Conductivity of dissimilar metals

[Phaedrus]

So I'm looking into an issue we're running into where a connector is overheating and causing the plastic shroud to burn in some instances, at lower current than it should be doing. Part of this is due to greater sustained load than anticipated during design, and partially due to the factory not being picky about connector vendor or materials.

During my investigation I found that the male connector pin is made of steel, while the female connector pin is aluminum. I'm wondering if the conductivity between these two different metals is lower than would be the case if they were both steel or both aluminum. Also I'm wondering if the conductivity of the connector as a whole would be improved if the pins were eletroplated with copper. Ideally of course we could use something like beryllium copper for both pins, but I don't know if that's cost effective.


Finding information on this is difficult, as googling mostly returns lists of the conductivity of various metals, which is not useful to me.

Basically, what can be done to improve the conductivity and ampacity of a connector without changing pin size or arrangement?

[calexanian]

As I understand it aluminum in any electrical connection is always the first suspect in interface heating. I believe copper cladding or plating does help as theoretically the interface oxide layer of the aluminum does not form as readily due to the bonding process giving you the copper surface to make a cleaner connection. It was very popular in residential wiring some decades ago to use Aluminum and they learned all of this the hard way. Copper itself has issues such as oxidizing and its general softness leads to galling. Typically it is alloyed with something else to give it better properties. Gold it a typical plating agent as is silver.. Tin works very well in many applications, as most pin/sockets used these days are tin plated copper or steel.

Just my thoughts.

[IanB]

Dissimilar metals in electrical contact can be subject to galvanic corrosion if moisture is present. This could produce a layer of insulating oxide material between the mating surfaces and increase the resistance. The amount of moisture present does not have to be visible, it could just be a microscopic layer adsorbed on the surface. Steel and aluminum in contact seems like a bad combination from this perspective.

I think connectors like spade connectors ought to have a strong mechanical force between the mating surfaces to exclude air and moisture and give a close metal to metal interface. Steel is good for this due to its strength and springiness. Aluminum is not good at all. It is weak and it suffers from creep. That's why aluminum wires are bad in screw terminals. They suffer plastic deformation under the screw pressure and become loose over time.

At a guess, I would say aluminum is the bad actor here. I would think steel or brass (or brass plated steel) are the materials of choice.

[scientist]

Aluminum can weld to other aluminum if left in contact for too much time - definitely the aluminum that's causing your problems.

[Phaedrus]

Thanks! I'm suggesting to management that we switch to steel-on-steel and consider using copper electroplate if the cost is not too bad.

[chickenHeadKnob]

@Phaedrus

Suggestion: Take the connector you are using and subject it to excessive heat with soldering iron. See at what temp the plating starts to separate from the steel core. Initially the damage  can be very subtle. When I was a young lad starting out wire-wrapping   high-rel proto boards for telecom and industrial applications I discovered this the hard way. We used to have this practice of fixing a socket into place by soldering two pins to their respective pads, before starting wrapping so the sockets wouldn't fall out.

[PedroDaGr8]

Congrads, you just created a variant on the lasagna cell (I'll explain in a minute) Aluminum metal being a very easily oxidized (in the REDOX sense) metal can cause huge issues. What typically saves it in the fact that the oxide layer makes the underlying aluminum impervious to air. That being said, aluminum is often used as a sacrificial anode on steel buildings because it will preferentially be oxidized. Based upon the table here (top of pg 6) of electrochemical potentials generated, aluminum generates over -1V of potential in contact with stainless steel.

 About the lasagna cell. Basically it is an electrochemical cell comprised of lasagna cooked in a stainless steel pan covered in an aluminum foil.  The lasagna acts as an acidic electrolyte causing the aluminum to rapidly degrade.


So even if you weren't having this problem its not a great idea in general to mix aluminum and steel.  You would likely end up with reliability problems down the road. Especially in humid and/or high salt type environments.

[Phaedrus]

Agreed. Our engineers should have caught this. Note that these pins aren't stainless steel though, only mild steel. The difference is only about -0.2V, which is still significant but not as bad as stainless.

[kaindub]

PedroDaGr8 is almost correct.
Whilst aluminium is a good conductor, it easily forms a highly non conductive aluminium oxide skin when exposed to air.
When I worked in the power industry, aluminum cables and conductor ends were always cleaned (shiny) prior to joining (usually by a crimped or bolted connection) and then treated with Alcoa No 3 paste (If I remember correctly).

However is your problem more basic than this?
If you have a connector system ( I said system) from a manufacturer, and they rate it for a certain current, irrespective of the materials used, the connector should carry the current to it's ratings.
I'd go back to the manufacturer and ask why it does not perform to their specifications. That's what the field sales people are for (and you thought they were there just to annoy you).

Robert

[Phaedrus]

PedroDaGr8 is almost correct.
Whilst aluminium is a good conductor, it easily forms a highly non conductive aluminium oxide skin when exposed to air.
When I worked in the power industry, aluminum cables and conductor ends were always cleaned (shiny) prior to joining (usually by a crimped or bolted connection) and then treated with Alcoa No 3 paste (If I remember correctly).

However is your problem more basic than this?
If you have a connector system ( I said system) from a manufacturer, and they rate it for a certain current, irrespective of the materials used, the connector should carry the current to it's ratings.
I'd go back to the manufacturer and ask why it does not perform to their specifications. That's what the field sales people are for (and you thought they were there just to annoy you).

Robert

The connectors are rated to 28A, and I'd expect a cable with that rating to have overhead to 30A, and handle peaks of up to 32-36A.

Instead, they're failing at 24-28A. The problem we have is that our specification is simply for a cable rated to 28A, with x dimensions using this style connector. Our Taipei team passes that spec to the PSU vendor, who pass that spec to their cable vendor, who then purchase the connectors from their vendors. It's way up the supply chain, and we didn't specify x materials. So it's basically been uncontrolled.

The problem could as well be that the cable vendor or the connector vendor are skimping; but I don't have evidence of that. What I do have is evidence that aluminum sucks for our application. So by nailing down a specification on what metal is to be used, we'll also be able to force the cable and PSU vendors to perform basic QA to make sure the connector is meeting that spec; and they should therefore also catch any other skimping that would prevent the connector from handling a constant 28A. Plus if the problem is deformation and galvanic corrosion (which does seem reasonably likely), that problem will be fixed.


It really is the cable vendor who should be handling this; but because they're handling it extremely slowly, with little transparency, and in Chinese, I did my own investigation.

[miguelvp]

Could it be due to a thermoelectric effect like the Peltier effect?

http://en.wikipedia.org/wiki/Thermoelectric_effect

The Seebeck Coefficient of Steel is:
SS = 2.5269 - 0.0143 * (T) + 0.000008 * (T)^2
where T is the temperature at the time.

Aluminium based on wiki, is 3.5 not sure what the Thomson effect would be because I didn't dig on how the Seebeck Coefficient of Aluminium changes with temperature.

http://en.wikipedia.org/wiki/Seebeck_coefficient

But both being P type you'll figure that not much would happen. And if so, then copper/steel should be worse right?

Maybe it's not due to this at all after all, just saying that dissimilar metal junctions have thermoelectrical properties.

Edit: Aluminum is p-type but steel is no charge but still very close to each other.

[G7PSK]

You will get a rectifier effect with Aluminium and steel so if the system is AC you will have heating from that. With Aluminium you need to abrade the contact surface and then cover with some form of grease.