Log in to the online community

Want to post a reply? You'll need to log in
Thermal imaging of cables
Jam 1100158471
56 Posts
Question
Hi all,
So it's a little sunny out, and I have a phone-mounted IR camera, so I thought I'd leave some offcuts of larger cable out to cook in the garden and sense-check my derating assumptions for cables in direct sunlight.

However, I've only really used the IR camera for differential inspections (i.e. that phase is running hotter than the other two) and not worried terribly much about the absolute values, which do change significantly depending on the material emissivity setting.

Can anyone recommend a source for aluminium and copper conductors and black PVC and MDPE / HDPE sheaths after exposure to normal installation treatment? (I.e. sheath roughened a little from pulling in, metal dulled from oxidiation, but neither burnt nor mirror-polished)

The difference between the sunny and shaded side of a 45mm OD single core is quite telling; might try a bundle in trefoil tomorrow!

Jam
4 Replies
OMS 615869
646 Posts
PVC would be about 0.85 to 0.95

MDPE wouldn't be be too dissimilar

Metals are much more problematic depending on surface finishes - there could easily be two orders of magnitude or more between polished and anodized aluminium - oxidized aluminium is around 0.2 to 0.3

Similar numbers for copper but if it's had heating, could easily be double or more.

As you've found out, absolute values are a nightmare - normally if you are looking at say a DB, then anything hotter than say 60C is a trigger value for investigation

Regards

OMS

 
Jam 1100158471
56 Posts
That’s grand thank you.

FYI, My Seek Thermal has options of 0.3, 0.6, 0.8 and 0.97 which isn’t exactly fine-tuned (but then again it doesn’t pretend to be a fully featured device either), but taking the closest would suggest a sheath and conductor surface temperature appx 30oC above ambient in free air, though the shaded side of the sheath is 10ish lower. Offcuts so no gain from load. Slightly higher than I’d expected, but that’s surface temperature not ambient air.
OMS 615869
646 Posts
Sounds about right - could even be substantially higher.

Basically, what you are dealing with is difference between what is effectively the air temperature and the sol air temperature along with the resultant temperature rise due to incident radiation.

If we take a transformer as an example, used somewhere hot and dusty. It's pretty easy to get a unit that can deliver the required power at say an ambient 50C air temperature. You could dump it on a slab and it'll be perfectly happy until the sun comes out - the incident solar radiation will quickly start to raise the temperature of the metal casing to easily 90C or more - it'll take the skin off your hand. If you add a simple shade over the top (usually a dual layer of wriggly tin) then that incident radiation has no effect on the transformer. We tend to use a dual layer at a slope of say 15 degrees as the outer sheet gets red hot, heats the air space between it and the lower sheet by radiation, which induces a convective air current, which keeps the lower sheet much cooler and therefore has much less ability to put incident radiation on the transformer - Happy days, if you need to give that transformer a bit of hard work.

Ideally, you'd put it inside with mechanical cooling  - but simple shading is quite effective

Regards

OMS

 
mapj1 80733779
2070 Posts
indeed,  and many things other than transformers also benefit from a sun hat, even a very crude one.  and if the "wiggly tin" can be painted UN white, rather than cammo coloured, even desert sand, that helps a lot.

In  warmer climes than the UK, we have seen green boxes of radio kit on the floor outdoors reach 70C internally before they even get switched on  - a small version of the effect when you park your car in the sun for the day, and then cannot bear to sit in it.
We do a lot of calibration with temperature indicating labels, and if we want to use IR for an accurate temp test, we usually  stick a label of known emissivity on the thing.

Log in

Want to post a reply? You'll need to log in