity (32% to 56% IACS). The high-zinc brasses have an electrical conductivity of 27% to 28% IACS, and can be both spot and projection-welded over a wide range of conditions. Electrode pickup can be a problem, unless weld time, welding current, and electrode force are properly selected.

Excessive electrode pickup and blowthrough of the weld may occur when long weld times, high energy input, and low electrode forces are used. Yellow brasses (C268 and C270) are less susceptible to electrode pickup than cartridge brass except when long weld times and high energy input are used. Electrode force should be sufficient to prevent arcing or expulsion of molten metal, to which these alloys are subject because of their 30% to 40% Zn content, which boils at 905C (1665F). As shown in table 19, the recommended electrode force, when using electrodes having a face diameter of 4.8mm (3/16-inch) is approximately 1.8 kN (400-lbf)...

...electrode forces lower than those needed for welding low-carbon steel are used. But extremely low forces, which can cause electrode pickup and weak welds, should be avoided. Low electrode force can also cause high-zinc alloys to flash or burn through

Many of the sunstone spot-welding youtubes are short on raw info, but one of them stated their machine was using 5-lbs of force before it would pulse on copper...

Selection of process
Weldability of the work metal often determines which process should be used for a given application. Some of the coppers and copper alloys can be spot-welded, but not seam-welded because of high conductivity, and not projection welded because of low compressive strength of the projections at elevated temperature...spot and seam welds can be made in work metal as thin as 0.025mm (0.001-inch). Spot-welding of metal as thick as 3.2mm (0.125-inch) has been reported for copper alloys. Projection welding is best suited for work thicker than 0.50mm (0.020-inch).

Electrodes
The current used for resistance welding of copper alloys is much higher than used for welding low-carbon steel. Therefore, the electrode must have high electrical conductivity to minimize heat build-up. The Resistance Welding Manufacturers Association (RWMA) Class-1 electrode materials (typically tungsten or molybdenum alloys, containing copper and cadmium) are sometimes used for welding copper and high-conductivity brass. Class-2 materials containing copper and chromium, are used on low-conductivity brass, bronze, and copper-nickel alloys. Electrodes must be sufficiently liquid-cooled to minimize sticking to the work material, and to prolong their life. Tip contours must be carefully prepared, and the electrodes must be properly aligned.

The use of projection welding frequently can increase the quality of joints in high-conductivity alloys because current can be concentrated where needed. Distortion and electrode pickup are minimized because the electrode contacts a large area of the work metal. Projection welding may be preferred when the components are self-locating, or to simplify fixturing, or improve dimensional accuracy

 

[spinningmagnets]

There are two brasses that look to be useful to us as paralleling strips. Red brass has more copper in it, (so it has more conductivity and less resistance) and it is likely to respond to spot-welding methods and tools in a way that is similar to copper.

Which tungsten electrodes should work best for copper? I am ready to buy. If tungsten (or molybdenum, or "X") turn out to be unworkable, I will gladly send them on to anyone who is experimenting with copper spot-welding for further experiments...(for the cost of postage)

edit: IF...you are interested in 2% Thoriated Tungsten rods as spot-welding probes, here is some info on the diameters of the available rods.

inches____inches_____mm
.040______3/64______1.0mm
.062______1/16______1.6mm
.093______3/32______2.4mm
.125______1/8_______3.2mm

Fechter has suggested that inserting a short tip of tungsten into a thicker copper rod would aid in conductivity and also cooling the tungsten tip, therefore I "think" a useful first experiment would be with the thinner tungsten rods. For the thinnest rods, perhaps cut a lengthwise slice in the copper rod tip, and then clamp the tungsten section in the groove? (rather than drill a 1.0mm hole)

.040
http://www.ebay.com/itm/10-pcs-040-7-1-0-175mm-RED-WT20-2-Thoriated-Tungsten-TIG-Electrode-/231646210749?hash=item35ef30aabd:g:h~oAAOSw3ydVxaew

1/16
http://www.ebay.com/itm/5-PCS-of-1-16-7-RED-WT20-2-Thoriated-Tungsten-Welding-TIG-Electrodes/231655653682?_trksid=p2047675.c100005.m1851&_trkparms=aid%3D222007%26algo%3DSIC.MBE%26ao%3D2%26asc%3D38530%26meid%3Dbee92e3e882f4604b498c85cfa2a07ca%26pid%3D100005%26rk%3D2%26rkt%3D6%26sd%3D171967679595

Sampler, 4 pieces, different diameters
http://www.ebay.com/itm/8-PCS-1-8-3-32-1-16-040-7-RED-WT20-2-Thoriated-Tungsten-TIG-Electrodes-/331629382540?hash=item4d36a6cb8c:g:1OUAAOSwgQ9Vzy4U

 

[spinningmagnets]

I owe a great debt of gratitude to ES member agniusm, for the time that he posted pics of his copper series bars that had been nickel-plated. I believe he had them sent out to be plated, and that sounded expensive and time-consuming. The actual plating process only uses a very tiny volume of nickel, so the price of nickel would no longer be a concern (projected to rise sharply in 2018), however...I assumed that home-plating of nickel would be a PITA. After some research, I now believe that home nickel-plating would not be difficult. It looks like it is easy, cheap, and fast.

Pure nickel strips = high resistance, very corrosion-resistant, easy to spot-weld, expensive in the future.

Pure copper = great conductivity, corrodes easily, hard to spot-weld, cheap and readily available plus should remain cheap

As you may know, I have also been investigating the use of brass as a conducting material, which has qualities halfway between nickel and copper...with its main benefit being it is cheap and available, but still easier to spot-weld than pure copper. When looking at the differing compositions of copper alloys, some are rated as "excellent" at spot-welding, but conductivity is lower than pure copper...so I wondered why, and looked for trends in their composition.

If an alloy is 98% copper, it has very good conductivity regardless of additives, but...its "spot weldability" goes up to excellent if the additives include a trace of Nickel, Iron, or Chromium (or a mix of any of these three). Hardly anyone stocks these obscure alloys, and even if you commissioned a foundry to mix up a batch, and then have it rolled into sheet-metal...it's expensive and time consuming. But...lets get back to home-plating copper...

If a sheet-metal has some iron, nickel, or chrome in it...the resistance means spot-welding produces high temps in a very tiny and localized area, meaning the "spot" welds rapidly, without heating the interior of the 18650 cell. The series connections were never an issue, anyone with a salvaged transformer can make that connection with no fear of damaging the cells, even if using pure copper. For me? the issue has always been making the spot-welded connection from the paralleling strips to the cell ends (without using pure nickel). So...what common material is corrosion-resistant, and also has some nickel, iron, and chrome inside it in order to make spot-welding easy?...stainless steel

If the contact area is large enough between the cell-end and the paralleling strips, and also between the paralleling strips and the series flat-bars (higher current there), then voltage drop shouldn't be too bad if the cores of these materials are pure copper with only a thin plating of nickel or stainless steel.

"DIY plating nickel onto copper"
https://www.youtube.com/watch?v=Q8Xo43sfLgY

 

[kdog]

RTL had a crack at it, the results were not so consistent and didn't help much ( sorry RTL, you did a great job!) we found out later that nickel plating can be a hinderance bc that nickel vaporises so much earlier than the copper, and it blows the nuget out with considerable force.
I think really thick professionally done plating would help but not so much home plating. I don't really have much experience though- just a few attempts.
(Anyone with any interest please read my post on brass spot weld thread. Trying to gauge interest in how many would want a good nickel alternative)
K

 

[spinningmagnets]

Cross-post from the brass strip thread, by ES member "volts"

https://endless-sphere.com/forums/viewtopic.php?f=14&t=85059&p=1248076#p1248048

I worked out how to get copper to weld onto battery's

1. Clean everything with vinegar especially the welding surfaces
2. Flux the area with a flux gel
3. Cut a small piece of solder, I like using silver solder you can use what you want
4. Put that small piece of solder on the welding area
5. Place the copper or metal piece on top of the solder
6. Press the spot welder very lightly onto the metal weld area and Weld

The trick is to make sure not to press down to hard as you will push the solder out, this method is soldering using the spot welder takes practice but works

 

[Hummina Shadeeba]

Anyone with success copper welding? Seems hit or miss. A friend has what he describes as a 3000watt welder we're going to try with 20mil copper, tungsten tips, and slotting. Doesn't sound promising.
If fails I guess I'll try the solder/welding approach above

 

[kdog]

Imperial! Gawd! How do you work in such ways.
Any way if I did my conversion correctly 20mils equals 0.5mm which is hell thick.
Try 0.1 mm as a starter ( 4mils)- even 0.15mm (6mils) is too hard.
Try copper electrodes if you have no luck with the tungsten. It can be done it just takes practice and working out what you can/can't do. But .5mm copper isn't going to happen- sorry
K

 

[Hummina Shadeeba]

Yea I figured it's too thick and will get some other stuff. Is there an equivalent nickel strip thickness that a welder can do to do 10mil copper...if the welder I'll use can do 15mil nickel at its thickest and the copper is a 1/4 the resistance it doesn't sound good

 

[kdog]

Get some 0.1mm copper, cheap easy to buy. Try it out. I use about the same power for it as I would .2mm nickel. It's weak so getting just the right amount of power without destroying it is the key. Given that it carries the equivalent amps as .4mm nickel, it's pretty good for high current builds.

 

[Hummina Shadeeba]

will do that! thanks

 

[kdog]

Also, laying down nickel first then welding copper on top of that helps. It also protects the cell from the inevitable blow out. the lower resistance of the nickel helps pull current down rather than across the copper making a better weld- If you get my drift...

 

[Hummina Shadeeba]

i don't get your drift on the pulling current down vs across the copper. Id think the current would travel much easier across the copper than the nickel but whatever not important and what Im more interested in is your using nickel and then the copper! so spot weld the nickel tab on and then do the copper right on top? that seems the best of both worlds.

any suggestion on a good quick place to find .1mm copper. I'm looking but it's seeming rare so thin. want to do it this weekend


here:
https://www.onlinemetals.com/merchant.cfm?pid=4541&step=4&showunits=inches&id=129&top_cat=87
.1mm sheet

 

[kdog]

eBay aliexpress have it by the truck load. Just search copper foil/sheet.
Nickel for P group connects then copper on top for series connects.
K

 

[spinningmagnets]

That may end up being the next best step, spot-weld thick nickel strips as usual, and then bond copper over the series part of the connections. current takes the path of least resistance, so the resistance of the nickel will only cause waste-heat and wasted watts (voltage drop) for the tiny distance between the cell-tip and the series bars.

If copper is overlaid on the series strips and spot-welded...before...the nickel is spot-welded onto the cells, then there will be no chance of any damage to the cells if you are using a huge MOT welder to bond the copper to the nickel.

Now that I think about it, the copper mass can even act as a heat-sink for whatever heat is generated at the cell tips.

C41410 is still the ideal parallel strip material (IMHO), but its only special-order, so...expensive.

 

[Hummina Shadeeba]

I imagine the tiny points of contact between the cells and where it's welded to the nickel are the weakest link and wonder how much the copper on top could help resistance.

my friend is going to have an even stronger spot welder this weekend and since I can't resist going the whole hog I'll try the copper to the cells first. I'll tell you how it goes with the .1mm I'll try welding some copper strips to each other fand then to the cell too.

 

[spinningmagnets]

I am now wondering for the parallel strips...whether copper or brass...if two long and thin strips would work better, instead of just having a slit over the cell tip?

4S / 2P, the orange strips are brass parallel strips (or copper?), the red is the copper series bars that have to carry the full pack current. using two thin brass strips is only to help the spot-welding of brass onto the cell-tips, the parallel connections don't carry high current...

 

[Hummina Shadeeba]

I did retarded math and thought .1mm was much thinner than it is.

http://basiccopper.com/thicknessguide.html
thickness guide and one of many places to get the thicknesses desired.
I'll use the 2mill (.002 inches) i already have. = .05mm

 

[kdog]

HS the weld isn't the only point of contact so not the only current path. if you get a nice flat strip and a few good welds it actually has a large SA in contact.
0.05 is going to be super weak but easier to weld obviously.
SM two long strips works well in terms of the slot effect, but it is sooo fiddly its not worth the effort.
I still cant find a supplier that will help me with some samples of various copper alloys... they're out there, they've got the gear, but they're not interested in small fry
k

 

[vex_zg]

Like this I mean:




That way you would need (much?) less current to weld and might be able to spot weld copper strip using the simple spot welders if you insert tungsten tips in the electrodes.

Tried this - one electrode on battery casing and the other on the copper strip. This is likely to cause blowout in the battery casing.

200Ws (1F@20V) is in my experience not enough to weld 0.1mm copper. Tried with copper, molybdenum, tungsten, 30%-70% Tungsten-Copper.

The issue is that with voltages higher than some 14-15V blowouts become inevitable.

I will try to go to 2F/400Ws and see if it helps.

 

[litespeed]

Copper is tuff. Can't wait to see if you can figure this out.

Tom

 

[vex_zg]

Copper is tuff. Can't wait to see if you can figure this out.

Tom

Yep. Even with 200Ws it is difficult to cope with transient voltages, and evaporating metal from tabs and electodes.
I wonder what will I see on the scope when I increase the energy to 400Ws. Now waiting for another 1F capacitor which I bought.

 

[litespeed]

My welder is 800 WS with 3 farrad of true 25 volt computer caps. Wonder why yours is less? My voltage is 23.xx Something, I forget right now cause I just popped some pain pills......

Tom

 

[vex_zg]

My welder is 800 WS with 3 farrad of true 25 volt computer caps. Wonder why yours is less? My voltage is 23.xx Something, I forget right now cause I just popped some pain pills......

Tom

I love pain pills

I built my own welder, of course didn't know how the project would turn out so I didn't invest in multiple caps.

200Ws gives me borderline 0.3mm nickel weldability but not 0.1mm copper.

With your 800Ws I am unsure why you are unable to weld, the energy should be sufficient.

 

[kdog]

I figure there is so sort of exponential power requirement with respect to thickness. There is both the increased conductivity, and increased thermal wicking/mass.
I have found .2 mm to be a ceiling of sorts with the power levels the can can handle. over this thickness and it gets a bit sketchy with heat and weak welds etc.
with a slot I get reliable welds with .3mm in nickel/brass but not so much as a tack without the slot (at this thickness). I seem to need more power for .3Ni than .1Cu. I'm guessing the thermal mass is the difference.
For me .2 is the go too, except in pure copper where it's .1mm.

 

[vex_zg]

I figure there is so sort of exponential power requirement with respect to thickness. There is both the increased conductivity, and increased thermal wicking/mass.
I have found .2 mm to be a ceiling of sorts with the power levels the can can handle. over this thickness and it gets a bit sketchy with heat and weak welds etc.
with a slot I get reliable welds with .3mm in nickel/brass but not so much as a tack without the slot (at this thickness). I seem to need more power for .3Ni than .1Cu. I'm guessing the thermal mass is the difference.
For me .2 is the go too, except in pure copper where it's .1mm.

kdog - can you tell a bit more about how you are welding copper

- which welder you are using
- energy
- voltage
- which electrodes you are using
- weld timing (pulse lengths)