All articles

CFRP Installation Defects: Blisters, Delamination, and Resin Shortage

Blisters, delamination, and resin-starved laminate are preventable — and graded for accept/inject/cut-and-patch by objective code limits. How to tell the cause, detect it (pull-off >200 psi + substrate-mode), and repair by size. Per ACI 440.2R.

CFRP Installation Defects: Blisters, Delamination, and Resin Shortage

At a glance

  • Delamination is graded by size: a spot under 2 in² is acceptable only if total delaminated area stays under 5% and there are ≤10 spots per 10 ft²; 2–25 in² needs epoxy injection; ≥25 in² must be cut out, patched, and signed off by the engineer.
  • Pull-off passes only on two conditions: >200 psi (1.4 MPa) and failure inside the concrete substrate (ASTM D7522 Mode G).
  • A resin shortage mid-application is not recoverable — once the first pour gels, topping up hours later only films over dry fibre. Pre-calculate the full resin quantity.

Most wet-layup CFRP defects — blisters, delamination, and resin-starved (dry-fibre) laminate — are preventable at installation and are graded for acceptance or repair by objective, code-defined limits. This guide explains what each defect is, how to tell the cause, how it is detected, and the size thresholds that decide accept, inject, or cut-and-patch, grounded in ACI 440.2R and ACI SPEC-440.12-22. For getting it right the first time, see how to install carbon fiber fabric.

What are the common wet-layup defects?

Four conditions cover almost everything an inspector finds. Air voids / blisters — trapped air or vapour lifting the laminate off the substrate or between plies. Delamination — a loss of bond over an area, the defect most tightly graded by the code. Resin-starved (dry fibre) — fibre that was never fully wetted out, from too little resin or resin used past its pot life. And resin-rich — excess unreinforced resin that adds brittleness and wastes material without adding capacity. The first three matter structurally; knowing which one you are looking at drives the repair.

Cross-sections of four CFRP laminate conditions: a good laminate, an air void/blister, a delamination with moisture-vapor path, and a resin-starved dry-fiber laminate. Good concrete full wet-out Air void / blister trapped air — hollow on tap Delamination / moisture moisture vapour (§6.2) Resin-starved dry, uncoated fibre Air voids lift while resin is wet; moisture blisters appear during cure. Resin-starved fibre = pot life exceeded or too little resin. (ACI PRC-440.2-23 §4.1/§6.2/§6.5.)
The four laminate conditions an inspector maps. Air voids and moisture blisters differ in when and why they appear; resin-starved fibre comes from stalled or under-quantified resin. (Schematic; ACI 440.2R.)

Why do blisters form — and how do you tell the cause?

Two mechanisms, distinguished by timing. A bubble that lifts while the resin is still wet is entrained air or an unfilled surface pinhole/void — a rolling and surface-filling problem. A scattered field of small bubbles appearing during cure is moisture vapour driving out of the concrete through the not-yet-cured resin, which ACI 440.2R-23 §6.2 names explicitly as a bond-destroying mechanism. A puncture test confirms which: gas only with dry fibre beneath means trapped air; moisture present means vapour drive — and that must be fixed at the source (test substrate moisture per ACI 503.4, switch to a moisture-tolerant primer) before any repair will hold.

Injecting a moisture blister without fixing the moisture is wasted work. If the blisters are moisture-driven (§6.2), the vapour will simply lift the repair too. Address substrate moisture first — see curing: temperature, humidity, and time.

What causes resin-starved (dry-fibre) laminate?

Running short of resin, or running past pot life. The resin’s job is to wet out and bind the fibre and transfer load between filaments; once it passes its pot life, viscosity keeps climbing and it can no longer penetrate the substrate or saturate the fabric (ACI 440.2R-23 §6.5). That is why a mid-panel resin shortage is not recoverable — once the first pour gels, resin added hours later only skins over the surface and cannot reach fibre already locked in place. The fix is discipline before the pour: pre-calculate the full resin quantity for one continuous wet application, mixing at the correct ratio (e.g. FSE322 at 2:1, ~70 min pot life at 23°C) and never stretching a batch past its window.

How is a defect detected and accepted?

Two field methods, both in the specification. Continuous visual and acoustic tap-testing maps delamination (the method must reliably resolve areas ≥2 in²), and pull-off testing (ASTM D7522) verifies bond. The pull-off acceptance is a two-part test:

Pull-off outcomeDisposition
>200 psi (1.4 MPa) AND substrate-mode failurePass
≤200 psi with substrate-mode failureEngineer may accept
>200 psi but non-substrate failureEngineer’s call / retest

Test frequency is 3 pull-offs per 1,000 ft² of installed FRP, minimum 3 per day; where witness panels are specified, one 12×12 in. panel per 5,000 ft² (minimum 2 per day). Fibre-orientation deviation beyond ±5° from the specified axis triggers an engineer’s capacity review.

How do you repair by size?

Delamination area sets the repair, per ACI SPEC-440.12-22 §3.5.3.

Repair decision flow by delamination size: under 2 square inches accept, 2 to 25 inject, 25 or more cut and patch with engineer approval. Measure delamination area < 2 in²Accept ONLY if:total area < 5% of laminateAND ≤10 spots / 10 ft²→ ACCEPT(else inject) 2 – 25 in²Low-pressureepoxy injection→ INJECT + re-test ≥ 25 in²Cut out ply, patch withequal-ply lapped overlay→ ENGINEER SIGN-OFF
Repair is governed by delamination area: small spots accepted within limits, mid-size injected, large areas cut out and patched with the engineer’s approval. Pull-off must independently pass (>200 psi + substrate-mode). (ACI SPEC-440.12-22 §3.5.3.)

What prevents these defects in the first place?

Prevention is a short discipline list. Verify substrate pull-off ≥1.4 MPa (ACI) — FidStrong’s resin data sheets require ≥1.5 MPa — and check substrate moisture before priming. Roll out entrained air before the resin gels for full wet-out, and lay each ply before the previous one fully cures (a cured ply needs mechanical abrasion and a solvent wipe to bond the next). Pre-calculate resin so a panel is completed in one continuous wet application, and hold fibre orientation within ±5°. Skip the prep and the pull-off fails in the resin or at the interface rather than in the concrete — see surface preparation and common installation mistakes.

FAQ

My tap test found a hollow spot the size of a business card — do I need to fix it?

It depends on the totals, not the one spot. A single delamination under 2 in² is acceptable only if the total delaminated area stays under 5% of the laminate and you find no more than 10 such spots per 10 ft². Exceed either limit and it needs low-pressure epoxy injection (ACI SPEC-440.12-22 §3.5.3.1).

A pull-off read 250 psi but peeled at the resin-concrete interface — does it pass?

Not on its own. Acceptance requires both >200 psi and failure inside the concrete substrate (a D7522 Mode G failure). A high reading with a bond-line or interface failure is a non-substrate mode — the engineer may accept it at their discretion or require a retest, but it is not an automatic pass.

We ran out of resin halfway through a panel and finished three hours later — is that a problem if it looks fine?

Yes — that section is under-saturated regardless of appearance. Once the first pour gels, added resin only films over the surface and cannot wet fibre already locked in place (§6.5). Map the dry-fibre extent by tap-test or thermography; small spots can be injected, larger areas must be cut out and re-laminated in one continuous wet application, lapped onto sound FRP.

Blisters appeared a day after installation, scattered across a patch — what’s the cause?

A scattered field of small bubbles appearing during cure points to moisture vapour driving out of the concrete through the uncured resin (ACI 440.2R-23 §6.2), not trapped air. Puncture-test to confirm moisture, then fix the substrate moisture (test per ACI 503.4, use a moisture-tolerant primer) before any injection repair — otherwise the vapour lifts the repair too.

What test frequency does the spec require beyond visual inspection?

Pull-off testing (ASTM D7522) at 3 tests per 1,000 ft² of installed FRP, minimum 3 per day, passing at >200 psi with substrate-mode failure. If witness panels are specified, one 12×12 in. panel per 5,000 ft², minimum 2 per day. Tap-test delamination mapping is continuous, not sampled.

Is a resin-rich (excess resin) area a defect?

It wastes material and adds brittleness without adding capacity, but the specification does not assign it a numeric acceptance limit the way it does delamination. The structurally graded defects are delamination, and by extension the air voids and dry-fibre areas that a pull-off and tap-test will catch. Aim for full, uniform wet-out rather than a resin-heavy surface.

FidStrong manufactures FSC carbon fabric, saturating epoxies, primers, and injection epoxies for repair, under ISO 9001, 14001, and 45001 systems. Acceptance and repair thresholds here follow ACI 440.2R and ACI SPEC-440.12-22; confirm every value and the repair method against the governing specification and the engineer of record.

All articles