Minimum conduction voltage between aluminium plates exposed to air?

[cmb]

Hi, just thinking about a possible project and it may involve bolting aluminium plates together and putting a very small voltage across them.

If I take two plates of aluminium, leave them to oxidise in air (we all know Al forms a surface layer) then is there a minimum voltage that will flow between the plates?

I have read various accounts that say this layer is broken as an electric field forms across it, but does that mean there is a minimum that does this? Or is the aluminium oxide metallic and will conduct?

 

[Rive]

is the aluminium oxide metallic and will conduct?

By default, it does not conduct. Some PCBs are even based on aluminium plates and (chemically strengthened) AlOx layer as insulator, making some hilariously high thermal conductivity for power components.

When naturally thin AlOx layer gets mechanical damage (by another Al plate, for example) then this layer breaks and current can flow. With the usual disorganized bolting methods it is almost sure that such plate pile will conduct, but it is hard to predict the exact resistance, especially for long term.

A side note: if Al and bolting (possible non-Al metal) and small currents then you have to be careful with possible electrochemical corrosion.

 

[cmb]

OK, but if I put two AL plates together, then they *do* conduct when I put a battery voltage across it. What's the smallest voltage that *doesn't* conduct in that situation? Is it 0.000001V, or less, because ANY conduction means the surfaces have become 'damaged'?

 

[Rive]

ANY conduction means the surfaces have become 'damaged'?

That's the practical meaning, yes. If the AlOx layer is made then there is usually a breakdown voltage specified (with plenty of text in the footnote about mechanical damage as exclusion from any guarantee). But if it is just natural, then this layer just makes things uncertain.

 

[cmb]

I was wondering if anyone could add to this?

I have found a couple of papers on this, one I seem to have lost which appeared authoritative, but this one agreed with the basic figures

What I am trying to get at is if I go to the metal store, buy two plates of aluminium, clamp them together firmly and attempt to pass 1000A through them, I'm not going to expect to find that they are insulated from each other.

My question is what resistance will I find if, say, I turn up a 1000A power supply, across them, to the max and I have created a 100cm^2 contact area between the two pieces?

The graph above suggests the resistance will be 10^-7 Ohm, after exposure to air, so the voltage between the pieces will be 1mV. Is that correct?

 

[anorlunda]

What I am trying to get at is if I go to the metal store, buy two plates of aluminium, clamp them together firmly and attempt to pass 1000A through them, I'm not going to expect to find that they are insulated from each other.

My question is what resistance will I find if, say, I turn up a 1000A power supply, across them, to the max and I have created a 100cm^2 contact area between the two pieces?

Your question oversimplifies. In real world cases, almost any handling and clamping will scratch or scrape away the oxide in some places creating a low resistance points of contact.

If you want to dig deeper in your study, this NASA report on electrical bonding is very good.

https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19980201283.pdf

 

[cmb]

Thanks but it doesn't appear to tell me anything. It talks about treated aluminum. I'm talking about untreated aluminium.

I am asking what the likely resistance is that I find if I clamp two pieces of aluminum together and try to put 1000A through them.

 

[Rive]

I am asking what the likely resistance is that I find if I clamp two pieces of aluminum together and try to put 1000A through them.

Al is actually one of the most inconvenient materials if it is about 'just clamp them and see'. To get anything really predictable that 'clamping' should be made following some industrial rules/requirements: anything else will contain a big dose of luck.
Sorry.

 

[cmb]

I've never ever known two pieces of aluminium to present electrical isolation when held together. I never got the impression it was 'luck' that they conducted. The graph I have shown indicates 60 Angstroms of thickness builds in atmosphere, and that a current will tunnel through that.

The sorts of dielectric layers that may be put in place in certain applications I have to presume are deliberately formed from thicker layers. I notice on that graph once above 120 Angstroms then it starts flattening out, so I presume there is a thickness that will be insulating, but doesn't seem to form from air.

I thought this would be a simple question for a DIY forum; take two pieces of metal, what is the surface resistance to expect?

 

[Rive]

I thought this would be a simple question for a DIY forum; take two pieces of metal, what is the surface resistance to expect?

To avoid presence of the 'luck' you should follow relevant factory requirements: apply proper surface preparations (adequate surface clean, lubricant) and mechanical load (torque on the bolt according to requirements for the given system and bolt size). As I've just checked some random resources about Al busbar systems it seems that a good contact resistance would be <250uOhm/cm2.

Good luck.