Microstructure Optimization of Mg-Alloys by the ECAP Process Including Numerical Simulation, SPD Treatments, Characterization, and Hydrogen Sorption Properties

Abstract

Both numerical simulation and hardness measurements were used to determine the mechanical and microstructural behavior of AZ31 bulk samples when submitted to the Equal Channel Angular Pressing (ECAP) technique. Billets of this representative of Mg-rich alloys were submitted to different numbers of passes for various ECAP modes (anisotropic A, isotropic B_C). The strain distribution, the grain size refinement, and the micro-hardness were used as indicators to quantify the effectiveness of the different processing routes. Structural characterizations at different scales were achieved using Scanning Electron Microscopy (SEM), micro-analysis, metallography, Small Angle Neutron Scattering SANS, X-Ray Diffraction (XRD), and texture determination. The grain and crystallite size distribution and orientation as well as defect impacts were determined. Anelastic Spectroscopy (AS) on mechanically deformed samples have shown that the temperature of ECAP differentiate the fragile to ductile regime. MgH₂ consolidated powders were checked for using AS to detect potential hydrogen motions and interaction with host metal atoms. After further optimization, the different mechanically-treated samples were submitted to hydrogenation/dehydrogenation (H/D) cycles, which shows that, for a few passes, the B_C mode is better than the A one, as supported by theoretical and experimental microstructure analyses. Accordingly, the hydrogen uptake and (H/D) reactions were correlated with the optimized microstructure peculiarities and interpreted in terms of Johnson-Avrami- Mehl-Kolmogorov (JAMK) and Jander models, successively.

Keywords: ECAP process; anelastic spectroscopy; grain and crystallite size; hydrogen uptake and kinetics sorption; magnesium alloys; numerical simulation; severe plastic deformation; texture.

Figure 1

A picture of ECAP processed billet shows transformation of a reference square array drawn on the four 50 × 10 mm faces at the initial state.

Figure 2

Simulated distribution of strain intensities: (a) at 1/3 pass, (b) at 3/4 pass of the 1st ECAP pass (relative scales: blue = minimum strain, green = intermediate strain, red = maximum strain).

Figure 3

Microstructure of a pure Mg sample after 1 ECAP pass at RT. The white bar scale is for 20 µm.

Figure 4

Strain intensity distribution after 2 ECAP passes done: (a)—along mode A, (b)—along mode B_C, after 3 ECAP passes done: (c)—along mode A, (d)—along mode B_C. The relative scale is between minimum strain = blue, intermediate strain = green and maximum strain = red without any physical meaning.

Figure 5

Strain level (%) versus the strained relative volume of sample (%) after different ECAP processes—open circles (O): after a 1st pass, open squares (□): after a 2nd pass, billet is not rotated reference to the 1st pass, black circles (●): after a 2nd pass, with a rotation of φ = 90° reference to the 1st pass, black triangles (▲): after a 2nd pass, with a rotation of φ = 180°.

Figure 6

Strain level (%) versus the strained relative volume of sample (%) after the 3rd pass with different rotations relative to the 2nd pass. Black dots (●): optimal ECAP process (Figure 5) with rotation of φ = 90° relative to the 1st pass (В_С mode). Open squares (□): billet was not rotated relative to the 2nd pass. Black diamonds (♦): rotation is φ = 90° relative to the 2nd pass. Open triangles (∆) rotation is φ = 180° relative to the 2nd pass.

Figure 7

Impact of the number of ECAP passes on the grain size in a AZ31 billet as processed at 200 °C. It is worth it to note that, after 3 passes, a minimum grain size of 3 µm was achieved and then stabilized.

Figure 8

AZ31 microstructure: (a)—initial state, (b)—after 2 B_C passes at 200 °C.

Figure 9

Microhardness of the AZ31 alloy vs the increasing number of ECAP passes at 200 °C.

Figure 10

Dependence of the stress/strain response during ECAP compression of the AZ31 samples at RT: 1, as-cast (♦); 2, after 1st ECAP pass (■); 3, after 2nd ECAP pass (▲).

Figure 11

Strain level (%) versus the strained relative volume of the sample (%) versus the ECAP passes: diamonds (♦): 1st pass, triangles (▲): 2nd pass at φ = 90°, being the optimal combination, mode B_C. Black squares (■): 2nd pass at φ = 180°, mode A, black circles (●): 3rd pass at φ = 90° relative to the 2nd pass at φ = 90°, mode B_C.

Figure 12

Microstructure versus temperature of AZ31 billets after 2 B_C ECAP passes: (a)—as received, (b)—150 °C, (c)—200 °C, (d)—250 °C, (e)—300 °C.

Figure 13

Microstructure versus the number of B_C ECAP passes of AZ31 billets after ECAP B_C passes at 250 °C: (a)—2 passes, (b)—4 passes, (c)—5 passes, (d)—8 passes, (e)—9 passes.

Figure 14

Microstructure and twinning: (a): ZK60 for 1 pass A mode at 150° C: various fields of twins affecting large and well oriented grains, (b): AZ31 for 2 passes B_C mode at 150 °C where zones can be seen as twin orientations along the 2 successive passes directions of (1) and (2) and (3) perpendicular bi-twinned grains.

Figure 15

Williamson-Hall [44] plots allow for the study of the impact of temperature processing AZ31 billets submitted to A-type ECAP passes: (a)—marked strained material with a mean small size of crystallites of 300-400 nm for 1 pass at RT, (b)—2 passes at 250 K, which revealed almost no strain in spite of a marked anisotropic distribution and a minimum mean size of crystallites (<100 nm).

Figure 16

Williamson-Hall [44] plots allow for the determination of the crystallite size and the level of stresses leading modulation of the linewidth of XDR patterns versus the Bragg angle (scattering vector s) for ECAP-treated AZ31 billets at 250 °C: (a)—mode A - 9 passes, red: (dots and squares: see text), 9 passes + 1 pass at RT (blue), (b)—mode B_C - 9 passes (red), 9 passes + 1 pass at RT (blue).

Figure 17

Stereographic projections of pole figures for the planes: (002), (100), (101), and (110).

Figure 18

Pole figure of the (002) plane obtained after 2 passes mode B_C.

Figure 19

Texture integrated and normalized intensities of the four main crystallographic planes. (a)—after 1 pass (~A), (b)—after 9 passes B_C. Colors are for: blue (002), red (100), black (101), and green (110) lines.

Figure 20

Small angle neutron scattering records (halo) of 1 pass AZ31 at RT for (a)—39 and (b)—8 m as distances sample to detector.

Figure 21

Anelastic spectroscopy traces recorded under 5 Hz excitation, of the elastic energy dissipation factor Q⁻¹ (dots) and Young modulus E (open circles) versus temperature (heating = black, cooling down = red) shown e.g., for (a) (left) Mg#1 sample (225 °C, B_C ECAP treated) and (b) (right) for Mg#2 sample (175 °C, B_C ECAP treated).

Figure 22

Relative height of the two peaks found on cooling down branch Mg#2 4 B_C-175 °C. Lines are guides for the eyes (origin dots have no physical meaning).

Figure 23

Hydrogen absorption rate (H/M w%) versus time for the samples treated in mode A at 175 °C (black): (a)—for 3 passes, (b)—for 8 passes.

Figure 24

Hydrogen absorption rate (H/M w%) versus time at 275 °C (red) for 3 passes (triangles).

Figure 25

Hydrogen desorption rate (H/M w%) versus time of samples: (a)—for 3 passes (triangles) and for 8 passes (squares) as treated at 175 °C (black).; (b)—for 8 passes as treated at 275 °C (red).

Figure 26

Hydrogen absorption rate (H/M w%) for B_C ECAP- treated billets: (a)—3 passes at 175 °C (black) and 3 passes at 275 °C (red). (b)—for 8 passes at 275 °C (red). Relative gains (kinetics and uptake) from three to 8 passes is less than for the A mode as shown in Figure 24 because of a more homogeneous micro-structure in the billet.

Figure 27

Hydrogen desorption rate (H/M w%) of samples ECAP-treated at 175 °C—2 passes (triangles) and 8 passes (squares).

Figure 28

Hydrogen desorption rate (H/M w%) for: (a)—A-mode ECAP treated at 275 °C for 3 passes (b)—B_C- and A-mode ECAP-treated for 8 passes at 275 °C.

Figure 29

Comparison of the traces in terms of the 1st and 2nd absorption rates (H/M w%) of 8 passes ECAP-treated samples at 275 °C: (a)—mode A, (b)—mode B_C.

Figure 30

Second absorption rate traces (H/M w%) of: (a)—3 passes ECAP treated in B_C route at 175° and 275 °C: (b)—difference between traces A-B_C routes for 8 passes at 275 °C and for 3 passes B_C between 175 & 275 °C for similar results whatever the treatments are.

Figure 31

Comparison of the 2nd desorption rate traces (H/M w%) of: (a)—three passes B_C route ECAP treated at 175 and 275 °C. (b)—Eight passes ECAP-treated at 275 °C for A and B_C routes.

Figure 32

Numerical fits using Equation (1) at the beginning absorption for (2) and (3) at the end of absorption: (a)—for eight passes A mode treated at 275°C, the 1st cycle absorption, (b)—for three passes B_C mode treated at 175 °C, the 2nd cycle absorption.

Figure 33

Limit value ξ_L (%) of the change in kinetic process vs. log(DF). Blue dots are for ECAP mode A, red dots are for ECAP mode B_C. (a): for the 1st hydrogenation (b): for the 2nd hydrogenation.

Figure 34

ECAP tool used for SPD treatments to Mg alloys, as developed by POINSARD Design and Tool SAS (Besançon, France). Up and left: the open channeling die, down-left: tools forming different angle channels, Φ_E = 95°, 105°, 115°, 125°, and 135°, right: ECAP system (die, punch) [36].