Quantification and modelling of the microstructure/strength relationship by tailoring the morphological parameters of the T₁ phase in an Al–Cu–Li alloy

Abstract

We present a systematic study of the relationship between precipitate microstructure and resulting yield strength in an Al–Cu–Li alloy. By varying the thermomechanical ageing treatments applied to the AA2198 alloy (pre-deformation, heat treatment temperature and duration), T₁ microstructural parameters (thickness, diameter and volume fraction) are varied over a wide range, and the corresponding yield strength is determined. The resulting database of microstructure–strength relationships is used to establish a strengthening model based on interfacial and stacking fault strengthening. This model successfully describes the strength evolution from under-aged to over-aged conditions without the need for a transition from precipitate shearing to by-passing.

Introduction

The high specific strength, good damage tolerance and excellent property stability of Al–Cu–Li alloys makes them highly attractive for aerospace applications [1], [2], [3]. Recently developed Al–Cu–Li alloys can be found, for instance, under the name AIRWARE® [4]. A complex precipitation sequence that relies heavily on thermomechanical treatment is involved in these alloys [5], [6]. Cold working prior to the final ageing treatment has been shown to promote precipitation of the T₁-Al₂LiCu phase at the expense of other phases [7], [8], [9], [10]. The T₁ phase is known to provide the highest strength [9], [11]; it forms as semicoherent platelets along the {111}_Al planes and exhibits an hexagonal structure [12], [13].

The strengthening response associated with the formation of precipitates in Al alloys has been widely studied in the literature [14], [15], [16], [17]. In the case of Al–Cu–Li alloys, the T₁ plates were originally thought to be shear resistant [18] and thus attempts have been made to use modified versions of the Orowan equation for precipitate by-passing [18] to model the yield strength increment associated with the T₁ phase [19], [20]. The applicability of these early attempts is, however, questionable as T₁ precipitates have been recognised as being shearable for a wide range of heat treatments [21], [22], [23], [24]. An interfacial strengthening model, considering the energy required to create a new precipitate–matrix interface when shearing a T₁ plate, has been developed by Nie and Muddle [25]. This model has been only once compared to a limited set of microstructural data [25]. Therefore there is a shortage of quantitative microstructure data available to critically discuss the applicability of the different strengthening models, as well as their robustness and domain of validity. We present in this paper a systematic study in which a wide range of T₁ microstructures are generated by varying the parameters of the thermomechanical treatment, in order to generate a complete database of microstructure–strength relationships. This database will be used to evaluate and to improve the existing strengthening models.

The existing literature indicates that the choice of thermomechanical treatment parameters has a major effect on the resulting precipitation microstructure in Al–Cu–Li alloys [7], [8], [9], [10], [26], [23], [27]. The thermomechanical parameters that strongly influence precipitation are: the pre-age stretch, the heat-treatment temperature and the heat-treatment duration. Varying these parameters is a suitable approach to achieve a large variety of T₁ microstructures and provides a strategy for varying the parameters of the precipitate microstructure independently. The heat-treatment temperature has recently been shown to be a key parameter that controls the activation of the T₁ thickening process [27]. Namely, it was shown that at 155 °C the thickness of T₁ precipitates remains constant for very long ageing times at the minimum value of 1.3 nm, and that when temperature is increased from 155 to 190 °C T₁ thickening is quickly activated. In addition, since T₁ nucleates on dislocations, increasing the level of pre-deformation leads to a greater density of smaller precipitates [9]. The influence of pre-deformation on the competitive precipitation of T₁, θ′ and δ′ has been addressed by Gable et al. [9]. They revealed that the same strength of 450 MPa could be reached for different pre-deformations (0%, 2%, 4% and 6%) that result in significantly different T₁ microstructures: the mean T₁ diameter was found to range from 40 to 160 nm, highlighting that the relationship between the T₁ morphological parameters and the yield strength is not straightforward. However, in the alloy studied by Gable et al., θ′ and δ′ were also present in considerable amounts, making the analysis complex. We will show in the present study that with an appropriate choice of alloy composition and thermomechanical treatment, the microstructure becomes dominated by T₁ precipitates only.

We recently developed a method to characterise the T₁ phase quantitatively in terms of morphological parameters and volume fraction [27]. This method first uses small-angle X-ray scattering (SAXS) to measure the mean diameter and thickness of T₁ precipitates. SAXS can be carried out in situ during the heat treatments, giving access to the precipitation kinetics [28]. Secondly, this method uses differential scanning calorimetry (DSC) to measure the volume fraction of T₁ precipitates. It will be applied systematically in the present investigation to record the T₁ microstructures for a wide range of thermomechanical treatments. In parallel, the yield strength will also be measured in all the studied heat-treatment conditions. The resulting microstructure–strength characterisation will be then used to challenge the existing models for platelet strengthening; a modified approach to T₁ strengthening will then be proposed.