1 - 15 of 15 Posts

#### MathisLaurant

· Registered
Joined
·
47 Posts
Discussion Starter · ·
Hey guys,

So I am trying to get a better understanding of how to design bus bar for a 12s12p module using 18650's. i understand that bus bars current carrying capacity is also dependent on temperature rise. but how do you know which cross sectional area to use?

Let's assume that the bus bar is 100mm L x 20 mm W x 5mm thick.
Also, let's say that the cells are connected long way along the 100mm distance with the bus bar in between.
(+)=======(-).
From my reading, my assumption in my gut says i should do 20x5x1.2= 120amp
With 1.2 amp/mm^2 for copper
But this doesn't account for temperature rise
Also, why would we not do 100x5x1.2=600amps? Is it because the current is going to find the shortest path from Pos to Neg? Lastly, is there a larger calculation that can be made that includes temperature rise and thermodynamic effects?

#### remy_martian

· Registered
Joined
·
5,855 Posts
Your connectivity scheme doesn't make any sense, nor does "let's design it for the longest distance".

What are the dimensions of the pack in cm/mm/inches, take your pick?

Number of arrayed cells....x cells by y cells?

#### MathisLaurant

· Registered
Joined
·
47 Posts
Discussion Starter · ·
Front and back shown below
the bus bar in question is the one in red.
everything is done in mm.

#### Attachments

• 443.9 KB Views: 79

#### remy_martian

· Registered
Joined
·
5,855 Posts
I'd go with an empirical approach by building a small test rig if you want to be thorough. You're doing what Tesla does, so using their bus plate thickness should get you close enough, I'd think. There are tools out there, but not worth the time or trouble vs just setting a plate up on a power supply and measuring it.

How were you planning to attach your fusible links and what are you using for those?

#### MathisLaurant

· Registered
Joined
·
47 Posts
Discussion Starter · ·
i am thinking of doing Ribbon strips or wire bonding and laser weld to the terminals.
the design is still pending. my concen is just that one bus bar design at the moment and how to estimate based on the cross sectional are and the thermodynamics

#### remy_martian

· Registered
Joined
·
5,855 Posts
"Thermal design".

"Thermodynamics" comes into play when you undersize it and it blows up.

That aside, empirical is the way I would do it, unless you're an Ansys or Comsol jockey.

#### brian_

· Registered
Joined
·
8,647 Posts
You're doing what Tesla does...
Yes, or more specifically, what Tesla did for the Model S and X. The Model 3 and Y have a different bus plate design, which might be interesting to look at since this is still at the design stage.

Other than that, you two seem to have a handle on this subject and I don't have anything useful to contribute.

#### Frank

· Registered
Joined
·
573 Posts
It seems to me that because current flows in the 'series' direction i.e. up-and-down in the pics you posted, not left-to-right, if the busbar is 100 mm wide then you should use that cross-section area i.e. 100 * 5 mm^2.

#### MathisLaurant

· Registered
Joined
·
47 Posts
Discussion Starter · ·
It seems to me that because current flows in the 'series' direction i.e. up-and-down in the pics you posted, not left-to-right, if the busbar is 100 mm wide then you should use that cross-section area i.e. 100 * 5 mm^2.
so you are saying the direct of current is the the cross section to use in the calculation, Correct?

#### Frank

· Registered
Joined
·
573 Posts
so you are saying the direct of current is the the cross section to use in the calculation, Correct?
Yes. If the top group of cells is "+" and it's connected to the group of "-" cells below it, then the cross section area of the copper between them is what is carrying current (or holes, depending on which model of current flow you use).

Another way to think of it is to replace the busbar with a piece of wire of equivalent area. You may be surprised...

#### Ghey Thunder

· Registered
Joined
·
10 Posts
This is a table a dude on endlessphere made. I think it might not be completely accurate however but more on the concervative going by my remarks highlighted in yellow. I had to convert it from Excel to pdf so it might be difficult to read.

#### Attachments

• 473.4 KB Views: 1,460

#### remy_martian

· Registered
Joined
·
5,855 Posts
This is a table a dude on endlessphere made. I think it might not be completely accurate however but more on the concervative going by my remarks highlighted in yellow. I had to convert it from Excel to pdf so it might be difficult to read.
Mind putting up the Excel as well?

#### Ghey Thunder

· Registered
Joined
·
10 Posts
I am unable to attach excel files through the paper clip attachment button. Any other way of doing it?

#### remy_martian

· Registered
Joined
·
5,855 Posts
You can filehost it and point to the link to the file if nothing else on here works.

Or, just change the file extension to .jpg, post it with the paperclip as a "picture" and we can manually change it back to the excel extension (tell us which one it was). I'm hoping it won't corrupt the file by compressing it as a jpg, of course. I don't expect it to.

#### MathisLaurant

· Registered
Joined
·
47 Posts
Discussion Starter · ·
yeah, the Excel would be extremely helpful!

1 - 15 of 15 Posts