Optimal thread style?

[beretta24]

I am trying to pick the best connection for some test equipment. I am running two copper axial leads (1.25" OD copper bar cut in half) down the center of a shaft. At both ends I will have a flat machined into the bars to accept a copper radial lead. The radial lead will thread into the axial leads through a hole in the shaft. When all is said in done I will have .373" of material in the axial lead to drill and tap a hole. My question is what is the best thread to use?

I ran through some numbers from the Machinery's Handbook for 1/2-13 UNC 2B threads. It suggests the desired length of engagement is .335". My concern is that in all reality I will not get the desired engagement due to threads at the top and bottom of the hole being incomplete. The applied load to the connection is 1585 lbf. I need to keep the preload greater than this in order to maintain contact between two surfaces to ensure electrical conductivity.

Should I be looking outside of UNC series threads given the small length of engagement? I'm not very familiar with any metric thread, would they or anther style be a better choice? This connection will only be be reused a dozen or so times so fatigue isn't a big concern. Thanks.

Brian

 

[deanc]

http://www.boltcouncil.org/2nd Edition Guide.pdf

http://www.hexagon.de/dose/dose1e.htm

Have fun

 

[EdStainless]

Don't tap the copper. It isn't strong enought and if the connection gets warm it will creep and relax more. Use a through hole and a SiBronze bolt and nut.

= = = = = = = = = = = = = = = = = = = =
Rust never sleeps
Neither should your protection

http://www.trent-tube.com/contact/Tech_Assist.cfm

 

[gfbotha]

Hard or half hard copper (hardened via cold work) might be OK but only if your calcs show the stress levels to be acceptable AND the parts are not be subjected to substantial heating.

But your margin seems a bit slim! - and I don't think thread type is gonna save you this time. I guess going slightly finer with the thread (say M12 x 1.5) might only help a tiny bit - but beyond a certain point you'l start loosing again...

 

[beretta24]

EdStainless,

I don't have experience with copper as far as creepage goes. I reviewed the mechanical properties of the copper being used for this application and there is a sf > 2 with respect to the yield strength of the material after the desired preload is achieved.

I have also incorporated a flat where the two pieces come together in order to increase the contact area.

Could there still be an issue with creepage?

 

[gfbotha]

OK, so you have a substantial SF (not simply 373/335).

Talking creep, you need to tell us how long at what typical temperatures.

 

[beretta24]

Still working on the temps with the electrical guys. Part of the trouble is there a lot running through the connection, 1000V at 1000A, but only for about 3s so I don't have a good feel since the connection won't reach a ss temp. I will supply more info when I get it. Thanks.

 

[deanc]

Buss bar:

http://www.stormcopper.com/design/index.htm

http://www.patentstorm.us/patents/6040525-description.html

Check the links on this page. Both have design guides.

 

[EdStainless]

You need to silver plate both parts of the connector. The connection will heat more than the conductors. Mass and thermal conductivity will help some, but I am concerned.

= = = = = = = = = = = = = = = = = = = =
Rust never sleeps
Neither should your protection

http://www.trent-tube.com/contact/Tech_Assist.cfm

 

[beretta24]

Thanks for all the info, in the end we will be getting samples and testing the connection in our tensile tester. We will be testing for joint strength and to ensure the contact pressure is maintained between the lead and conductor faces.

I have called out silver plating for the threads. In addition I compared the current density through the connection to the current density elsewhere in the machine's DC path. I expect the current density in the connection to be less than a third what it is elsewhere.

I will try to give you an update after testing, thanks.

Brian

 

[gfbotha]

Hi Brian

Would be interesting to know: how do you plan to heat the joint to the right temperature for only about 3s? Time at actual temperature is important. Or, are you going to isolate the test pieces and actually use electrical current?! In the real application the actual bulk/length of (surrounding) material will influence the heating and subsequent cooling...

Or did you meanwhile established that the heating will be low?

 

[Goahead]

Could you consider Ultrasonic Welding?

Several Ultrasonic Equipment manufacturers (from a Google search) could probably weld the contacts for you.

Regards,

http://www.welding-advisers.com/