A double cantilever beam (DCB) specimen is two bonded composite laminate arms with a pre-implanted delamination of length a at one end. Pulling the arms apart opens the crack in pure Mode I (opening mode). Each arm behaves as a cantilever beam fixed at the crack tip, so beam theory (ASTM D5528 "modified beam theory") gives the compliance:
C = δ/P = 2a³ / (3·E₁₁·I), I = b·h³/12
Energy release rate (Irwin–Kies):
G_I = (P²/2b)·dC/da = 3·P·δ / (2·b·a)
The crack is stationary as long as G_I is below the material's Mode I interlaminar fracture toughness, G_Ic — the energy per unit new crack area the interface can absorb. As the crosshead keeps opening the specimen, P and δ rise together (compliance-controlled), and G_I rises with them. The instant G_I reaches G_Ic, the Griffith energy criterion is satisfied and the crack advances just far enough to bring G_I back down to G_Ic (resistance-controlled growth) — solving the compliance relation for the equilibrium crack length at the current opening:
a_eq(δ) = ( 9·E₁₁·I·δ² / (4·b·G_Ic) )^(1/4)
This produces the classic saw-toothed DCB load trace: load climbs, drops as the crack jumps, then climbs again on the freshly exposed (still bonded) interface ahead. Real interlaminar toughness is what keeps a wing spar, a wind-turbine blade, or a pressure-vessel wall from peeling apart ply-by-ply under service loads — this is the exact test aerospace and wind-energy labs run to measure it.
- Ply system — swaps E₁₁ (fiber-direction modulus) and G_Ic for three real material systems.
- a₀, h — initial crack length and arm thickness set the starting stiffness.
- Opening rate — how fast the virtual crosshead pulls the grips apart.
Deflection and arm separation are visually amplified for legibility (true openings are a few millimeters over a 140 mm specimen); the underlying δ, P and G_I readouts are the real computed values.