Quench sensitivity in 6xxx series aluminium alloys
Quench sensitivity, the dependence of strength and hardness on the cooling rate after extrusion or solution treatment, limits the applications of extrusions of heat treatable aluminium alloys. Reduced properties in slow cooled samples have long been attributed to the loss of solute on heterogeneous nucleation sites during cooling. While minimum cooling rates to avoid precipitation of Mg-Si-containing phases and reduced properties have been mapped out for a broad range of commercial alloys, little is understood about the physical mechanisms leading to quench sensitivity. In this investigation the difference in hardness evolution between slow cooled and fast quenched samples of various commercial and model Al-Mg-Si alloys has been determined for a broad range of natural and artificial ageing times. This has demonstrated that the artificial ageing response of fast quenched samples differs greatly from the artificial ageing response of slow cooled samples. In contrast to fast cooled samples, where natural ageing has a negative effect, the artificial ageing response of slow cooled samples is less dependent on natural ageing time. As a result, hardness differences between fast and slow cooled samples is large during artificial ageing after short natural ageing times, but small after long natural ageing times, such that quench sensitivity strongly varies. In order to explain this phenomenon, the cooling rate dependent microstructural development has been investigated using hardness measurements, electron microscopy (SEM and TEM), differential scanning calorimetry (DSC) and positron annihilation lifetime spectroscopy (PALS). From TEM investigations and theoretical considerations of the diffusion range it has been determined that solute is only lost from a limited region around the non-hardening precipitates forming during cooling on secondary dispersoids. The increased age hardening rate in fast cooled samples after short natural ageing times is therefore the result of a high vacancy supersaturation rather than a high solute supersaturation compared to slow cooled samples. Non-equilibrium vacancies appear to act as precipitation sites for the precursor of the strengthening β”-precipitates. In slow cooled samples the formation of β”-precipitates seems to be suppressed. In lean alloys with a low solute content, such as the investigated AA6060, the age hardening response is almost independent of the cooling rate. Small additions of Cu or changes in the Mg:Si-ratio in alloys with a combined Mg+Si-content greater than 1 wt.% have little effect on precipitation of non-hardening phases during cooling, but affect natural and artificial ageing behaviour. The cooling rate after extrusion or solution treatment was also found to affect whether natural ageing has a negative effect or not. In fast cooled samples of Al-Mg-Si alloys with combined Mg+Si-contents greater than 1 wt.% the decrease in ageing kinetics due to the decrease of non-equilibrium vacancies with increasing natural ageing time contributes to the negative effect. In slow cooled samples the initial number of quenched-in vacancies is low, such that ageing kinetics are independent of the natural ageing time.