Fig. 1. Tensile creep test at constant load on pure 2Al at initial stress of 15 MPa and 473 K: (a) (true) strain εε versus time t, (b) strain rate ε ? versus strain εε; gray dots: original data, black: reduced and smoothed by SmooMuDS[23] with Δε≈0.002Δε≈0.002, (c) slope n′ versus strain εε; gray and black: without and with averaging using SmooMuDS with Δε=0.01Δε=0.01; dotted gray lines mark minimum and inflection points of ε ? ; transient n′-decrease in 0.2<ε<0.40.2<ε<0.4 due to work softening at decreasing rate; relative n '-minimum at ε≈0.35ε≈0.35 due to increasing influence of fracture.

Fig. 2. Examples of grain structures of 4Al at ambient temperature after ECAP in (a) SEM, thick and thin lines for θ > 15° and > 2°, respectively, (b) TEM.

Fig. 3. Examples of grain structures of 8Al at ambient temperature after ECAP in (a) SEM, thick and thin lines for θ > 15° and > 2°, respectively, (b) TEM.

Fig. 4. Boundary spacings d (θ > 15°), w2 (θ > 2°) and w (all boundaries) in pAl at ambient temperature as function of p collected from present and previous [12,13,33,34] work in Brno (black symbols); gray symbols from Cabibbo et al.[3]; spread of d (black circles) for p = 8 and 4 demonstrates experimental uncertainty due to spatial inhomogeneity and observation method.

Fig. 5. Grain structures of 8Al after (a) 1 h, (b) 3 h of annealing at 473 K; the low-angle subgrain structure in several of the large grains (no thick black lines between patches of different color) indicates that discontinuous grain coarsening took place during deformation by ECAP.

Fig. 6. Grain structure of 8Al (all LAGB and HAGB visible) after 6 h of annealing at 473 K (TEM).

Fig. 7. HAGB spacings d and subgrain sizes w (see Fig. 6) after annealing of 4Al and 8Al at 473 K for a time interval tann (with data from Ref. [12]); gray lines: from Eq. (4) for different starting values d0; w0: estimate of subgrain size at start of creep.

Fig. 8. Selected curves of ε ? as function ofε for CG Al in compression at constant stress.

Fig. 9.  ε ? versus σ/G for creep of CG Al in tension at constant load Fc = σ0 S0 with σ0 = 15 MPa; gray field: data of normalized ε ? (right ordinate scale) from compression and tension tests on Al99.99 [37,38]; bar in lower left shows tensile strain scale (see text).

Fig. 10. Tensile curves of CG Al at constant load with highest measured deformation resistance and compression data for final deformation resistance at σ and ε≈0.3ε≈0.3; gray field: see Fig. 9; open circles: compression results for Al99.99 [5,39].

Fig. 11. Average ε ? -ε curves in compression at 473 K and 15 MPa for CG Al and p Al; wiggles in average curves for 0.29<ε<0.350.29<ε<0.35 have no meaning, but result from variations of final strains of individual curves.

Fig. 12. Average ε ? -σ curves of 1Al in tension at constant load compared to typical curve of CG Al; dotted: qs deformation strength for CG Al from Fig. 10.

Fig. 13. Boundaries in gauge of 1Al after tension at 473 K and σ0 = 15 MPa to ε=0.41ε=0.41; red: misorientation ≥15°, green: misorientation above ≈ 2° (experimental limit); the mean HAGB spacing d determined from the limited fields of view are (a) 8.6 × 10-5 m and (b) 3.7 × 10-5 m; the largest, only partly visible grain on the left of (a) extends over > 3 × 10-4 m.

Fig. 14. Compression of 8Al: (a) evolution of deformation resistance with strain εε at constant stress σ, (b) boundaries after compression at σ = 11 MPa for 205 h to ε=0.35ε=0.35 (dashed curve in (a)); red and green lines: HAGB and LAGB, respectively; compression direction: horiziontal.

Fig. 15. Tension of 8Al: (a) gray: typical ε ? -σ(ε) curves at constant load; black: all curves for σ0 = 15 MPa: white-filled black circles exemplarily mark begin of significant fracture influence (see section 3); dashed: specimen of Fig. 15; (b, c) boundaries in (b) head, (c) gauge after 60.3 h of creep at initial stress σ0 = 15 MPa till fracture at ε=0.72ε=0.72 (dashed line (a); red: HAGB, green: LAGB, tensile direction: vertical.

Fig. 16. Spacings d (HAGB, θ > 15°) and w2 (θ > 2°) after creep of 4Al and 8Al as function of normalized stress in relation to estimated subgrain size w0 (TEM) at start of creep (Fig. 7) and qs subgrain size wqs (Eq. (1)).

Fig. 17. Minimal (ε≈0.1ε≈0.1) and final (0.3<ε<0.60.3<ε<0.6) creep rates as function of stress for (a) p = 1, (b) p = 2, (c) p = 4, (d) p ≥ 8; label ‘?’ indicates probable influence of fracture.

Fig. 18. Schematic of variation of (average) HAGB-fractions fHAGB,ECAP after ECAP and fHAGB,0 at creep begin after heating to T and fHAGB,fin at end of creep tests at 4 × 10-4 < σ/G < 1.5 × 10-3; for comparison: wqs/d0 in investigated stress range.

Fig. 19. Overview of fit lines from Fig. 17 for (a) maximal and (b) final deformation strength of pAl; the final (approximately qs) strength for p ≥ 8 was fitted (gray dashed line in lower left of (b)) with the model of qs deformation under HAGB-control of recovery [46], using stress-dependent d = wqs(σ), wqs = 1 × 10-6 m at σ = 0.0114 G, fit values c = 0.2, and dislocation interaction factor α = 0.3; see text.