Xian Chen, chen1561@umn.edu Co-authors: Yintao Song, Richard D. James From compatibility conditions to stress-free microstructure for martensitic transformation The cofactor conditions (CC) are the conditions of super compatibility between phases for martensitic transformation. By satisfying CC, austenite and variants of martensite can fit together without elastic transition layers for any twinning volume fraction between 0 and 1. Here we discuss different forms of CC in Type I/II, Compound twins and domains, followed by the prediction of their possible microstructures. Then we calculate the geometric linear case of CC. Finally, we show real examples whose lattice parameters were tuned to satisfy CC closely for both Type I and II twin system. Materials undergoing reversible phase transformations are desirable for applications in medical devices, microelectronics and energy conversion devices [1]. The conditions of elastic compatibility of the phase-transforming materials have profound impact on these applications. Because they dominate the hysteresis, the reversibility and the resistance to functional degradation upon cyclic operation [2, 3]. It has been proven experimentally that the degree of hysteresis and the cycling ability of the materials can be optimized [4] by making the lattice parameters satisfy a set of geometric compatibility conditions λ2 = 1 a · Ucof(U2 − I)n = 0 trU2 − det U2 − |a||n| 4 ≥2 (1) (2) (3) where λ2 is the middle eigenvalue of the transformation stretch tensor U, the vector a, n are the twinning parameters. Microscopically, the satisfaction of these conditions means a perfect lattice points match at the interface between the initial and final phase, which results in stress-free microstructure in the material. Therefore, it allows for materials transforming back and forth freely without paying internal elastic penalty. This is why the materials satisfying CC exhibit tremendous reversibility. References [1] Srivastava, V., Song, Y. Bhatti, K. and James, R. D. Adv. Eng. Mater. 1, 97-104 (2011) [2] Ball, J. M and James, R. D. Arch. Ration. Mech. Anal. 100, 12-52 (1987) [3] Chen, X., Srivastava, V., Dabade, V. and James, R. D. J. Mech. Phys. Solids, 61, 2566-2587 (2013) [4] Song, Y., Chen, X., Dabade, V., Shield, T.W. and James, R. D. Nature, 502, 85 (2013) [5] Chen, X., Song, Y. and James, R. D. in prep. From compatibility to stress-free microstructure for martensitic materials Xian Chen! ! Aerospace Engineering and Mechanics University of Minnesota Support by MURI project (AFOSR) and NSF-PIRE Martensitic transformation Applications Aerospace engineering Microelectronics Medical device Flexinol®! actuator! wire geometry adjustable Chevron stent Durability Hysteresis Reversibility Materials undergoing phase transformation Irreversible Tin pest Reversible A material made by! satisfying some mathematic conditions Structural transformation em 3 initial structure! (austenite) final structure! (martensite) F for some em 2 ea3# ea2# = I ea1 • U stretch matrix depends on lattice parameters of initial and final lattices. • For martensitic transformation em 1 c U Ø a b Û Û determines the variants of martensite b c a i.e. cubic to monoclinic, there are 12 variants. Ø Ø c b a Structural transformation em 3 initial structure! (austenite) final structure! (martensite) F = RU ! for some R∈SO(3) em 2 ea3# ea2# = I ea1 • U is the transformation stretch matrix that only depends on lattice parameters of initial and final lattices. • For martensitic transformation em 1 c U Ø a b Û Û determines the variants of martensite b c a i.e. cubic to monoclinic, there are 12 variants. Ø Ø c b a Structural transformation em 3 initial structure! (austenite) final structure! (martensite) F = RU ! for some R∈SO(3) em 2 ea3# ea2# = I ea1 • U is the transformation stretch matrix that only depends on lattice parameters of initial and final lattices. • For martensitic transformation em 1 c U Ø a Ûi = Qi U b Ûj = Qj U QjT QiT b the symmetry-related Û = Q U QT, for Q ∈ 𝒫a determines the variants of martensite c a i.e. cubic to monoclinic, there are 12 variants. Ø Ø c b a In the case of Tin pest Initial phase: tetragonal! a0 = 5.83! c0 = 3.18 . = Final phase: cubic! a = 6.48 . . . = = . . I . . U1 or U2 I . U3 condition for a compatible interface Deformation gradients are differed by a rank1 matrix A-B=b⊗m A B for some b, m ∈ ℝ3 Conditions of compatibility in general case twinned martensite Let U and Û be symmetry related variants austenite and they satisfy QÛ - U = a ⊗ n ! for some Q ∈ SO(3) f n 1-f n a U QÛ U Favg = Rf [f U + (1 - f )QÛ] 200μm mf ( + ) = has solutions for some f ∈ [0, 1] iff the C(f) = (U + f n ⊗ a) (U + f a ⊗ n) has middle eigenvalue 1. J. Ball and R. D. James, Rat. Mech. Arch. 1987" RV. Kohn and S. Muller, Phil. Mag., 1992 If C(f) has middle eigenvalue 1 for some f twinned martensite austenite type I twin with twinning plane n = (0, 1, -1) f = 0.208 f = 0.792 n QÛ - U = a ⊗ n ( + ) m = (0.52,0.01,-0.85) m = (0.52,-0.85,0.01) m = (-0.52,0.01,-0.85) m = (-0.52,-0.85, 0.01) = There are 4 solutions per twin system (U, a, n) {, {, { , { , + + , , + , , + + } } + } } J. Ball and R. D. James, Rat. Mech. Arch. 1987 A degeneracy case: if λ2 = 1 where λ2 is the middle eigenvalue of U a perfect interface for f = 0 " ( + ) = R0U - I = b0 ⊗ m0 or a perfect interface for f = 1" R1U - I = b1 ⊗ m1 Hysteresis and λ2=1 80 TiNiPt TiNiCu ZnAuCu 72 TiNiPd TiNiAu mf Hystersis (degree Celcius) 64 56 48 40 # of strains reduces 32 24 austenite 16 single variant! martensite 8 0 0.950.960.970.980.99 1 1.011.021.031.041.051.061.071.08 middle principle stretch of transformatio matrix J. Cui et al., Nature Mater., 2006" Z. Zhang et al., Acta Mater., 2009" X. Chen et al., JMPS, 2013" Y. Song et al., Nature, 2013 Further degeneracy for (U, a, n) " Cofactor Conditions (CC) Let U = UT and detU > 0, and define Û = (-I + 2 e ⊗ e) U (-I + 2 e ⊗ e) for some |e| = 1, so that Q Û - U = a ⊗ n for some Q ∈ SO(3) and a, n ∈ ℝ3, then the equations of crystallography theory of martensite ( + ) = has solutions {Rf , bf ⌦ mf } for every f ∈ [0, 1] if and only if ▪ (CC)1 λ2 =1 Click to play ▪ (CC)2 ▪ (CC)3 · ( ) = | | | | There are infinitely many strains ! cofactor conditions in different twin systems Let U and Û = (-I + 2 e ⊗ e) U (-I + 2 e ⊗ e) define a twin system. physically, e, |e|=1 is a 2-fold symmetry axis of the austenite. (CC) (CC) (CC) X Type I satisfied simultaneously Type II Compound λ X e ⋅ v 2 = 0, e ⋅ v2 = 0 ⊥ (2 1− (e ⋅ v1)2 (e ⋅ v3 )2 +1) detU2 ≥ 2 where XI = |U-­‐1e|, XII = |U e| denote the inverse and forward stretches along the 2-­‐fold axis e. X.Chen et. al., JMPS, 2013 microstructure by satisfying CC compound twin: plane strain Compound twin • λ2=1" • e⋅v2=e⊥⋅v2=0" • inequality condition for every f ∈ [0, 1]" • Rf (U + f a ⊗ n) - I = bf ⊗ mf " • mf ⋅ v2 = 0 A general case " of transition layer v2 mf e e⊥ type I twin: stress-free triple-junction • λ 2 = XI = 1 Rf (U + f a ⊗ n) - I = bf ⊗ mf! Rf is the same for every f ∈ [0, 1] for half of the solutions All rank-1 connected triple-junctions" RU - I = b0 ⊗ m0" RQÛ - I = b1 ⊗ m1" RQÛ - RU = Ra ⊗ n n I RU Perfect interface transition layer m0 m1 RQÛ X. Chen et.al, JMPS, 2013 Diamond nucleation of austenite The austenite (red lattice) in a diamond shape can grow from a slim band of M variant (blue lattice) in another M variant (green lattice) by satisfying the Cofactor Condition for Type I twin This nucleation process occurs at zero elastic energy. Au ste nit e M2 M1 X. Chen et.al, JMPS, 2013 type II twin: parallel phase boundaries • λ2 = XII = 1" ➞ Twinning plane || habit plane for every f ∈ [0, 1] mf nII X. Chen et.al, JMPS, 2013 CC satisfied for both Type I and Type II twins simultaneously • λ2 = XI = XII = 1 Six rank-1 connections = Triple junction Quad junction = =ˆ )=ˆ ( )=ˆ ( ( a te s u M1 e nit M3 M1 M2 M2 )=ˆ …result in six perfect interfaces! M4 X. Chen et.al, JMPS, 2013" Y. Song et.al, Nature, 2013 Combination of tri-quad junctions with varying the volume fractions continuously au s ite n te M1 M2 austenite marte M1 M2 nsite M3 M4 X. Chen et.al, JMPS, 2013" Y. Song et.al, Nature, 2013 the CC material has infinitely many flexibility to form zero energy microstructure " these are not just math… A family of alloys satisfies CC closely Differential Scanning Calorimetry " for an alloy family Zn45AuxCu55-x 3.1℃ Trends of hys. 2.4℃ 6 Hysteresis = 0.5℃ 0.8℃ 4.2℃ 4.8 5.1℃ Hysteresis ℃ 2.6℃ 3.6 2.4 1.2 0 24 26 28 30 Au at% 32 Structural determination by XRD Austenite 17℃ Martensite -50℃ L21 M18R" (P2) Y. Song et.al, Nature, 2013 Lattice parameters and Cofactor Conditions (CC)2 for Type I twin (CC)2 for Type II twin Y. Song et.al, Nature, 2013 1 Cycling the phase transformation ~16,000 Shift of trans. temps. °C 0 -2 -4 -6 Au30 -8 Au27 Au25 As Af Ms Mf 64 12 8 25 6 51 2 16 32 8 4 2 1 -10 Number of cycles Y. Song et.al, Nature, 2013 210 211 212 213 214 Irreproducible microstructure from cycle to cycle Zn45Au30Cu25 λ2 = 1.0006" XI = 1.0081" XII = 0.9996 Click to watch the movie of reversible phase transformation Y. Song et.al, Nature, 2013 Ongoing work Theoretical prediction Optical microscopy Ongoing work Theoretical prediction Quantitative verification of the theory Optical microscopy SEM EBSD Ongoing work Theoretical prediction Quantitative verification of the theory Optical microscopy SEM EBSD Summary • formulation of the cofactor conditions (CC) for martensitic phase transformation • geometric interoperation of CC in type I, type II and compound twin systems • possible stress-free microstructure by satisfying CC in different twin systems • discovery of materials that satisfy CC closely, which show enhanced reversibility and unusual microstructure Thank you This work is done by University of Minnesota group lead by Prof. Richard James and ongoing work is collaborating with CMU group lead by Prof. Marc de Graef and LBNL advanced light source. " Supported by ASFOR MURI project: Managing the Mosaic of Microstructure and NSF-PIRE.