Creating Coils from Copper Rod: Bending Process

[trini]

I'm working on a project, and I need to make coils out of 5/8" diameter copper rod, ideally with center line radii ranging from 3 - 10 cm, and with turn heights of 5 cm. Most metal shops near me don't have the capacity to do this, because there rolling machines are based off of cold bending.

I have formulated a design with which to make something close to these coils in shape, though it remains to be seen whether or not they would then be able to handle my currents. I plan to feed the rod through a flame (i would prefer an induction heater but cost is against me) heat the rod up to about 95% of its melting point (about 1000 C), then immediately feed the hot end to a cylinder to which I will attach a clamp to hold the end in place. I will use a gear system to control the bending ( when the cylinder rotates once, it moves 5 cm horizontally).

A problem I can see is that the copper will oxidise, and the granular structure of the rod (which is helically planed) will be altered, which may affect conductive properties. I would like to hear your thoughts on this as well as any other practical alternatives which you may have to my method.

Also I wasn't sure which forum to put this in, please move it if necessary.

 

[Q_Goest]

Hi trini,
I wouldn't bother with heating the copper like that. I'm assuming the copper rod is being used just for its electrical properties because if you heat it to close to the melting point and let it air cool, the yield strength will go down to about 6 ksi. And yes, it will scale badly if you do that.

Here's a suggestion. Get the copper bar fully annealed if you can. If you can't, just anneal it yourself by either heating to red hot and air anneal or preferably have it done in a furnace using an inert gas so no scale gets on it. If you want to do it by annealing using a torch, then dip it or wipe it in a weak sulfuric acid solution (4% to 15% as I recall) to remove scale. This acid trick works great, leaves it as shiny as a new penny.

To roll it, just get a cylindrical bar of the diameter you want to roll it around, then simply wrap it. I can imagine doing this on a low speed lathe for example. You might even be able to do it by hand if the copper bar is fully annealed. If you don't feel comfortable doing it yourself, there have to be shops that do this regularly. I'm just not familiar enough with what those shops are, though if it were me I'd look for a large machining and manufacturing place.

 

[Bob S]

Is this project for a dc or ac application? If the required pulse risetime is less than a second, you should calculate the skindepth of the Fourier-transformed current in the copper to ensure full current penetration. If you need more conductor surface area to reduce ac losses, then consider a 7-conductor twisted copper conductor. See
http://www.copperinfo.co.uk/busbars/pub22-copper-for-busbars/sec4.htm
Is the purpose of the coil to get a magnetic field, or for ac inductance, or for dissipating resistive heating? The resistance of copper increases with temperature, so the I-squared R losses increase with temperature. Another thought is to use a square 1/2" or 9/16" copper conductor, or maybe 1/2" by 3/4" rectangular copper busbar. This might decrease the bending forces required to make the coil.
There are hollow square and rectangular copper conductors available in many sizes if water cooling is required. Hollow rectangular copper busbar is often used in the manufacture of large magnets for research, and bending this into special coil shapes is routine. I suspect this busbar is delivered full soft. How long a conductor do you need?

 

[trini]

This project requires dc current, so skin depth is not a factor. also the purpose of the application is to create a rare Earth permanent magnet, so i need a strong magnetic field. The required current is going to be in the range of a 16kA pulse, which is why i am uncertain whether or not a hollow rod will be able to handle this current, as well as wheter or not a hollow rod will produce a uniform field. my total conductor length is just over 10m, so I will have to weld together multiple rods to get my desired shape.

 

[trini]

hey guys, just want to know, what would be the strongest way to join copper rods together end to end? would welding be strongest and if so, what kind of weld should i be using?

 

[Mike_In_Plano]

Surprisingly, I had a problem like this about 3 weeks ago using AWG4 wire. We managed that with a lathe, but I'm certain from that experience that you cannot use this technique with 5/8". At least with any "normal" lathe. The required torque increases to the square of the thickness, and we barely made it.

You mentioned that the application was a pulsed application, so I'd look into the nature of the pulse and whether you can absorb the loss in a smaller conductor. Say loose 400 watts, but only do it for 20ms. This may turn out to be a good solution if you don't have to do it often.

Also, you may be able to use copper strips as laminates. Simply form one, and form the next over it, etc... We had a welder that used this technique to carry thousands of amps through a pliant connection. You can also use wire, but keep in mind that square wire likes to twist and form a trapezoidal shape if you bend it tightly.

As for making the part out of solid pieces, you may want to try brazing with silver solder. It's easy torque with and makes strong joints.

Just a word of caution. When I was in college, our EE department was conducting experiments which required large pulsed currents (100's of kA). The forces involved could get very large. An associated university had a bus bar come loose and crack the neighboring wall. We used a lot of solid wood, thick (1/2") Plexiglas, and brass screws to build shields and thankfully nothing came loose. If you're dealing with a high-energy magnetizer, you might do well to keep it contained.

 

[marcusl]

This project requires dc current, so skin depth is not a factor. also the purpose of the application is to create a rare Earth permanent magnet, so i need a strong magnetic field. The required current is going to be in the range of a 16kA pulse, which is why i am uncertain whether or not a hollow rod will be able to handle this current, as well as wheter or not a hollow rod will produce a uniform field. my total conductor length is just over 10m, so I will have to weld together multiple rods to get my desired shape.

A pulsed current is, by definition, not DC. You definitely should do the skin depth calculation. If your pulse is short, no current will penetrate to the center of your 5/8" diameter wire and you might as well use tubing. You'll then need to calculate whether your tube is safe at 16 kA.