Heat Treatment Optimization in Al-Cu-Mg-Si Alloys, with or without Prior Deformation

: The properties of Al-Cu-Mg alloys simultaneously depend on dissolving the maximum amount of Cu and Mg in the solution treatment and on achieving optimal ageing. The aim of this study is to analyze the effect on an Al-Cu4.5-Mg1.5-Si0.75 alloy, manufactured by continuous casting and hot rolling until achieving a reduction in its cross-section greater than 90%, of the dwell time at the solution temperature, 495 ◦ C (4, 8, and 24 h), of the different ageing temperatures (160, 180, and 200 ◦ C), and of cold rolling prior to ageing. The microstructural variations underwent the material during its manufacturing process, form its continuous casting to subsequent hot rolling, were analyzed by means of optical microscopy (OM) and scanning electronic microscopy (SEM), with characteristic energy dispersive X-ray (EDX) microanalysis. The crystalline phases present after the different solution and natural ageing treatments were identiﬁed and quantiﬁed by means of X-ray diffraction (XRD), concluding that Mg is easier to dissolve than Cu. The transient states associated with the Mg 2 Si phase are the most abundant. However, the longer the dwell time at the solution temperature, the greater the weight percentage of the transient states associated with the Al 2 Cu phase during ageing. The higher the ageing temperature, the faster the peak hardness is reached, but the lower its value. The ageing temperature that allows the highest hardness to be obtained was 160 ◦ C. The maximum hardness value reached was 162 HV, obtained after a solution treatment at 495 ◦ C for 4 h and ageing at 160 ◦ C for 50 h. By means of prior cold rolling, the peak hardness values are reached more quickly and their values slightly exceed those obtained without this deformation. With ageing at 180 ◦ C, 168 HV are reached after 6 h at this temperature. 300 g load. The results correspond to the average value obtained from 10 indentations in each specimen. The hardnesses were measured in an intermediate zone between the periphery and the center of one of the cross-sections of the sample.


Introduction
Al-Cu-Mg alloys are widely used both in the aeronautical industry and in the automotive industry to manufacture structural elements. In addition to depending on their chemical composition, their mechanical properties are also determined by the solution heat treatment, quench rate, and ageing heat treatment employed. In the solution heat treatment, the aim is to dissolve the phases containing Cu and Mg and eliminate possible interdendritic chemical segregations derived from a non-equilibrium cooling. The maximum temperature at which this treatment can be carried out will depend on the content in Cu and Mg. When the percentage by weight of Cu exceeds 4%, and the percentage of Mg exceeds 1.5%, the ternary eutectic reaction of equilibrium between the liquid phases, the α solid solution, CuMgAl 2 , and CuAl 2 may take place at 508 • C [1].

Materials and Methods
The alloy is cast in a tundish at a temperature of 700 • C, from which the molten metal is channeled into a U-shaped circular copper die cooled with water, producing a bar with a rectangular cross-section. This is then transported by rollers to the rolling mills where the bar is transformed into 15 mm diameter wire rod, working within a temperature range of between 470 and 450 • C. Once this process is finished, the material is water cooled to a temperature of around 50-60 • C. The resulting microstructure was analyzed using specimens obtained from the finished product.
Subsequently, three solution treatments were carried out at 495 • C to avoid incipient fusion [19], employing different dwell times at this temperature: one of 4 h, another of 8 h, and third of 24 h. The aim of this last dwell time was to carry out a long treatment and compare the solution of the phases containing Cu and Mg with the two previous treatments. Cooling was carried out in water at 15 • C. The resulting microstructure was analyzed in the three cases. The phases present after 1000 h of natural ageing were identified and quantified by means of X-ray diffraction (XRD, PANalytical, Almelo, The Netherlands). The diffractograms were obtained on a Philips X'Pert Pro diffractometer (PANalytical, Almelo, The Netherlands) belonging to the Scientific-Technical Services, University of Oviedo (Spain). A fine focus Cu ceramic tube located in the fixed primary arm was used to produce X-rays, operating under 45 kV × 40 mA working conditions. The diffracted radiation was analyzed on an X'Celerator detector (PANalytical, Almelo, The Netherland) equipped with a secondary monochromator that filters the signal to the K doublet (1.54056-1.54439 Å) of the Cu anode. The recordings were made in continuous mode, over the two ranges between 10 • and 90 • , employing an angular step and count time of 0.0084 • and 128 s, respectively. Conventional reference-intensity ratio (RIR) method [20] was used to estimate the weight fraction of the crystalline phases by means of semi-quantitative analysis.
Several ageing treatments were carried out at three different temperatures on specimens treated for 4 and 8 h at 495 • C: 160, 180, and 200 • C. Other specimens that had previously been treated for 4 and 8 h at 495 • C were rolled at room temperature to obtain a true deformation of 0.25 (25%). They were then subjected immediately (in a period of less than 20 min) to ageing treatments at 160, 180, and 200 • C employing different dwell times at these temperatures. The deformation was carried out on 10 mm thick specimens by means of eight rolling passes in 75 mm diameter work-rolls rotating at a linear velocity of 65 mm/s. Hardness was determined after the different dwell times in order to determine the hardness peak for each ageing temperature and compare its value and time of appearance with the homologous samples that had been previously rolled. Figure 1 shows a schematic of the process that has been followed.
Metallographic inspection was carried out by means of optical microscopy (Nikon, Tokyo, Japan) and scanning electron microscopy (JEOL, Nieuw-Vennep, The Netherlands). A semi-quantitative analysis of the main elements present in the precipitated phases was carried out by means of characteristic energy dispersive X-ray (EDX, JEOL, Nieuw-Vennep, The Netherlands) microanalysis. The optical microscope employed was a NIKON Epiphot 200, the images being obtained using the Omnimet Enterprise Image Analysis System (Enterprise, BFuehler, Lake Bluff, IL, USA). The electron microscope used was a JEOL JSM-5600. The images were taken in an intermediate zone between the periphery and the center of the cross-section of the wire rod. Preparation of the metallographic samples was carried out via the process of cutting with a SiC disc, cold-drawing in plastic resin, roughing on SiC paper using different sizes of abrasive grain, from grit 240 to 600, and, finally, polishing with textile cloths spread with 6 and 1 micron diamond paste and lubricant. The samples were inspected without chemical etching, solely in the polished state.
Vickers hardnesses were obtained under the application of a 300 g load. The results correspond to the average value obtained from 10 indentations in each specimen. The hardnesses were measured in an intermediate zone between the periphery and the center of one of the cross-sections of the sample.  Table 1 shows the chemical composition of the Al-Cu4.5-Mg1.5-Si0.75 alloy under study.  show the microstructure obtained after continuous casting. In Figure 5, the Mg2Si and Al2Cu particles are identified as the main phases detected by the semi-quantitative analysis carried out by EDX. Figure 6 shows the microstructure obtained after hot rolling, during which a 90% reduction in cross-section is obtained. A finer microstructure can be seen as a result of the hot dynamic restoration and the hot static recrystallization the material underwent [21], which implies a refining of the grain size [22]. Furthermore, during this process, the fragmentation of the particles that had precipitated during the solidification takes place, losing their continuity in the grain joints. Figure 7 identifies, by EDX, the Mg2Si and Al2Cu phases as the main phases present after rolling. Dark spots appear within the Al2Cu network that were identified as Mg2Si via the semi-quantitative analysis by EDX. This phase precipitates at a higher temperature than the Al2Cu phase, so it may be assumed that its precipitation has ceased when the precipitation of the Al2Cu network begins.    Table 1 shows the chemical composition of the Al-Cu4.5-Mg1.5-Si0.75 alloy under study.  show the microstructure obtained after continuous casting. In Figure 5, the Mg 2 Si and Al 2 Cu particles are identified as the main phases detected by the semi-quantitative analysis carried out by EDX. Figure 6 shows the microstructure obtained after hot rolling, during which a 90% reduction in cross-section is obtained. A finer microstructure can be seen as a result of the hot dynamic restoration and the hot static recrystallization the material underwent [21], which implies a refining of the grain size [22]. Furthermore, during this process, the fragmentation of the particles that had precipitated during the solidification takes place, losing their continuity in the grain joints. Figure 7 identifies, by EDX, the Mg 2 Si and Al 2 Cu phases as the main phases present after rolling. Dark spots appear within the Al 2 Cu network that were identified as Mg 2 Si via the semi-quantitative analysis by EDX. This phase precipitates at a higher temperature than the Al 2 Cu phase, so it may be assumed that its precipitation has ceased when the precipitation of the Al 2 Cu network begins.   Table 1 shows the chemical composition of the Al-Cu4.5-Mg1.5-Si0.75 alloy under study. Figures  2-5 show the microstructure obtained after continuous casting. In Figure 5, the Mg2Si and Al2Cu particles are identified as the main phases detected by the semi-quantitative analysis carried out by EDX. Figure 6 shows the microstructure obtained after hot rolling, during which a 90% reduction in cross-section is obtained. A finer microstructure can be seen as a result of the hot dynamic restoration and the hot static recrystallization the material underwent [21], which implies a refining of the grain size [22]. Furthermore, during this process, the fragmentation of the particles that had precipitated during the solidification takes place, losing their continuity in the grain joints. Figure 7 identifies, by EDX, the Mg2Si and Al2Cu phases as the main phases present after rolling. Dark spots appear within the Al2Cu network that were identified as Mg2Si via the semi-quantitative analysis by EDX. This phase precipitates at a higher temperature than the Al2Cu phase, so it may be assumed that its precipitation has ceased when the precipitation of the Al2Cu network begins.             Microstructure of the material after continuous casting, obtained by scanning electron microscopy. Particles of Mg 2 Si (black coloring) and Al 2 Cu (white coloring) identified by semi-quantitative analysis by characteristic energy dispersive X-ray (EDX) microanalysis.   In the wire rod form, in addition to the α phase, the main identified phases were Al2Cu () and Mg2Si (β), although the probable presence of AlFe2 and AlCu3 was also detected. The solution treatment and natural ageing led to a different texturing of the constituent matrix (α phase), which is reflected in the variation in the relative heights of its Bragg peaks associated with an increase in the percentage in weight of the previously identified phases and the appearance of the Al7Cu2Fe phase. Table 2 shows the percentage in weight obtained by means of the reference-intensity ratio (RIR) method. Ageing leads to an important increase in the percentage in weight of the Al2Cu and Mg2Si phases due to the precipitation of their transient ", ' and β", β' states, respectively. It should be noted that the greater the dwell time at 495 °C, the greater the percentage in weight of the transient states associated with the Al2Cu phase. It can thus be deduced that Mg has a greater facility to dissolve during the solution treatment than Cu, the latter needing more time. The omega-Al7Cu2Fe phase was found to precipitate during the heat treatment as a result of the dissolution and atom diffusion of Cu [23], which is a transition phase during aging [24].   Figure 8 shows the diffractograms for the specimens in the form of 'wire', as well as after their solution heat treatment and natural ageing treatment. Figure 8a shows the diffractogram in the form of wire. Figure 8b-d respectively show the diffractograms obtained after dwell times of 4, 8, and 24 h at 495 °C and 1000 h of natural ageing. In the wire rod form, in addition to the α phase, the main identified phases were Al2Cu () and Mg2Si (β), although the probable presence of AlFe2 and AlCu3 was also detected. The solution treatment and natural ageing led to a different texturing of the constituent matrix (α phase), which is reflected in the variation in the relative heights of its Bragg peaks associated with an increase in the percentage in weight of the previously identified phases and the appearance of the Al7Cu2Fe phase. Table 2 shows the percentage in weight obtained by means of the reference-intensity ratio (RIR) method. Ageing leads to an important increase in the percentage in weight of the Al2Cu and Mg2Si phases due to the precipitation of their transient ", ' and β", β' states, respectively. It should be noted that the greater the dwell time at 495 °C, the greater the percentage in weight of the transient states associated with the Al2Cu phase. It can thus be deduced that Mg has a greater facility to dissolve during the solution treatment than Cu, the latter needing more time. The omega-Al7Cu2Fe phase was found to precipitate during the heat treatment as a result of the dissolution and atom diffusion of Cu [23], which is a transition phase during aging [24].   Figure 8a shows the diffractogram in the form of wire. Figure 8b-d respectively show the diffractograms obtained after dwell times of 4, 8, and 24 h at 495 • C and 1000 h of natural ageing. In the wire rod form, in addition to the α phase, the main identified phases were Al 2 Cu (θ) and Mg 2 Si (β), although the probable presence of AlFe 2 and AlCu 3 was also detected. The solution treatment and natural ageing led to a different texturing of the constituent matrix (α phase), which is reflected in the variation in the relative heights of its Bragg peaks associated with an increase in the percentage in weight of the previously identified phases and the appearance of the Al 7 Cu 2 Fe phase. Table 2 shows the percentage in weight obtained by means of the reference-intensity ratio (RIR) method. Ageing leads to an important increase in the percentage in weight of the Al 2 Cu and Mg 2 Si phases due to the precipitation of their transient θ", θ' and β", β' states, respectively. It should be noted that the greater the dwell time at 495 • C, the greater the percentage in weight of the transient states associated with the Al 2 Cu phase. It can thus be deduced that Mg has a greater facility to dissolve during the solution treatment than Cu, the latter needing more time. The omega-Al 7 Cu 2 Fe phase was found to precipitate during the heat treatment as a result of the dissolution and atom diffusion of Cu [23], which is a transition phase during aging [24].   Figure 9 shows the hardness profile obtained during the natural ageing, up to 1700 h, of the quenched specimens following the solution treatment at 495 °C. A rapid increase in hardness in the first hours of ageing is observed in all three cases. From a certain point onward, however, this increase leveled out at a much lower rate. This higher initial rate of increase in hardness is greater when the dwell time at 495 °C is 24 h. After 1700 h, the hardness obtained in the three samples was around 134 HV, the maximum hardness not seeming to be obtained in any of the three cases.   Figure 9 shows the hardness profile obtained during the natural ageing, up to 1700 h, of the quenched specimens following the solution treatment at 495 • C. A rapid increase in hardness in the first hours of ageing is observed in all three cases. From a certain point onward, however, this increase leveled out at a much lower rate. This higher initial rate of increase in hardness is greater when the dwell time at 495 • C is 24 h. After 1700 h, the hardness obtained in the three samples was around 134 HV, the maximum hardness not seeming to be obtained in any of the three cases.  Figure 10 shows the hardness profiles obtained in the four specimens kept at 495 °C for 4 h. It can be observed that the peak hardness is reached first at the highest ageing temperature. Nevertheless, the maximum value of hardness is obtained at the lowest temperature (160 °C). Figure  11 shows the hardness profiles of specimens kept at 495 °C for 8 h. It can likewise be observed that the peak hardness is reached first at the highest ageing temperature. In this case, however, the maximum hardness value at 180 °C coincides with that obtained at 160 °C. The maximum hardness value achieved was 162 HV, after a solution treatment at 495 °C for 4 h and ageing at 160 °C for 50 h.  Figures 12 and 13 show the hardness profiles of the specimens that were rolled prior to the artificial ageing treatments. Figure 12 shows the results obtained on the 4 specimens kept at 495 °C for 4 h, while Figure 13 shows the results obtained on the specimens that remained at this temperature for 8 h. In both cases, ageing does not take place at 200 °C, a continuous reduction in hardness being produced due to the recovery softening of the material at this temperature. The density of dislocations increases during cold rolling deformation, while two mechanisms compete during the subsequent ageing treatment: one in which the dislocations act like nucleating agents for  Figure 10 shows the hardness profiles obtained in the four specimens kept at 495 • C for 4 h. It can be observed that the peak hardness is reached first at the highest ageing temperature. Nevertheless, the maximum value of hardness is obtained at the lowest temperature (160 • C). Figure 11 shows the hardness profiles of specimens kept at 495 • C for 8 h. It can likewise be observed that the peak hardness is reached first at the highest ageing temperature. In this case, however, the maximum hardness value at 180 • C coincides with that obtained at 160 • C. The maximum hardness value achieved was 162 HV, after a solution treatment at 495 • C for 4 h and ageing at 160 • C for 50 h.  Figure 10 shows the hardness profiles obtained in the four specimens kept at 495 °C for 4 h. It can be observed that the peak hardness is reached first at the highest ageing temperature. Nevertheless, the maximum value of hardness is obtained at the lowest temperature (160 °C). Figure  11 shows the hardness profiles of specimens kept at 495 °C for 8 h. It can likewise be observed that the peak hardness is reached first at the highest ageing temperature. In this case, however, the maximum hardness value at 180 °C coincides with that obtained at 160 °C. The maximum hardness value achieved was 162 HV, after a solution treatment at 495 °C for 4 h and ageing at 160 °C for 50 h.  Figures 12 and 13 show the hardness profiles of the specimens that were rolled prior to the artificial ageing treatments. Figure 12 shows the results obtained on the 4 specimens kept at 495 °C for 4 h, while Figure 13 shows the results obtained on the specimens that remained at this temperature for 8 h. In both cases, ageing does not take place at 200 °C, a continuous reduction in hardness being produced due to the recovery softening of the material at this temperature. The density of dislocations increases during cold rolling deformation, while two mechanisms compete during the subsequent ageing treatment: one in which the dislocations act like nucleating agents for  Figures 12 and 13 show the hardness profiles of the specimens that were rolled prior to the artificial ageing treatments. Figure 12 shows the results obtained on the 4 specimens kept at 495 • C for 4 h, while Figure 13 shows the results obtained on the specimens that remained at this temperature for 8 h. In both cases, ageing does not take place at 200 • C, a continuous reduction in hardness being produced due to the recovery softening of the material at this temperature. The density of dislocations increases during cold rolling deformation, while two mechanisms compete during the subsequent ageing treatment: one in which the dislocations act like nucleating agents for precipitation, and another one in which the dislocations are reorganized promoting recovery softening [21]. Cieslar et al., report 200 • C as the temperature above which the latter mechanism could prevail [25]. Furthermore, the peak hardness is reached before with ageing at 180 • C than at 160 • C, both values being very similar at just over 165 HV regardless of the duration of the solution treatment at 495 • C. These values slightly exceed those obtained without prior deformation. However, it should be kept in mind that the hardness of the material after 25% deformation is 150 HV; hence. This means an increase in hardness of over 50%, which is related to the increase in dislocation density. The increase in hardness due to ageing treatment is less than 20 HV in all cases. If we compare this result with those obtained in the samples that did not undergo deformation, this very low increase in hardness can be attributed to competition between the two mechanisms described above and to the decrease in the concentration of vacancies resulting from the increase in dislocations, thereby decreasing the density of GP zones [13]. No less important is the finding that if ageing were carried out at 180 • C, the dwell time needed to reach the peak hardness would be 6 h. However, to obtain the maximum hardness without prior deformation, 50 h at 160 • C ( Figure 9) would be necessary. precipitation, and another one in which the dislocations are reorganized promoting recovery softening [21]. Cieslar et al., report 200 °C as the temperature above which the latter mechanism could prevail [25]. Furthermore, the peak hardness is reached before with ageing at 180 °C than at 160 °C, both values being very similar at just over 165 HV regardless of the duration of the solution treatment at 495 °C. These values slightly exceed those obtained without prior deformation. However, it should be kept in mind that the hardness of the material after 25% deformation is 150 HV; hence. This means an increase in hardness of over 50%, which is related to the increase in dislocation density. The increase in hardness due to ageing treatment is less than 20 HV in all cases. If we compare this result with those obtained in the samples that did not undergo deformation, this very low increase in hardness can be attributed to competition between the two mechanisms described above and to the decrease in the concentration of vacancies resulting from the increase in dislocations, thereby decreasing the density of GP zones [13]. No less important is the finding that if ageing were carried out at 180 °C, the dwell time needed to reach the peak hardness would be 6 h. However, to obtain the maximum hardness without prior deformation, 50 h at 160 °C ( Figure 9) would be necessary.   precipitation, and another one in which the dislocations are reorganized promoting recovery softening [21]. Cieslar et al., report 200 °C as the temperature above which the latter mechanism could prevail [25]. Furthermore, the peak hardness is reached before with ageing at 180 °C than at 160 °C, both values being very similar at just over 165 HV regardless of the duration of the solution treatment at 495 °C. These values slightly exceed those obtained without prior deformation. However, it should be kept in mind that the hardness of the material after 25% deformation is 150 HV; hence. This means an increase in hardness of over 50%, which is related to the increase in dislocation density. The increase in hardness due to ageing treatment is less than 20 HV in all cases. If we compare this result with those obtained in the samples that did not undergo deformation, this very low increase in hardness can be attributed to competition between the two mechanisms described above and to the decrease in the concentration of vacancies resulting from the increase in dislocations, thereby decreasing the density of GP zones [13]. No less important is the finding that if ageing were carried out at 180 °C, the dwell time needed to reach the peak hardness would be 6 h. However, to obtain the maximum hardness without prior deformation, 50 h at 160 °C ( Figure 9) would be necessary.

Conclusions
The effect on the hardness of an Al-Cu4.5-Mg1.5-Si0.75 alloy, manufactured by continuous casting and hot-milled until achieving a reduction in its cross-section greater than 90%, of the dwell time at a temperature of 495 °C (4, 8, and 24 h), of the ageing temperature (160, 180, and 200 °C), and of milling prior to ageing was analyzed.
As regards the dwell time at the solution temperature, it is concluded that: 1. The greater the dwell time at 495 °C, the greater the percentage in weight of the transient states associated with the Al2Cu phase during natural ageing. 2. Mg is easier to dissolve during the solution treatment than Cu, the latter needing more time. 3. Precipitation of the Al7Cu2Fe phase is observed during natural ageing, its content being reduced with increasing dwell time at the solution temperature. 4. Regardless of the dwell time at the solution temperature, the maximum hardness obtained during natural ageing only slightly exceeds 130 HV.
As to the artificial ageing temperature, it is concluded that: 1. The higher the ageing temperature, the faster the peak hardness is reached, but the lower its value. The ageing temperature that allows a higher hardness value to be obtained is 160 °C. 2. The maximum hardness value obtained was 162 HV, after a solution treatment at 495 °C for 4 h and ageing at 160 °C for 50 h.
If the material is cold milled with a true deformation of 25% prior to the ageing treatment, it is concluded that: 1. During artificial ageing in the 160-180 °C range:


The peak hardness values are reached more quickly.  The hardness values exceed those achieved without prior deformation. However, this result is due to the hardness associated with the deformation, as the increase in hardness due to ageing was less than 20 HV in all cases.

Conclusions
The effect on the hardness of an Al-Cu4.5-Mg1.5-Si0.75 alloy, manufactured by continuous casting and hot-milled until achieving a reduction in its cross-section greater than 90%, of the dwell time at a temperature of 495 • C (4, 8, and 24 h), of the ageing temperature (160, 180, and 200 • C), and of milling prior to ageing was analyzed.
As regards the dwell time at the solution temperature, it is concluded that: 1. The greater the dwell time at 495 • C, the greater the percentage in weight of the transient states associated with the Al 2 Cu phase during natural ageing.

2.
Mg is easier to dissolve during the solution treatment than Cu, the latter needing more time.

3.
Precipitation of the Al 7 Cu 2 Fe phase is observed during natural ageing, its content being reduced with increasing dwell time at the solution temperature.

4.
Regardless of the dwell time at the solution temperature, the maximum hardness obtained during natural ageing only slightly exceeds 130 HV.
As to the artificial ageing temperature, it is concluded that: 1.
The higher the ageing temperature, the faster the peak hardness is reached, but the lower its value. The ageing temperature that allows a higher hardness value to be obtained is 160 • C.

2.
The maximum hardness value obtained was 162 HV, after a solution treatment at 495 • C for 4 h and ageing at 160 • C for 50 h.
If the material is cold milled with a true deformation of 25% prior to the ageing treatment, it is concluded that:

1.
During artificial ageing in the 160-180 • C range: • The peak hardness values are reached more quickly.

•
The hardness values exceed those achieved without prior deformation. However, this result is due to the hardness associated with the deformation, as the increase in hardness due to ageing was less than 20 HV in all cases.

•
The maximum hardness values are the same regardless of the ageing temperature. The maximum hardness obtained was 168 HV when employing a dwell time of 8 h at