Question about flares
I had always believed that there were two kinds of flares on line ends, perfect and wrong.
I just had an experienced HVAC tech tell me that a flare which has an irregular ridge at the outside edge of the mating surface, as if there was a burr from cutting the copper line and the flaring cone had flattened it out, would "squish right out" when torqued down against the brass fitting in a mini split, and would never leak, not even with 410A. Is he right? Most of the flared surface looked fine, but the raised and slightly jagged ridge at the outer edge would be the first thing to make contact with the fitting on the equipment.
I don't have a digital camera with the macro focusing needed to show you what I'm talking about.
He would be wrong. The tubing needs to be reamed before flaring.
Originally Posted by taxman
Thanks. It was a factory made line set. One end looked perfect, the other end looked like the deburr step got skipped.
I tried to pay the guy to cut the flare off and redo it, and he refused.
The funny part is, this is an outfit with a very good reputation, they claim to always flood nitrogen while brazing, and they pressure test everything with nitrogen.
Maybe they'd rather redo a bad flare during pressure testing than before?
I've made a LOT of bad flares over the years and used them.
Would I knowingly use a bad flare today? No. Not for the time that it would take to cut, deburr and reflare it.
It isn't worth the hassle of a leak down the road somewhere, IMO.
That's what I was thinking.
Originally Posted by jpsmith1cm
10 minutes x $$ per hour to reflare the lines vs possibly dealing with a leaky Fujitsu mini split 2-3 years later...
Tomorrow's Saturday, maybe I'll drive to every HVAC shop in town with this lineset and see who the perfectionists and anti-perfectionists are.
The funny thing is, I called ahead, described it just as I described it here, they said to bring it in and they'd fix it for me. Then they said there was nothing wrong with it when I got there.
Last edited by jpsmith1cm; 11-09-2012 at 05:38 PM.