[paul.fr]

does the current flow through the whole thickness of the wire, through the middle or mainly along the surface? Would it be different for AC and DC current?

[lyner]

DC travels through the whose cross section of the wire. AC is different; as a consequence of Maxwell's equations (long may his tribe increase) we find that, for a perfect conductor, the current all flows on the surface and with a real metal, there is a 'skin depth' below which no current flows.
This is because of the induced fields around the wire and the , consequently, induced currents back in the wire. (This is connected with the explanation of why a metal, which is a good conductor, reflects electromagnetic waves so well.)
To get the lowest possible RF resistance, for winding inductors for use in radio sets, they used to use Litz wire which was made of many strands of, individually insulated, thin copper wire. It was quite expensive and a real pig to solder to when the enamel was not the self-fluxing type they use today.

[another_someone]

It is certainly different for high frequency AC, which I believe will be more on the surface of the wire; but for DC, it should be through the bulk of the wire (hence why making a wire thicker will reduce its resistance, and improve its current carrying capacity).

OK - 9 minutes late again!

[lyner]

The skin depth at 50Hz is quite a lot, actually, for the low resistivity metals - several  cm  before the current is 1/2 value. They often use multiple strands on big current conductors, and I think that is part of the reason.
The effective AC resistance (and power factor considerations, of course) makes it better to transmit over long distances (point to point) using DC.
It is at RF that it becomes really significant.

[another_someone]

The skin depth at 50Hz is quite a lot, actually, for the low resistivity metals - several  cm  before the current is 1/2 value. They often use multiple strands on big current conductors, and I think that is part of the reason.
The effective AC resistance (and power factor considerations, of course) makes it better to transmit over long distances (point to point) using DC.
It is at RF that it becomes really significant.

From what I have heard, multi-strand wire has little effect on the skin effect unless the strands are separately insulated.  As I understand it, the major benefit to multi-stranded wire (where not separately insulated) is mechanical rather than electrical (i.e. improved flexibility; resistance to cracking across the whole wire; and breaking a single strand will continue to allow other electrical paths to continue carrying current with little impact on overall performance).

What I have heard they do sometimes do with big conductors carrying AC is to have an inner core of high strength steel which may not be an ideal resistance, and have a lower resistance metal on the outside which may not carry much mechanical strength.

Ofcourse, when one is talking about RF, that is why coax cables are used, since at those frequencies, anything other than the thin skin has no meaning to the actual resistance of the wire.

[lyner]

I would agree with the mechanical advantages of 'flex' and also that (overhead) cables are composite, for strength and conductivity. I don't know the maximum thickness of conductor, used for high power / high current cables but the thickest I have seen was, perhaps 5cm across. They tend to use higher voltages, of course, where extremely high powers are transmitted so the current would always be kept to a reasonable value.
And - yes - you would need to insulate the strands for my point to be relevant. That is done, in the case of Litz wire.
There is also an issue with alternating current where superconductors are used, I believe. Anyone else know about that?

[syhprum]

What is really meant by 'current flowing down a wire' I always understood the flow of energy was due to the electromagnetic field guided by the conductor which dragged along Electrons albeit slowly thru the conductor.

[lyner]

Yes - the wave idea is more precise but it would be a real headache to do your circuit calculations using waves. Kirchoff's laws work very well most of the time - they just refer to amps and volts!
Treating 'current' as a separate entity works very well, in practice.
 
So many of the posts on this forum are obsessed with what is 'really happening'. It's entertaining but it really does get in the way of getting things to work.

The main message about current in a wire is the incredibly low speed at which the conduction electrons actually move, as you say, syphrum.

[techmind]

What I have heard they do sometimes do with big conductors carrying AC is to have an inner core of high strength steel which may not be an ideal resistance, and have a lower resistance metal on the outside which may not carry much mechanical strength.
to the actual resistance of the wire.

I have recently had cause to talk to engineers from the National Grid. You might be interested to know that on 132kV pylons (the smallest of the proper pylon-shaped metal pylons) the cables are rated at about 360 Amps each...

[lyner]

It's not surprising that the starter lead of a car drops a significant voltage when it takes a similar amount of current, when you think how relatively skinny it is. These cables go for miles and miles but can afford to drop a few volts when they start off with 132kV.