Blisters in Fiberglass Pipes

Here is a good site for downloadable papers regarding fibreglass that may help you

http://www.derakane.com/derakaneControllerAction.do?method.goToFabCenterPage.ash.nes.tierMenuNavID.fabCenter

It sounds like Mr. or Miz 'Guest' directed you to a pretty comprehensive site. I noticed the first sentence in Section 2 of that site reads, "Air bubbles in a laminate reduce its strength and may seriously affect its corrosion resistance properties. Air bubbles in the corrosion liner1 can be more damaging than those in the structural layer." (so it appears you are certainly prudent in your desire to prevent these things!) I noticed also from Section 9 "RECOGNIZING FRP DEFECTS" of that site the following defects are described/defined,

"Air bubbles, voids" – "Air entrapment in and between plies; non-interconnected spherical voids. Whipping action during mixing. Poor or dirty surface. Air bubbles not rolled out during lay-up."

"Blisters" - Rounded, sometimes sharply defined elevations of laminate surface resembling blisters. Too rapid cure. Moisture in resin, filler or glass."

It appears you thus could have at least two factors going against you in a "high temperature" (that I would think will normally accelerate cure) as well as "humid" (moisture)-laden environment, so good luck in implementing it sounds rather complex other recommendations from this site and others).

[I suspect one reason for sensitivity to such defects or damages is the quite small "critical crack (or flaw) length" indicated for polymers/plastics in general (e.g. see the …Journal article "Fracture from Inherent flaws in Polymers" excerpt from the mid-second column page 4202 now posted at http://www.springerlink.com/content/m187024812048527/ that says the following,

"A lower limit on the flaw size for which brittle failure may be expected can thus be calculated from a = ……

For polymers the yield stresses are much lower, … so that a ~260 um. "

(for us English-type folks I calculate 260 micrometers = 0.000260 m = 0.000260mx39.37 in/m = 0.010 in, or only ten thousandths of an inch.)

As well as a quite low strain value (before at least some cracking occurs) for reinforced plastic composites in general, as is at least in part explained in the book "Plastics Institute of America Plastics Engineering" on page 1116 at http://books.google.com/books?id=PUtTPElqgzkC&pg=PA1116&lpg=PA1116&dq=allowable+strain+steel+pipe&source=web&ots=pow5nfZobm&sig=S9IKemx8TMJ5KaEsJetOiGi3Y2E&hl=en&sa=X&oi=book_result&resnum=1&ct=result#PPA1116,M1, with the words, , "The weep point or strain-to-first crack in a wall for filament-wound pipe constructed using isophthalic plastic is currently found to be not less than 0.009 in./in." ]

Another cause of blisters is poorly saturated layers of reinforcing fiber. This allows water to penetrate and lift the fiber layer off the one below. This is sometimes found on sailboats in warm water, where the boat is in the water for several years with only brief haul-outs for maintenance.