Non-equilibrium reactions in 6xxx series Al alloys

Non-equilibrium reactions in 6xxx series Al alloys

Materials Science and Engineering A304–306 (2001) 119–124 Non-equilibrium reactions in 6xxx series Al alloys C. Hsu a , K.A.Q. O’Reilly a,∗ , B. Cant...

712KB Sizes 0 Downloads 3 Views

Materials Science and Engineering A304–306 (2001) 119–124

Non-equilibrium reactions in 6xxx series Al alloys C. Hsu a , K.A.Q. O’Reilly a,∗ , B. Cantor a , R. Hamerton b a

Oxford Centre for Advanced Materials and Composites, University of Oxford, Parks Road, Oxford OX1 3PH, UK b Alcan International Limited, Banbury Laboratories, Southam Road, Banbury OX16 7SP, UK

Abstract Laboratory scale techniques have been developed to investigate particular aspects of nucleation and growth during solidification. These techniques have been used to isolate and study under controlled conditions the nucleation and growth aspects of the solidification that may occur in industrial processes such as direct chill (DC) casting, where cooling rates and growth velocities can be relatively high. The resultant intermetallic phase selection has been studied in Al alloys, where a wide range of equilibrium and metastable intermetallics can form. This paper describes the use of the entrained droplet technique together with intermetallic phase identification by transmission electron microscopy (TEM) to study the solidification sequence in a model 6xxx series Al alloy. In the entrained droplet technique, 1–1000 nm liquid droplets are entrained in a solid matrix which acts as a heterogeneous nucleation catalyst, and their solidification is monitored using differential scanning calorimetry (DSC). In the majority of commercial Al alloys, solidification takes place by the formation of primary Al, followed by secondary eutectic and peritectic reactions to form small quantities of intermetallic particles in the interdendritic regions. In the model 6xxx series Al alloy, the solidification sequence observed using the entrained droplet technique after partially melting to 640◦ C and re-solidifying at a cooling rate of 5 K min−1 , showed that the cubic ␣c -AlFeSi phase was formed via three reactions: (i) the predicted equilibrium peritectic reaction L + Al13 Fe4 → ␣-Al + ␣c -AlFeSi; (ii) a non-equilibrium eutectic reaction L → ␣-Al + ␣c -AlFeSi; and (iii) a ternary eutectic reaction L → ␣-Al + ␣c -AlFeSi + Mg2 Si. Only the peritectic and non-equilibrium eutectic reactions were readily observed during solidification at a cooling rate of 5 K min−1 from the fully molten state. The literature suggests that all the three reactions may operate during commercial casting operations. © 2001 Elsevier Science B.V. All rights reserved. Keywords: 6xxx series Al alloy; Transmission electron microscopy; Differential scanning calorimetry

1. Introduction The nucleation of solidification usually takes place by a heterogeneous process. There are serious experimental difficulties in studying heterogeneous nucleation, because of the difficulty of excluding extraneous impurities which can act as uncontrolled catalysts leading to irreproducible results [1]. The entrained droplet technique was first devised by Wang and Smith [2] to improve the reproducibility of liquid undercooling measurements. Alloys are thermomechanically manipulated to produce a microstructure of low melting point particles embedded in a high melting point matrix, and are then heat treated to melt the particles and monitor their solidification behaviour during subsequent cooling. Cantor and co-workers [1,3] have shown that rapid solidification often produces entrained droplets, <100 nm in size, leading to temperature and undercooling measurements reproducible in some cases to better than 0.2◦ C. Early work using the entrained droplet technique focused on immiscible alloys [4], but more recently the technique ∗

Corresponding author.

has been used to study eutectic and peritectic systems [4], and now commercial alloys, where it has been primarily used to investigate the influence of trace alloying elements on heterogeneous nucleation [5,6]. 6xxx series Al alloys are commonly used in extruded form, and are being adopted increasingly in sheet form for automotive vehicle bodies [7–9]. The main alloying elements in 6xxx series Al alloys are Si and Mg. These alloying elements are partly dissolved in solid solution in the primary ␣-Al matrix, and partly present in the form of intermetallic phases. A range of different intermetallic phases may form during solidification, depending on alloy chemistry and solidification conditions [10]. Fe is present as an impurity in all commercial alloys, and forms a variety of Al–Fe and Al–Fe–Si intermetallic phases during solidification. Any Si which is not incorporated in the ␣-Al matrix or the Al–Fe–Si intermetallic phases combines with Mg to form Mg2 Si at a later stage in the solidification process. The type, size, morphology and distribution of the intermetallic particles are very important in determining the subsequent material properties. The nature of the intermetallic particle population in the final product will be determined not

0921-5093/01/$ – see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 1 - 5 0 9 3 ( 0 0 ) 0 1 4 6 7 - 2

120

C. Hsu et al. / Materials Science and Engineering A304–306 (2001) 119–124

only by the as-cast state, but also by subsequent ingot homogenisation and thermomechanical processing. This paper presents the results of an investigation into the use of the entrained droplet technique to study the intermetallic phases formed during solidification of a model 6xxx series Al alloy.

2. Experimental A model 6xxx series Al alloy was investigated which contained ∼0.3 wt.%Fe, 0.6 wt.%Si and 0.8 wt.%Mg. The alloy was manufactured from 99.997% pure Al, 99.999% pure Si, 99.9% pure Mg and a master alloy of Al–5 wt.%Fe by melting in a graphite crucible, followed by casting as discs 5.5 cm in diameter and 9 mm in thickness. Specimens of the alloy were then rapidly solidified by melt spinning to provide chemical homogeneity and to ensure that a uniform distribution of fine scale intermetallic phases would be obtained for the entrained droplet technique. Alloy charges of ∼4 g were induction remelted in a He atmosphere in a quartz nozzle using a heating rate of ∼100 K s−1 to a temperature of ∼800◦ C, and were then ejected with an Ar overpressure of 200 mbar through a ∼0.85 mm diameter orifice onto the outer surface of a polished Cu wheel rotating with a tangential surface speed of 30 m s−1 . This produced ductile crystalline melt spun ribbons, typically 50–150 ␮m in thickness and 2.5 mm in width. The solidification and melting behaviour of the alloy was investigated in a Mettler Toledo DSC821e heat flux compensated differential scanning calorimeter (DSC). Individual ∼5 mg samples of as-melt spun ribbon were sandwiched between two Mo discs and crimped into a Cu DSC pan. A similar pan containing two Mo discs was used as a reference. All experiments took place at atmospheric pressure under a dynamic Ar atmosphere. The fine, rapidly solidified melt spun microstructure was partially melted in the DSC by heating from 550 to 640◦ C at a heating rate of 5 K min−1 with the matrix remaining solid, i.e. the intermetallic phases were melted, producing a dispersion of liquid droplets entrained in the primary ␣-Al solid matrix. The heterogeneous nucleation of solidification of these dispersed droplets was then monitored calorimetrically during cooling to 550◦ C at a cooling rate of 5 K min−1 . Quench experiments were carried out to interrupt the cooling procedure in order to identify the particular intermetallic phases corresponding to DSC peaks during solidification. The quench experiments were limited by the time taken to open the DSC, remove the specimen and put it into the water quench medium. Hence, only two temperatures, 610 and 580◦ C, were selected. A cooling rate of 40 K min−1 was used in conjunction with a quench temperature of 580◦ C. DSC treated specimens were then jet-electropolished in a solution of 25 vol.% nitric acid and 75 vol.% methanol at about −40◦ C for microstructural examination and diffraction analysis in JEOL 200CX and Philips CM20 trans-

mission electron microscopes (TEMs) fitted with energy dispersive X-ray (EDX) detectors.

3. Results Fig. 1 shows the solidification DSC traces at cooling rates of 5 K min−1 from the alloy after full melting and after partial melting to 640◦ C at heating rates of 5 K min−1 . After full melting, four solidification peaks were observed with peak temperatures of approximately 651, 631, 627 and 611◦ C. The full melting and solidification experiments were repeated five times (not shown) and were found to give reproducible results to within ±0.2 K for the first and second peaks, and ±0.7–0.8 K for the third and fourth peaks. The first large solidification peak corresponds to solidification of the ␣-Al matrix, and the three lower temperature peaks correspond to the solidification of intermetallic phases. After partial melting, at first sight there appears to be five solidification peaks. However, the first observed peak is a DSC transient arising from switching from heating to cooling, and can be ignored. Hence there are four solidification peaks, with peak temperatures of approximately 625, 617, 593 and 586◦ C, corresponding to the solidification of intermetallic phases. The partial melting and solidification experiments

Fig. 1. Re-solidification DSC traces at a cooling rate of 5 K min−1 after full melting (grey) and after partial melting up to 640◦ C (black) at a heating rate of 5 K min−1 .

C. Hsu et al. / Materials Science and Engineering A304–306 (2001) 119–124

121

Fig. 2. A typical Al13 Fe4 particle in an ␣-Al grain observed after partial melting up to 640◦ C at a heating rate of 5 K min−1 , re-solidifying to 610◦ C at a cooling rate of 5 min−1 and then quenching into water: (a) bright field TEM micrograph; (b) SADP from the particle; and (c) diffraction pattern analysis.

Fig. 3. A typical faceted ␣c -AlFeSi particle in an ␣-Al grain observed after partial melting up to 640◦ C at a heating rate of 5 K min−1 , re-solidifying to 610◦ C at a cooling rate of 5 min−1 and then quenching into water: (a) bright field TEM micrograph; (b) SADP from the particle; and (c) diffraction pattern analysis.

were repeated three times (not shown) and were found to give reproducible results to within ±0.5–1.0 K. To facilitate identification of the transformations corresponding to each DSC solidification peak after partial melting, quench experiments were carried out from 610 and 580◦ C, i.e. after the second and fourth solidification peaks. Three types of particles were observed after quenching from 610◦ C and typical particles are shown in Figs. 2–4. Fig. 2a shows a particle inside a grain. The particle was ∼0.7 ␮m in width and ∼1.5 ␮m in length. EDX indicated that the particle was rich in Al, Fe and Si with the atomic ratio of Al:Fe:Si estimated to be approximately 10:4:1. Fig. 2b–c shows the corresponding selected area diffraction pattern (SADP) from the particle which indexes as a [0 0 1] zone of Al13 Fe4 . Fig. 3a shows another particle inside a grain. This particle was ∼0.6 ␮m in width and ∼2.5 ␮m in length. EDX again indicated that the particle was rich in Al, Fe and Si, with the atomic ratio of Al:Fe:Si approximately 8:5:1. Fig. 3b–c shows the corresponding SADP from the particle which indexes as a [1¯ 3 1] zone of ␣c -AlFeSi. Fig. 4a shows a cluster of particles, with an overall cluster size of ∼1 ␮m in diameter. Each particle in the cluster had an average size of ∼0.15 ␮m in diameter. Fig. 3b shows the EDX trace from a single particle in the cluster [arrowed in Fig. 4a], again indicating that the particle was rich in Al, Fe and Si, with the atomic ratio of Al:Fe:Si approximately 9:5:1. There is no corresponding SADP taken from the

Fig. 4. A typical cluster of ␣c -AlFeSi particles in an ␣-Al grain observed after partial melting up to 640◦ C at a heating rate of 5 K min−1 , re-solidifying to 610◦ C at a cooling rate of 5 min−1 and then quenching into water: (a) bright field TEM micrograph; (b) EDX from a single particle in the cluster (marked by arrow).

122

C. Hsu et al. / Materials Science and Engineering A304–306 (2001) 119–124

Fig. 5. A typical faceted ␣c -AlFeSi particle in an ␣-Al grain observed after partial melting up to 640◦ C at a heating rate of 5 K min−1 , re-solidifying to 580◦ C at a cooling rate of 40 min−1 and then quenching into water: (a) bright field TEM micrograph; (b) SADP from the particle; and (c) diffraction pattern analysis.

particle, because the overall cluster of particles was too thick to allow high quality diffraction patterns. At first sight, two types of particles were observed after quenching from 580◦ C and typical particles are shown in Figs. 5 and 6. Fig. 5a shows a faceted particle with an overall size of ∼1.6 ␮m in diameter. EDX again indicated that the particle was rich in Al, Fe and Si, with the atomic ratio of Al:Fe:Si approximately 7.5:4.5:1. Fig. 5b–c shows the corresponding SADP from the particle which indexes as a [0 1 0] zone of ␣c -AlFeSi. Fig. 6a shows a cluster of particles, forming a cluster structure of overall size of ∼1.0 ␮m in diameter. EDX from particle A in the cluster indicated that the particle was again rich in Al, Fe, and Si, with the atomic ratio of Al:Fe:Si approximately 7:4:1. Approximately nine such AlFeSi particles were observed in the cluster. Fig. 6b shows that particle B, with a diameter of ∼0.3 ␮m, in the cluster was rich in Mg and Si, and the ratio of Mg:Si was approximately equal to 1:4. Despite the high Si level, the particle was assumed to be Mg2 Si, with part of the Si signal resulting from the matrix surrounding the particle. Mg2 Si particles were not readily observed in the clusters analysed after quenching from 610◦ C. This result suggested that in the final stages of solidification Mg and Si were rejected into the regions between AlFeSi particles, favouring the formation of Mg2 Si.

Fig. 6. A typical cluster of ␣c -AlFeSi particles in an ␣-Al grain observed after partial melting up to 640◦ C at a heating rate of 5 K min−1 , re-solidifying to 580◦ C at a cooling rate of 40 min−1 and then quenching into water: (a) bright field TEM micrograph; (b) EDX from particle B in the cluster (marked by arrow).

Fig. 7. The number of particles within grains observed and analysed after partial melting to 640◦ C at a heating rate of 5 K min−1 , re-solidifying to (a) 610◦ C at a cooling rate of 5 K min−1 and (b) 580◦ C at a cooling rate of 40 K min−1 and then quenching into water.

C. Hsu et al. / Materials Science and Engineering A304–306 (2001) 119–124

Fig. 7a and b shows histograms summarising the number of different particles observed and analysed after quenching from 610 and 580◦ C, respectively. It should be noted that no Mg2 Si was observed after quenching from 610◦ C, and no Al13 Fe4 was observed after quenching from 580◦ C.

4. Discussion 4.1. Solidification sequence of 6xxx series Al alloys The solidification sequence of 6xxx series Al alloys has been proposed by Mulazimoglu et al. [11]. The phases formed include the ␣-Al matrix, ␣c -AlFeSi, ␤-AlFeSi and Mg2 Si, but there is controversy about the presence of Al13 Fe4 . Backerud [12] suggested that Al13 Fe4 particles are formed but easily transform into ␣c -AlFeSi, while Mulazimoglu et al. [11] proposed that Al13 Fe4 does not form at all because it is difficult to nucleate at high cooling rates. The presence or absence of Al13 Fe4 is important in determining the type of reaction which leads to the formation of the ␣c -AlFeSi. If Al13 Fe4 is present, the equilibrium phase diagram predicts that ␣c -AlFeSi will be formed via a peritectic reaction. However, if Al13 Fe4 is not present, ␣c -AlFeSi may form via the eutectic reaction proposed by Mulazimoglu et al. [11]. One possible interpretation is that ␣c -AlFeSi particles can form by either peritectic or eutectic reactions, in different regions, depending on the local melt composition. In other words, there are two possible solidification paths to form ␣c -AlFeSi particles as shown in Fig. 8. During solidification, the composition of some of the liquid can reach the L → ␣-Al + Al13 Fe4 eutectic valley, and Al13 Fe4 is then formed by a eutectic reaction. The composition of the remaining liquid follows the eutectic valley and reaches the peritectic point, L + Al3 Fe3 → ␣-Al + ␣c -AlFeSi to form ␣c -AlFeSi particles. However, the composition of some of the other liquid could reach the L → ␣-Al + ␣c -AlFeSi eutectic valley directly, and ␣c -AlFeSi particles could then form by the L → ␣-Al + ␣c − AlFeSi eutectic reaction. Both types of ␣c -AlFeSi particles, eutectic and peritectic, can transform into ␤-AlFeSi particles by a subsequent peritectic reaction

Fig. 8. Schematic liquidus projection near the Al corner in the Al–Fe–Si phase diagram showing two possible solidification paths to form ␣c -AlFeSi particles.

123

L + ␣c -AlFeSi → ␣-Al + ␤-AlFeSi. Finally, Mg2 Si particles can subsequently form by the eutectic reaction L → ␣-Al + Mg2 Si or by precipitation in the solid state during cooling. 4.2. Solidification after partial melting Overall, four solidification peaks were observed during re-solidification after partial melting as shown in Fig. 1. Two reactions took place above 610◦ C, and the other two reactions took place between 610 and 580◦ C. The phases corresponding to the re-solidification peaks above 610◦ C have been identified from the quenching experiments as shown in Figs. 2–4. The phases observed after solidification was interrupted by quenching from 610◦ C include Al13 Fe4 , faceted ␣c -AlFeSi and clustered ␣c -AlFeSi. Hence, the Al13 Fe4 phase can form during re-solidification after partial melting. According to the Al–Fe–Si phase diagram, the Al13 Fe4 particles should be the first secondary phase to form during solidification. Therefore, the first re-solidification peak almost certainly corresponds to the formation of the Al13 Fe4 particles. The clusters of fine scale particles shown in Fig. 4 were ␣c -AlFeSi particles. These clustered particles were also observed in a grain refined alloy, where they were identified to be ␣c -AlFeSi by microdiffraction [13]. The ␣c -AlFeSi phase can also be present as larger, separate faceted particles, as shown in Fig. 3. The same mixture of clustered and faceted ␣c -AlFeSi particles was also observed in Al–0.27 to 0.51 wt.%Fe–0.1 wt.%Si alloys [14]. The two morphologies of ␣c -AlFeSi particles indicate that ␣c -AlFeSi can form via two different reactions. The clustered morphology of the ␣c -AlFeSi particles may suggest that they formed by a eutectic reaction and that they nucleated on and then grew cooperatively with the ␣-Al matrix. During re-solidification after partial melting, the liquid droplets which formed clustered ␣c -AlFeSi particles solidified by the eutectic reaction, L → ␣-Al + ␣c -AlFeSi. On the other hand, the morphology of the isolated faceted ␣c -AlFeSi particles may suggest that they formed by a peritectic reaction. During re-solidification after partial melting, the liquid droplets which form faceted ␣c -AlFeSi particles reach the required composition to form Al13 Fe4 particles first, and then transform into ␣c -AlFeSi particles by the peritectic reaction, L + Al13 Fe4 → ␣-Al + ␣c -AlFeSi. From the phase diagram, Fig. 8, peritectic ␣c -AlFeSi is expected to form at higher temperatures than eutectic ␣c -AlFeSi, as already discussed in Section 4.1. The phases found when solidification was interrupted by quenching from 580◦ C included faceted ␣c -AlFeSi particles and clustered ␣c -AlFeSi particles containing Mg2 Si particles, as shown in Figs. 5 and 6. Al13 Fe4 particles were not observed in samples after quenching from 580◦ C. This indicates that the Al13 Fe4 particles found at higher temperatures had fully transformed into faceted ␣c -AlFeSi particles in the temperature range 610–580◦ C.

124

C. Hsu et al. / Materials Science and Engineering A304–306 (2001) 119–124

The average size of the particles in each cluster of ␣c -AlFeSi particles was smaller when quenched from 610◦ C than when quenched from 580◦ C. This indicates that the clustered ␣c -AlFeSi particles form rapidly during the quench from 610◦ C, but form slowly during cooling before the quench from 580◦ C. In other words, the L → ␣-Al + ␣c -AlFeSi eutectic reaction takes place between 610 and 580◦ C. This confirms that the peritectic faceted ␣c -AlFeSi particles form earlier than the clustered eutectic ␣c -AlFeSi particles, as expected from the phase diagram. The second re-solidification peak corresponds to the formation of faceted peritectic ␣c -AlFeSi particles and the third re-solidification peak corresponds to the formation of clustered eutectic ␣c -AlFeSi particles. Mg2 Si particles appeared within the clustered ␣c -AlFeSi particles in the sample quenched from 580◦ C but not in the sample quenched from 610◦ C. This suggests that the Mg2 Si particles formed in the final solidification stage. Therefore, the third re-solidification peak corresponds to the formation of clustered eutectic ␣c -AlFeSi particles, and the fourth solidification peak corresponds to the formation of Mg2 Si particles. There are three possible mechanisms of forming the Mg2 Si particles. One possible mechanism is via the ternary eutectic reaction, L → ␣-Al + ␣c -AlFeSi + Mg2 Si. During partial melting and re-solidification, eutectic liquid droplets solidify to form clustered ␣c -AlFeSi particles and, in the final stage, the residual liquid solidifies to form ␣-Al, ␣c -AlFeSi and Mg2 Si. A second possible mechanism of forming Mg2 Si is in residual Mg and Si rich liquid in the final solidification stage via the binary eutectic reaction, L → ␣-Al + Mg2 Si. A third possible mechanism is via precipitation in the solid state from the supersaturated ␣-Al matrix. The presence of the Mg2 Si particles in fine scale clusters associated with ␣c -AlFeSi, but not in separate large or fine-scale particles within the ␣-Al matrix, strongly suggests that the first mechanism is dominant.

4.3. Solidification after full melting The highest temperature peak corresponds to the solidification of the ␣-Al matrix. The other three peaks correspond to the first three peaks during re-solidification after partial melting, i.e. they correspond to the formation of Al13 Fe4 , peritectic ␣c -AlFeSi and eutectic ␣c -AlFeSi, respectively. However, the peak temperatures are higher than the corresponding re-solidification peak temperatures after partial melting as shown in Fig. 1, indicating that there is a kinetic element in nucleating the phases in partially melted droplets. The fourth peak is missing, so no Mg2 Si is observed on re-solidification after full melting, because Si tied up with Fe to form Al13 Fe4 , peritectic ␣c -AlFeSi and eutectic ␣c -AlFeSi. From the above discussion, it is suggested that the alloy solidifies by the following series of reactions:

L → ␣-Al L → ␣-Al + Al13 Fe4 L + Al13 Fe4 → ␣-Al + ␣c -AlFeSi L → ␣-Al + ␣c -AlFeSi L → ␣-Al + ␣c -AlFeSi + Mg2 Si

(after partial melting)

5. Conclusions A combination of the embedded droplet technique together with intermetallic phase identification by TEM has been used to study the solidification sequence in a model 6xxx series Al alloy. The solidification sequence observed using the entrained droplet technique, for a cooling rate of 5 K min−1 , showed that the cubic ␣c -AlFeSi phase was formed via three reactions: (i) the predicted equilibrium peritectic reaction L + Al13 Fe14 → ␣-Al + ␣c -AlFeSi; (ii) a non-equilibrium eutectic reaction L → ␣-Al + ␣c -AlFeSi; and (iii) a ternary eutectic reaction L → ␣-Al + ␣c -AlFeSi + Mg2 Si. Only the peritectic and non-equilibrium eutectic reactions were readily observed during solidification at a cooling rate of 5 K min−1 from the fully molten state. The literature suggests that all the three reactions may operate during commercial casting operations.

Acknowledgements The authors would like to thank Mr. John Worth of Alcan International Banbury Laboratories for casting the alloys, and one of us (C. Hsu) would like to thank the Chung Cheng Institute of Technology, Department of Defence, Taiwan, ROC, for financial support. References [1] K.I. Moore, D.L. Zhang, B. Cantor, Acta Metall. Mater. 38 (1990) 1327. [2] C.C. Wang, C.S. Smith, Trans. Metall. Soc. AIME 188 (1950) 136. [3] D.L. Zhang, K. Chattopadhyay, B. Cantor, J. Mater. Sci. 26 (1991) 1531–1544. [4] B. Cantor, K.A.Q. O’Reilly, Current Opinion in Solid State and Materials Science 2 (1997) 318. [5] C.R. Ho, B. Cantor, Acta Metall. Mater. 43 (1995) 3231. [6] C.M. Allen, K.A.Q. O’Reilly, P.V. Evans, B. Cantor, Acta Mater., in press. [7] T. Moons, P. Ratchev, P. De Smet, B. Verlinden, P. Van Houtte, Scripta Mater. 35 (1996) 939. [8] P.G. Bloeck, J. Timm, Aluminium 70th Annu. Publ. 1/2 (1994) 87. [9] J.T. Staley, D.J. Lege, J. Phys. IV 3C7 (1993) 179. [10] L.F. Mondolfo, Aluminium Alloys Structure and Properties, Butterworths, London, 1976, p. 787. [11] M.H. Mulazimoglu, A. Zaluska, J.E. Gruzleski, F. Paray, Met. Trans. A 27A (1996) 929. [12] L. Backerud, Solidification Characteristics of Aluminium Alloys 1 (1986) 75. [13] C. Hsu, D.Phil. Thesis, University of Oxford, 1999. [14] P. Liu, Key Eng. Mater. 44/45 (1990) 69.