Softening behaviour of Al–Zn–Mg alloys due to welding

Softening behaviour of Al–Zn–Mg alloys due to welding

Materials Science and Engineering A266 (1999) 198 – 204 Softening behaviour of Al–Zn–Mg alloys due to welding T. Ma, G. den Ouden * Department of Mat...

252KB Sizes 0 Downloads 17 Views

Materials Science and Engineering A266 (1999) 198 – 204

Softening behaviour of Al–Zn–Mg alloys due to welding T. Ma, G. den Ouden * Department of Materials Science and Engineering, Delft Uni6ersity of Technology, Postbus 5025, Rotterdamsweg 137, 2600 Delft GA, The Netherlands Received 1 September 1998; received in revised form 27 November 1998

Abstract The softening behaviour of two Al–Zn–Mg alloys (7020 and 7022) due to welding was studied by means of simulation and real welding experiments. It was found that the heat-affected zone of the alloys can be divided in two parts: the dissolution zone and the overageing zone. The dissolution zone is characterised by dissolution of the strengthening precipitates, whereas the overageing zone is characterised by growth of the strengthening precipitates. It appears that the 7022 alloy has a stronger softening tendency than the 7020 alloy in both the dissolution zone and the overageing zone. It was also found that the degree of softening depends on the heat input: a higher input leads to more severe softening and a wider softening area, located at a larger distance from the fusion boundary. The hardness in the heat-affected zone can be recovered by post-weld heat treatment, especially in the dissolution zone. As far as the recovery of the hardness is concerned, artificial ageing is more effective than natural ageing. © 1999 Elsevier Science S.A. All rights reserved. Keywords: Softening; Al–Zn–Mg alloys; Thermal simulation; Welding; Heat-affected zone; Precipitation

1. Introduction Al–Zn–Mg alloys are characterised by a specific precipitation behaviour. Generally speaking, these alloys have a relatively small tendency for overageing. Furthermore, they are assumed to be rather insensitive to quench rate [1], which is especially of importance in the case of welding. Within the Al– Zn – Mg alloy system the 7020 and 7022 alloys are relatively new aluminium alloys, which are being increasingly used in light structures, such as storage tanks of space rockets, military portable bridges, modern bicycle frames, etc. The 7020 alloy is a medium-strength alloy in the Al – Zn – Mg system and contains about 5.0 wt.% Zn and 1.5 wt.% Mg, while the high-strength 7022 alloy contains about 5.0 wt.% Zn, 3.5 wt.% Mg and 0.7 wt.% Cu. The main precipitates present in the 7020 and 7022 alloys are: Guinier – Preston (GP) zones, the transition phase h%, the intermetallic compound MgZn2 (h phase) and the intermetallic compound Al2Mg3Zn3 (t phase). * Corresponding author. Tel.: +31-15-2789111; fax: + 31-152786730.

The GP zones and the transition phase (h%), in the form of semi-coherent precipitates, are primarily responsible for hardening of the alloys. The intermetallic compound h and the intermetallic compound t are non-coherent precipitates. They are formed either by growth of the transition phase or by direct precipitation from the solid solution at high temperature and do not contribute to the hardening effect of the alloys [2–4]. During welding, different microstructural changes are likely to occur in the heat-affected zone (HAZ) of the weld, such as recrystallisation, grain growth and modifications of precipitates. With respect to strength, grain growth plays an important role in the case of non-heattreatable aluminium alloys, but is of minor importance in the case of heat-treatable aluminium alloys. In the case of heat-treatable aluminium alloys, the strength is mainly influenced by modification of precipitates [5]. The modification of precipitates involves the following four transformation reactions: 1. dissolution of precipitates; 2. growth or transformation of coherent precipitates to incoherent forms (in the following referred to as ‘overageing’);

0921-5093/99/$ - see front matter © 1999 Elsevier Science S.A. All rights reserved. PII: S 0 9 2 1 - 5 0 9 3 ( 9 9 ) 0 0 0 2 0 - 9

T. Ma, G. den Ouden / Materials Science and Engineering A266 (1999) 198–204

3. direct formation of coherent precipitates from the solid solution; 4. direct formation of incoherent precipitates from the solid solution at high temperature (in the following referred to as ‘harmful precipitation’). Each of these reactions has a specific effect on the mechanical properties of the material. Therefore, due to the existing relation between the mechanical properties and the metallurgical transformations, the mechanical properties can be used as an indicator of the transformation. Although the general transformation behaviour of Al–Zn–Mg alloys is known, detailed knowledge about HAZ softening in some individual alloys is lacking. The purpose of this paper is to obtain fundamental insight in the HAZ softening characteristics of the 7020 alloy and the 7022 alloy during arc welding, and to find out how the softening is affected by heat input and postweld heat treatment.

2. Experimental conditions

2.1. Materials The base materials used in this study are the commercial 7020 and 7022 alloys in the form of 12-mm thick plates in T6 condition. The chemical compositions of the materials are given in Table 1.

2.2. HAZ simulation HAZ simulation was carried out by imposing thermal cycles on a series of small specimens, having dimensions 10×10× 55 mm. The specimens were machined from the plate material, the longest dimension parallel to the rolling direction of the plate. Two sets of arc welding conditions were selected for simulation, corresponding to heat inputs of 590 and 1080 J/mm, respectively (Table 2). The HAZs in the case of both heat inputs were simulated by thermal cycles with peak temperatures of 150, 210, 230, 270, 300, 330, 350, 370, 390, 540 and 585°C respectively, which for the given heat input represent specific locations in the HAZ. The thermal cycles used in the HAZ simulation were calculated with the modified Rosenthal model (taking into account the finite plate width) for two-dimensional heat flow [6,7]. The simulations were carried

199

Table 2 Welding parameters Welding parameter

Low heat input conditions

High heat input conditions

Arc voltage (V) Arc current (A) Travel speed (cm/min) Shielding gas Gas flow rate (l/min) Filler wire, diameter (mm) Polarity Process efficiency (%) Heat input (J/mm)

27 200 55 Argon 15 AlMg5, 1.2 DCEP 65 590

27 200 30 Argon 15 AlMg5, 1.2 DCEP 65 1080

out in triplicate with the help of a physical simulator (Gleeble 1500/20). The properties of the HAZ are usually expressed either as a function of peak temperature or as a function of distance from the fusion boundary. In view of this, it is convenient to know the relation between the peak temperature and the distance from the fusion boundary. This relation was determined for the two heat inputs used in this work (590 and 1080 J/mm). The results are shown in Fig. 1.

2.3. Welding experiments To be able to compare the results obtained in the case of simulation with the results obtained in the case of real welding, a number of single V-butt welds were made in 12-mm thick plates, using the same welding conditions as those used to calculate the thermal simulation cycles (Table 2). Welding was carried out in a direction normal to the rolling direction of the plate, using a mechanised GMA welding system. Before welding the surface of the workpiece was cleaned by a stainless steel brush in order to remove the oxide layer.

Table 1 Chemical composition (wt.%) of the 7020 and 7022 alloys Alloy

Zn

Mg

Cu

Cr

Mn

Ti

Al

7020 7022

5.02 4.98

1.45 3.85

0.06 0.68

0.14 0.14

0.27 0.17

0.04 0.01

Bal. Bal.

Fig. 1. The relation between peak temperature and the distance from the fusion boundary for two heat inputs (590 and 1080 J/mm).

200

T. Ma, G. den Ouden / Materials Science and Engineering A266 (1999) 198–204

2.4. Heat treatment In order to examine the influence of heat treatment on the microstructure and properties of the materials, the simulated and welded specimens were subjected to one of the following heat treatments: 1. natural ageing (3 months at room temperature); 2. artificial ageing (24 h at 120°C); 3. step ageing (8 h at 100°C+24 h at 150°C).

2.5. Examination of hardness and microstructure Hardness measurements of both the simulated and welded specimens were carried out using a Vickers hardness machine (10 kg load). To study the microstructure of the simulated and welded specimens cross-sections were made, which were subsequently ground and etched using a suitable etchant (Keller). The microstructure was primarily examined by means of optical microscopy (Neophot 2). Additionally, transmission electron microscopy (TEM) was carried out using a Philips EM400. The thin foils required for the TEM examination were prepared by jet polishing.

3. Results and discussion

3.1. Softening of the HAZ In order to determine the effect of welding on the hardness of the 7020 and 7022 alloys, specimens of both alloys were simulated and examined following the procedure described in the previous section. In Fig. 2 the hardness of the alloys immediately after simulation is plotted as a function of the peak temperature for the situation corresponding with a heat input of 590 J/mm. The figure shows that for both alloys softening occurs when the peak temperature is higher than about 200°C, which is equivalent with softening in a zone having a width of about 13 mm adjacent to the fusion boundary. As already mentioned above, softening of the HAZ in heat-treatable aluminium alloys is directly related to modification of precipitates.

3.2. Modification of precipitates To be able to obtain additional information about the precipitation behaviour of the alloys due to welding, transmission electron microscopy (TEM) was carried out on both as-received and simulated specimens. Fig. 3 shows TEM micrographs of 7020 specimens simulated at different peak temperatures. In Fig. 3a the microstructure of the non-simulated parent metal is shown in which small precipitates are visible. During simulation up to a peak temperature of 210°C the

Fig. 2. Hardness as a function of peak temperature of simulated specimens directly after simulation: (a) 7020 alloy, (b) 7022 alloy. (heat input 590 J/mm).

precipitates remain virtually unchanged (Fig. 3b). However, simulation at a peak temperature of 300°C, leads to dissolution and growth of precipitates (Fig. 3c). When the peak temperature exceeds 390°C, all precipitates are dissolved (Fig. 3d). On the basis of these observations it may be concluded that the loss of hardness in the HAZ is due to dissolution and/or growth of precipitates. To separate the softening effect due to dissolution from that due to the growth of precipitates, simulated specimens of both alloys were given an artificial ageing treatment. The idea behind this approach is that dissolved precipitates will be re-formed to hardening precipitates by ageing. The hardness after artificial ageing is plotted as a function of peak temperature in Fig. 4. Fig. 4a shows the results obtained in the case of the 7020 alloy. It can be seen that at temperatures above about 350°C the hardness is fully recovered to the level of the parent metal by the ageing treatment. Apparently, hardening precipitates are formed from the solid solution under these conditions. In the temperature range between about 230 and 350°C the hardness is only partially recovered, which indicates that in this temperature range overageing (growth of precipitates) takes place. The most severe overageing occurs at peak

T. Ma, G. den Ouden / Materials Science and Engineering A266 (1999) 198–204

temperatures around 300°C. It should be noted that for both heat inputs overageing occurs within approximately the same peak temperature range, 230–350°C, and that the most severe overageing occurs at the same peak temperature (about 300°C). On the basis of the foregoing observations, it is possible to divide the HAZ into two sub-zones: the dissolution zone and the overageing zone. The boundary between the two sub-zones corresponds with a peak temperature of about 350°C. The results obtained in the case of the 7022 alloy are presented in Fig. 4b. The figure shows that overageing occurs in the peak temperature range between about 230 and 390°C. In the case of high heat input, the recovery of the hardness above 390°C seems to be incomplete, presumably due to harmful precipitation. The maximum rate of overageing occurs at a peak temperature of about 350°C. As in the case of the 7020 alloy, the HAZ of the 7022 alloy can also be divided into two sub-zones. However, it appears that the temperature dividing these sub-zones is less pronounced in the case of the 7022 alloy than in the case of the 7020

201

alloy, due to the fact that harmful precipitation occurs during the cooling part of the thermal cycle. The results obtained on both the 7020 alloy and the 7022 alloy show that the heat input does not have a measurable influence on the location and size of the sub-zones, but significantly affects the magnitude of overageing (especially in the case of the 7022 alloy). In order to find out at which peak temperature the precipitates of the parent metal will be completely dissolved, specimens of both alloys were simulated at high peak temperatures (above 330°C), cooled down at fast cooling rate to avoid the possible influence of harmful precipitation, and artificially aged. After this treatment the hardness was measured. It was found that complete recovery of the hardness occurs at peak temperatures higher than 350°C in the case of the 7020 alloy and at peak temperatures higher than 410°C in the case of the 7022 alloy. This implies that the temperature which divides the dissolution zone and the overageing zone is 350°C for the 7020 alloy and 410°C for the 7022 alloy.

Fig. 3. TEM micrographs of simulated 7020 specimens directly after simulation: (a) parent metal, (b) peak temperature 210°C, (c) peak temperature 300°C and (d) peak temperature 390°C.

202

T. Ma, G. den Ouden / Materials Science and Engineering A266 (1999) 198–204

3.4. Quenching rate sensiti6ity It is evident that the loss in hardness of the HAZ due to welding can be fully recovered by a solution treatment, followed by quenching and subsequent ageing. However, this treatment is only successful when the quenching rate is sufficiently high. It is generally assumed that Al–Zn–Mg alloys have a limited quenching rate sensitivity. However, on the basis of preliminary experiments it was found that in the case of the 7022 alloy, the quenching rate can play an important role. In order to check the quench sensitivity of the 7020 and the 7022 alloys, some simulated specimens were subjected to a solution treatment (500°C, 30 min), followed by air cooling or water quenching, respectively. After this, the specimens were treated by artificial ageing. The hardness of the specimens treated in this way is shown in Fig. 6. In this figure, the hardness is plotted as a function of the distance to the fusion boundary. For comparison, the hardness of the specimens treated only by artificial ageing is also plotted in the figure. The results presented in Fig. 6 clearly show that air cooling after solution treatment is not suitable

Fig. 4. Hardness as a function of peak temperature of simulated specimens for two welding conditions. After simulation, the specimens were treated by artificial ageing: (a) 7020 alloy, (b) 7022 alloy.

3.3. Influence of post-weld heat treatment As pointed out in the previous section, part of the hardness loss in the HAZ can be recovered by an ageing treatment. In order to sort out the effect of ageing in more detail the hardness of the simulated HAZ was measured after different ageing treatments. In Fig. 5 the hardness of the simulated HAZ is plotted as a function of the distance from the fusion boundary for different heat treatment conditions. The figures show that the different post-weld heat treatments have similar effects on both the 7020 and the 7022 alloys. Generally speaking, artificial ageing and step ageing result in the best hardness recovery in the high peak temperature area (close to the fusion boundary). It appears that the possibility of property recovery through natural ageing is limited compared with the other two artificial ageing treatments. Even after 3 months the hardness of the naturally aged specimen still has not reached the level obtained by artificial ageing. As can be seen, step ageing leads to some overageing of the unaffected zone, which is presumably due to the relatively high ageing temperature during the last step of this treatment.

Fig. 5. Hardness as a function of the distance from the fusion line for specimens simulated with fast cooling rate followed by artificial ageing: (a) 7020 alloy, (b) 7022 alloy.

T. Ma, G. den Ouden / Materials Science and Engineering A266 (1999) 198–204

203

input results in more severe softening and a wider softening area which is shifted farther away from the fusion boundary. This can be easily understood by realising that the cooling rate decreases with increasing heat input [8], as a consequence of which more time is available for overageing. Furthermore, a higher heat input causes a wider overageing zone, which is shifted farther away from the fusion boundary.

3.6. Verification of HAZ simulation To be able to compare the results obtained by means of simulation with the results obtained by real welding, a series of GMA welding experiments was carried out under identical thermal conditions (Table 2). In Fig. 8 the hardness profiles of both the simulated and the welded material for the two alloys are presented (low heat input conditions). It appears that excellent agreement exists between the results of simulation and the results of real welding, which can be considered as a justification of the simulation approach.

Fig. 6. Hardness as a function of the distance from the fusion boundary of simulated specimens after solution treatment followed by air cooling or water quenching, and subsequently artificially aged: (a) 7020 alloy, (b) 7022 alloy.

for these two alloys to improve the post-weld strength. In fact, air cooling seems too slow to maintain the supersaturated structure. Fig. 6 also shows that the hardness of the 7022 specimens cooled in air is even lower than that of the 7020 specimens cooled in air, which indicates that the 7022 alloy more easily forms the non-hardening phases and, hence, is more sensitive to quenching rate than the 7020 alloy.

3.5. Influence of heat input The influence of the heat input on the softening behaviour of the 7020 and 7022 alloys is illustrated in Fig. 4. In this figure, the hardness after simulation followed by artificial ageing is plotted as a function of the peak temperature of the simulated cycle for two different heat inputs (1080 and 590 J/mm). The influence of heat input on the softening profile becomes more apparent when the data are plotted as a function of the distance to the fusion boundary. This is done in Fig. 7. The results presented in this figure indicate that the heat input has a significant influence on overageing in terms of location, extent and severity (7022 alloy). More specifically, it can be concluded that a higher heat

Fig. 7. Hardness as a function of the distance from the fusion boundary after simulation at two different heat inputs followed by artificial ageing: (a) 7020 alloy, (b) 7022 alloy.

204

T. Ma, G. den Ouden / Materials Science and Engineering A266 (1999) 198–204 “

“

“

“

“

The dissolution zone is characterised by dissolution of precipitates and covers the peak temperature range above 350°C in the case of the 7020 alloy, and the peak temperature range above 410°C in the case of the 7020 alloy. The overageing zone is characterised by growth of precipitates and covers the peak temperature range between 230 and 350°C in the case of the 7020 alloy, and the peak temperature range between 230 and 410°C in the case of the 7022 alloy. The 7022 alloy has a stronger softening tendency than the 7020 alloy in both the dissolution zone and in the overageing zone. The heat input affects the degree of softening: a higher input leads to more severe softening and a wider softening area, located at a larger distance from the fusion boundary. HAZ hardness can be recovered by post-weld heat treatment. As far as the recovery of the hardness is concerned, artificial ageing is more effective than natural ageing.

References Fig. 8. Hardness profiles of simulated and welded material (low heat input conditions): (a) 7020 alloy, (b) 7022 alloy.

4. Conclusions The results presented in this paper lead to the following conclusions. “ The HAZ of the 7020 and 7022 alloys can be divided into two sub-zones according to their different mechanism of softening: the dissolution zone and the overageing zone.

.

[1] W.A. French, T.N. Baker, Weld. Res. Int. 8 (1978) 1. [2] N. Ryum, Z. Metallk. 66 (1975) 377. [3] M.P. Bartle, in: R.G. Baker (Ed.), Proceedings of Conference on Weldable Al – Zn – Mg Alloys, The Welding Institute, Cambridge, 1970, p. 24. [4] A. Umgeher, H. Cerjac, in: S.A. David, J.M. Vitek (Eds.), Proceedings of 2nd International Conference on Trends in Welding Research, ASM International, Materials Park, OH, 1989, p. 279. [5] G.T. Hahn, A.R. Rosenfield, Metall. Trans. 6A (1975) 653. [6] D. Rosenthal, Weld. J. 20 (1941) 220s. [7] D. Rosenthal, Trans. ASME 68 (1946) 849. [8] K.E. Easterling, Introduction to the Physical Metallurgy of Welding, 2nd edn, Butterworth Heinemann, London, 1992.