Effect of Composition and Thermomechanical Processingon The Intergranular Corrosion of AA6000 Aluminium Alloys
MetadataVis full innførsel
In order to achieve the desired strength, AlMgSi (6000-series) alloys are often alloyed with either small (fraction of a wt%) Cu or excess (in relation to that required to form Mg2Si) amounts of Si. Both approaches can introduce the desired strength after artificial ageing. However, these methods have also been reported to introduce susceptibility to intergranular corrosion (IGC), depending on the applied thermomechanical treatment procedure. There is disagreement in the literature whether additions of Cu or excess Si is the most harmful for IGC susceptibility, and the IGC mechanism of 6000-series alloys in this context is not well understood. This work investigates the effect of composition and thermomechanical history on IGC susceptibility and sheds light on the mechanism of IGC in 6000-series alloys in general. Intergranular corrosion (IGC) in 6000-series model alloys has been studied using accelerated corrosion tests, scanning electron microscopy (SEM), analytical field emission scanning transmission electronmicroscopy (FE(S)TEM) and In situ synchrotron X-ray microtomography. The alloying content of model alloys was Cu (0.02-0.3 wt%), Si (0.6-1.0 wt%), Fe (0.20-0.42 wt%), Mn (0.15- 0.56) and Cr (0.02-0.17 wt%) in the concentration ranges specified. These alloys were subjected to a range of thermomechanical treatment processes, specifically consisting of extruding, rolling and a range of heat treatment conditions. Important results obtained can be summarised as follows:Cu lean alloys (< 0.03 wt%) were not susceptible to IGC regardless of temper. Only slight superficial attack was seen in alloys with excess Si (Si:Mg ≈ 1.5).Cu containing alloys (0.2 - 0.3 wt%) were highly susceptible to IGC when they were air cooled after extrusion. The susceptibility was mitigated by artificial ageing to the near peak strength T6 condition.Cu containing alloys were resistant to IGC when rapidly quenched after extrusion. Underageing introduced IGC susceptibility that was gradually reduced upon by further artificial ageing towards the T6 condition.Fe did not introduce IGC susceptibility in the concentration range tested. However, it caused deeper and more numerous attacks in tempers otherwise susceptible.Increasing the amount of Mn had no significant effect on corrosion susceptibility and no visible effect on grain boundary precipitation.Small additions of Cr (~ 0.15 wt%) had a beneficial effect on corrosion performance of Cu lean alloys, but negligible effect on Cu containing alloys.Pre-aging cold work improved the ageing response, thus shortening the ageing time required to obtain an IGC resistant temper.Corrosion testing of a virtually Cu free alloy with high excess Si (Si:Mg ≈ 3) revealed limited IGC when the material was water quenched from high temperature and subjected to artificial ageing. This IGC was much less severe than that found in Cu containing material. Selective removal of surface intermetallics rendered the material completely resistant to IGC, whereas a copper containing alloy was still susceptible to IGC when subjected to the same procedure. However, the attack filament width was significantly smaller. In situ studies of IGC in a susceptible Cu containing material exposed to an aggressive solution was performed using X-ray microtomography. The tomography data revealed that IGC penetrated the grain boundaries very rapidly. Further corrosion occurred by widening of existing attacks, which was the stage that could be visualised by the technique due to its limited resolution. The corrosion rate as determined by corroded volume fraction vs. time showed that the corrosion rate was relatively constant. Careful analytical FE-(S)TEM characterization coupled with statistical analysis of EDS-maps showed a significant depletion of Si towards the grain boundaries of both Cu containing and Cu free alloys. A similar depletion was also found for Cu in a high Cu (0.7 wt%) alloy, but could not be resolved in alloys containing approximately 0.2 wt% Cu. This is believed to be caused by limitations in the sensitivity of the technique. The EDS data also revealed a significant enrichment of Cu at the grain boundaries of Cu containing, IGC-susceptible materials. The film was a few nanometres wide and contained significantly higher amounts of Cu than the base material. The structure of the film remains unknown, but it is assumed to be a highly efficient cathode. Based on the results in this work it is concluded that IGC in 6000-series alloys is caused by depletion of Si and Cu along the grain boundaries. This renders the depleted zones more active than the grain bodies. The cathodic reaction occurs to some extent on the surface intermetallics (external cathodes). However, the dominating cause of the rapid, knife-line attack, which is typical for the susceptible Cu-containing alloys is the microgalvanic process between the Curich film as the internal cathode and active aluminium in the depleted zone as the anode in the IGC filaments. This work provides new information regarding the grain boundary structure, the growth rate and morphology of IGC, and a comprehensive study of the effect of Mn, Cr and Fe on IGC in AlMgSi(Cu) alloys. A side-by-side comparison of the corrosion performance of alloys containing Cu and Cu-free, high Si alloys has provided new insight to the Cu vs. Si controversy encountered in the literature.