Is there an equation for Battery interconnects

[Nashton]

Hey guys, I don't post much, but I am always browsing the posts here on ES

I am working on a new battery project. I am spot welding copper to 21700 batteries to make a 4s13p modules.
The dimensions of this copper from the positive of cell group 1 to the negative of cell group 2 is 0.05mm thick, 25mm long, and 265mm wide.
Two layers of this 0.05 copper is being used. The total cross sectional area for the direction of current flow is 0.1mm x 265 mm = 26.5mm^2

So my Question: Is there an equation, based on the material properties, cross sectional area, and temperature that gives a relatively accurate value of current carrying capacity for cell to cell interconnects a with a certain temperature rise? I've been looking online and I don't really see anything useful. Also all the charts and table give something slightly different.

 

[amberwolf]

There's probably some info in the copper/nickel sandwich threads, but it's been a while since I read those, so not sure.

Keep in mind that the actual temperature you have (and current capacity that is based on that) depends on where the conductors are: If they are in the open with good cooling airflow, it'll be different than if they are covered up in layers of pack insulation and a housing that prevents any airflow from reaching anything.

In the latter case the only significant heatsink will be the cells themselves, which means heating those up more...and if they are generating more heat than the interconnects, then the interconnects have nowhere to send their heat and instead will heat up more from that coming from the cells.


In general, if you assume it's "pure" copper, you can use the conductivity per square mm of crosssection area of the copper itself for the interconnect carrying ability.

What the interconnect can get from the cell itself depends on how much of the interconnect actually contacts the cell and what the resistance of the connection is.

You can then use the resistance of this total crossection area times the length of the interconnect for the total interconnect resistance of it, and calculate the watts of power that will heat the interconnect based on the current expected to flow thru it.

Then you can use that heat along with whatever heat dissipation you expect possible over a set amount of time to calculate the temperature rise expected on that interconnect just from current. To be accurate you would need to add in whatever heat is expected from the cells connected to it, or subtract whatever heat it would send to them, and also recalculate the resistance of the interconnect at the higher temperature, if there is any significant difference in resistance from the starting temperature.

I don't have any of the math for doing these things (I know how they work but my brain doesn't work with numbers this way) but it should be "standard" equations.

 

[rafalg]

some electrician forums say 1mm2 of copper wire is good for 6-10A of current, so you should be able to handle 250A without problem (I think in case of such short connectors even more current can be handled without any heating).
For 13P battery it would mean about 20 amps of discharge current per single cell so at this point i'd rather be worried if these cells are able to withstand that.

 

[TDA78]

I use a trace ampacity calculator for high current printed circuit board design. You may find it useful. The source code could probably be modified for nickel or steel as well.

 

[Nashton]

some electrician forums say 1mm2 of copper wire is good for 6-10A of current, so you should be able to handle 250A without problem (I think in case of such short connectors even more current can be handled without any heating).
For 13P battery it would mean about 20 amps of discharge current per single cell so at this point i'd rather be worried if these cells are able to withstand that.

they are molicel p42a i am using. have any details you can share on thoes cells and its capability? I have the data sheet for them but if you have experience with them id love to know. I should note, I don't plan to pull more than 150~ish amps for 30sec

 

[Nashton]

I use a trace ampacity calculator for high current printed circuit board design. You may find it useful. The source code could probably be modified for nickel or steel as well.

this is quite helpful, ill dig into the and see how it works

 

[Nashton]

There's probably some info in the copper/nickel sandwich threads, but it's been a while since I read those, so not sure.

Keep in mind that the actual temperature you have (and current capacity that is based on that) depends on where the conductors are: If they are in the open with good cooling airflow, it'll be different than if they are covered up in layers of pack insulation and a housing that prevents any airflow from reaching anything.

In the latter case the only significant heatsink will be the cells themselves, which means heating those up more...and if they are generating more heat than the interconnects, then the interconnects have nowhere to send their heat and instead will heat up more from that coming from the cells.


In general, if you assume it's "pure" copper, you can use the conductivity per square mm of crosssection area of the copper itself for the interconnect carrying ability.

What the interconnect can get from the cell itself depends on how much of the interconnect actually contacts the cell and what the resistance of the connection is.

You can then use the resistance of this total crossection area times the length of the interconnect for the total interconnect resistance of it, and calculate the watts of power that will heat the interconnect based on the current expected to flow thru it.

Then you can use that heat along with whatever heat dissipation you expect possible over a set amount of time to calculate the temperature rise expected on that interconnect just from current. To be accurate you would need to add in whatever heat is expected from the cells connected to it, or subtract whatever heat it would send to them, and also recalculate the resistance of the interconnect at the higher temperature, if there is any significant difference in resistance from the starting temperature.

I don't have any of the math for doing these things (I know how they work but my brain doesn't work with numbers this way) but it should be "standard" equations.

Thanks for breaking this down. you actually pointed out an oversight i made. I never thought about the resistance created from the weld which could be variable from weld to weld if different pressures and and energy is used to create the weld.

 

[rafalg]

they are molicel p42a i am using. have any details you can share on thoes cells and its capability? I have the data sheet for them but if you have experience with them id love to know. I should note, I don't plan to pull more than 150~ish amps for 30sec

I don't have any experience with these and never needed such big currents, but it looks like they're pretty capable.

Test of Molicel INR21700-P42A 4200mAh (Gray)