I've used the ICA finishes before on bar tops but I was thinking of using the ML Campbell equivalent on an exterior door. What will happen to the relatively thick finish around where the raised panel profile meets the sticking when the panel expands and contracts? Will it crack, thus making is susceptible to failure? The door is a 6 panel, 48x96. The panels will be quite wide, approximately 15".
(Architectural Woodworking Forum)
From contributor J:
I really think two-part polyurethanes are far more trouble than they're worth. First of all, the cans themselves can easily cost ten times as much as the regular stuff (and you better buy twice as much as you think you're going to need). The life-span is very dependent on the number of applied coatings.
A two-part finish will indeed last about three times longer but you'll still need to re-coat about every year or so. If you do happen to "let it go", you'll have an extremely tough mess to remove (every square-inch with a heat gun, scraper and/or very coarse sanding). I believe the most flexible clear coating is still "marine spar varnish" and I would certainly go with that or perhaps Cetol.
Actually I am becoming a firm believer in Cetol. The newer formulas look great and last just as long as any two-part polyurethanes. Best of all, the yearly re-coat should never involve any heavy sanding or stripping of any kind (ever again). If you do "let it go" it eventually just flakes off with a brush.
I do have a door finished with two part poly that doesn't get much weather and it is holding up much better, but spar varnish probably would have been a better choice for that one also.
This exposure is not a warrantable situation for a door maker or finisher, but people still do it all the time. Neither the wood nor the finish can stand the expansion and contraction of the extremes. Some door designs will fare better, but all are at risk.