Guidance on wire gauge size for automotive wire in a bundle

[IntegraR0064]

Hi,

I'm designing a wire harness for class 8 heavy duty trucks. I need to size the wires. Everything is 12V DC. I have designed a first draft of a wiring harness and I have all of the currents and voltage drops over each wire mapped out. All the wires are in a bundle that will be run along the frame rails of the truck.

My question is - how do I determine the acceptable voltage drop and/or maximum current in each wire? I have seen tables online but they all assume open air. I have seen references to "compensating for bundles" but have not found how to do it. Of course they need to be sized so that each of the components has adequate voltage but I think the more limiting factor is overheating the bundle. Another complicating factor is I have pretty high peak amperages at startup, but after a fairly short time these amperages drop to very low levels. How do I deal with that?

I have also searched for a textbook or paper or something on this and have come up short - is there a good reference I should be looking at?

For more detail on the wiring, it is GXL wiring, 125 degrees C maximum insulation temperature, SAE J1228. Right now wire sizes are between 10 and 22 AWG. It will be bundled in groups of varying number of wires up to maybe 10 or 15 wires, but usually 3-5 wires, with a PVC coated woven yarn (max temperature 80 C) over the bundle.

Thanks for any input on any of my questions!

 

[Fish4Fun]

IntegraR0064,

I don't have an engineering degree, so take my thoughts as a simple layman's recommendation.

You might consider taking the current through each wire and calculating the I^2R losses as 100% heat addition to the interior of a PVC pipe similar in diameter to your bundle. You should be able to calculate the heat loss through a PVC pipe to ambient air based on the temperature differential. It should be fairly easy to write an iterative program that tests the heat added and lost in one second intervals, and as long as it can make it through 36,000 iterations w/o getting too close to 125C, you should be fine.

I am certain there are more rigorous engineering ways of proving safety, but I figured if I threw this out there, someone better qualified might give you the answer you are looking for.

Fish

 

[Fish4Fun]

IntegraR0064,

I don't have an engineering degree, so take my thoughts as a simple layman's recommendation.

You might consider taking the current through each wire and calculating the I^2R losses as 100% heat addition to the interior of a PVC pipe similar in diameter to your bundle. You should be able to calculate the heat loss through a PVC pipe to ambient air based on the temperature differential. It should be fairly easy to write an iterative program that tests the heat added and lost in one second intervals, and as long as it can make it through 36,000 iterations w/o getting too close to 125C, you should be fine.

I am certain there are more rigorous engineering ways of proving safety, but I figured if I threw this out there, someone better qualified might give you the answer you are looking for.

Fish

 

[IntegraR0064]

Good suggestion! I'll try that as an approximation.

 

[pantaz]

I've never found information specifically for bundled wires, but the following pages have some helpful tables and formulas for building harnesses:
http://www.stealth316.com/2-wire-resistance.htm
http://www.powerstream.com/Wire_Size.htm
http://donsnotes.com/home_garden/wire.html
http://amasci.com/tesla/wire1.html

 

[IntegraR0064]

Hey thanks for those links. From one of them I found a NEC bundled wire table - http://www.houwire.com/products/technical/article310_16.html [Broken]. I'm pretty far under those values so I think I'm good from a melting wires perspective. Thanks again.