Hey eastwood, nice build! A lot of information to unpack throughout this thread. Looks like you have been getting lots of quality information and luckily most things have been answered very well in my opinion. As others have said this is your build, and do it your way and what makes you feel most comfortable. Nonetheless ill drop my two cents in the bucket.
I have built about 10 battery packs now so by no means would I claim to be an expert. All 10 batteries were made using molicel p42a cells like you are using. Funny enough my most recent battery build is a 21s17p pack and it is oddly shaped, maybe even more so than yours. Ill get back to that….
Some people would say I am careless or reckless for doing so, but after doing basic internal resistance tests on a few hundred new cells, nothing really extensive as other have proposed, I stopped worrying about it. You have the whole internet coming at you from the most conservative battery builders to what some may consider careless and reckless builders like myself. Don’t get me wrong I use solid construction techniques. I make custom 3d printed cell spaces with a very tight fit, I use 0.2mm thick copper busbars with robust welds, and take a number of other measures to increase durability and reduce vibrations within the battery pack like anyone should. All of these packs have been used strictly for off road applications.
Despite my methods I have yet to have one pack experience failure of any kind. I guess some would call that luck. Point is some people are particularly anxious and paranoid and have fallen subject to extensive and extremely time consuming testing methods which are usually only reserved for packs being built with used cells.
I visited chi battery systems factory last year who have made thousands of custom aftermarket surron batteries…guess what cells they use. P42a. Guess what they don’t do…cell testing. Why? They don’t have time to waste (not that they don’t care about safety, they certainly do), and the cells are validated and brand new. Care to guess how many catastrophic battery failures they have had (not including people running into trees and cars)? Zero, none, zip. They dont test for ir, capacity, none of it. Sure they have testers and the capability to do so, but over time they realized it was a worthless endeavor. Their number one source of problems…the bms. 1 in 100 fail right out of the box. Loss of pack functionality, but nothing catastrophic. Next in line…balance wires shorting out or being damaged. No catastrophic results from those failures either. I think the owner said that once or twice out of thousands of batteries sold they have a had to replace a pack for someone because a cell went bad in a p group.
So maybe for some all this testing is worth it, not for me. Up to you. I just thought you and others could benefit from hearing the other side of all this testing. If you arent using brand new, genuine cells, yeah 100% go the extra mile.
For laying out odd cell configurations I optimize for the widest conductor path between each p group as possible. If you are using at least 0.2mm copper (probably 0.5 by the looks of it)…dont get too stuck in the weeds. If two or even three cells are passing your max current per cell of 26 amps through a single cell width sized cross section of 21mm and some change…you are going to be just fine. That 26 amp condition is also going to happen extremely infrequently and for very short time based on this discussion thus far. Good luck on the rest of your build. Im almost finished with mine, just waiting on my cl1000 to arrive!
P.S. in summary I believe that when using brand new, genuine cells from a reputable manufacturer and supplier it is all the other aspects of your build that are going to be the determining factor for the outcome of your build, its longevity, and safety. I havent seen anyone talk about cell level fusing? Ive seen multiple people attempt cell level fusing and ends in flames. Too many failure points, too time consuming, and mitigates a scenario that is extremely unlikely when compared to other failure modes. Similarly, yes you are mitigating some risk with extensive testing, but if I had 100 weights to place on a build for its level of safety, i would put very few (maybe 1) on testing when using new genuine cells. The majority of the weight goes to overcharge, overheat, overload, poor welds, poor vibration dampening, hitting trees, meteor strikes, etc.
