Screw terminal burnt 1

[CostasV]

Hi all,

During an annually check we have found that inside an ESD marshalling cabinet, one of the two parallel PSU 230VAC/24VDC-40A was not working because the 230 VAC feeder screw terminals and cables (6 mm2) were burnt. The terminal plastic insulation melted, phase and neutral came to contact and short-circuited, which caused the MCB (25A) to trip. The marshalling cabinet environment is kept controlled, no dust, temperature 20..24 degC.

My colleagues think that the cause was a high resistance connection inside the screw terminal blocks.

This marshalling cabinet was running continuously from 2012. Two months before, we were working inside the cabinet (but not on the damaged terminal) without noticing anything strange, so the burnt terminals must have happened in the last 2 months.

The terminals were rated for 50 A. If a well tightened terminal has a resistance of 0,0001 .. 0,001 Ohm , then maximum heat is 0,25 .. 2,5 W.
The PSU current was about 4,5 A so if the resistance from the loose connection increased from 0,001 to 1 Ohm, then the produced heat would be 20 W.
Any comments for this calculation? Does this make sense, or am I missing something?

Is a terminal connection possible to run with no problem for 10+ years and then start to increase its resistance?
What makes the resistance to increase with the time? Is this increase linear or more steeply ?

Are there any articles to explain how this type of fault happens ? I have searched and found only a Wikipedia article which is not very detailed.

Thank you for your feedback!

 

[IRstuff]

Is a terminal connection possible to run with no problem for 10+ years and then start to increase its resistance

You're asking as if it were a binary failure, but it could easily be a gradual decay in performance, and only reaching a failure threshold now.

It's not that different than corrosion, where it gradually gets worse, and only when your attention is called to it do you realize that "something" happened, seemingly overnight. I've had issues with breaker panels losing connections over time in a salt-air environment, literally 50 ft from the ocean.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[che12345]

[COLOR=]"... The terminals were rated for 50 A. If a well tightened terminal has a resistance of 0,0001 .. 0,001 Ohm , then maximum heat is 0,25 .. 2,5 W....The PSU current was about 4,5 A so if the resistance from the loose connection increased from 0,001 to 1 Ohm, then the produced heat would be 20 W..."[/color]
I have the following opinion for your consideration.
1. The original loose connection may be say 1 Ohm, producing heat 20W. The contact resistance increases due to over-heating increases from say from 1 Ohm to 10 Ohm, producing heat 200W !.
2. At say 20W the wire insulation and the terminal block material may show very little noticeable sign of over-heating. After a short period of time when the resistance increases say 1 to 10 Ohm, serious damage occurred. It may appear to occur "suddenly". Actually it is NOT, but a gradual damaging process which is not obvious/noticeable at the initial stage.
3. A good tight connection retaining low contact resistance with low heat would last for 30-50 years.
4. The other problem frequently encountered at site is a short created by lizard, snake, rats ... across the terminals. Look for evidence of carcass. Increase the IP rating if needed.
Che Kuan Yau (Singapore)

 

[bacon4life]

Worsening of a connection can be very nonlinear. One the connection develops a small amount of excess resistance, it begins operating above the rated temperature of the materials. This can lead to excessive thermal elongation of the bolt/screw that causes the metal to exceeding the elastic creep limit and permanent deformation. Over repeated heating/cooling cycles, this can lead to the connection becoming progressively looser. As the connection gets looser, the resistance will progressively increase, which leads to increased heat.

 

[CostasV]

Thank you all for your comments.

Update:
We have checked with a thermographic camera the wire connections inside all the rest marshalling cabinets and found no other issues.

Is there anything else to prevent this from happening again?

 

[RedSnake]

Here the rule was to go over all screw terminals one year after installation tightening them to required tightening torque while the machine just had been running being at working temperature.

For the reasons bacon4life has given.

The reason for this as I was told was that when you make the installation the machine is "cold" room temperature when you start running the machine everything gets warmer and the tightening torque isn't the same anymore, the screws can even screw themself out due to the changes from heat cold.
Off course doing this more than once you will eventually destroy the screw.

I saw that there were crimp bootlace used too, this can also create a problem if they are not applied with the right tool and force, the same thing will happen with them, they can also lose their shape if not applied right from changes in heat and cold making a gap between the crimp bootlace and the the wires.

One solution is using feed-through terminal with tension-clamp connection preferably without crimp bootlaces.

Like ZDU 10, Feed-through terminal, Tension-clamp connection, 10 mm², 1000 V, 57 A.

NAFO Sergeant Anna Gr 69th Sniffing Brigade

 

[jraef]

As IRStuff said, it’s rarely so linear as to be calculable. As materials heat up, their resistance changes slightly. So it’s a cascading effect that may take a long time to reach the point of failure, but may be undetectable to the naked eye for much of that time. This is why thermal imaging scans can be used to track changes in heating over time.


" We are all here on earth to help others; what on earth the others are here for I don't know." -- W. H. Auden