The room temperature yield strength of two alloys, Al-2.30 wt% Li-2.85 Cu-0.12 Zr (the 2–3 alloy) and Al-2.90 Li-0.99 Cu-0.12Zr (the 3-1 alloy), was investigated as a function of aging time at 160 and 190°C. Reversion experiments were used to separate the contributions of the δ' and T 1, precipitates to the overall strength of the aged samples. The reversion heat treatments nearly completely dissolved the δ' precipitates, leaving the dispersion of T 1 precipitates essentially undisturbed and enabling its contribution, Δτ T 1 , to be measured experimentally. Assuming linear additivity of the precipitation hardening contribution, Δτ p , and all the others (due to grain size strengthening, solid solution strengthening, etc.), the δ' contribution, Δτ δ' , was estimated using a generalized addition rule of the type Δτ p q = Δτ δ' q + Δτ T 1 q , where q is a constant (the shear stresses, Δτ, were obtained from the measured yield stresses using data on the textures of the alloys). These values of Δτ δ' , were then analysed according to a recent version of the theory of order strengthening which requires that the trailing dislocation be pulled through the sheared ordered precipitates immediately on encountering them. It was found that for the underaged alloys only q = 1.4 yielded values of the APB energy on {111}, γ apb , that were identical for both the 3-1 and 2–3 alloys (0.150 J/m 2). Using this value of γ abp , the theoretically predicted values of Δτ δ' at peak strength and the δ' particle size at peak strength were in very good agreement with the experimentally measured values, and the predicted spacings of the dislocation pairs were in excellent agreement with those measured. It is demonstrated that the value of q is not merely an empirical constant required to bring theory and experiment in agreement, but is justified by the results of computer simulation experiments on strengthening by point obstacles of two distinct strengths. The possible role of other strengthening mechanisms is discussed.
Read full abstract