Electronic Supplementary Materials Analytical Solutions of the Displacement and Stress Fields of the Generic Nanocomposite Structure of Biological Materials Gang Liu, Baohua Ji, Keh-Chih Hwang, Boo Cheong Khoo Appendix A: The unknown functions of the initial solutions 0 0 0 0 0 0 0 The unknown functions f i , gi , i , i , , p , q , and s 0 in the initial solutions (Eq.(22) and Eq.(25)) can be determined by applying appropriate m 0 p 0 m 0 p 0 boundary conditions. Firstly, based on the continuity conditions ( vI v , u I u , xym,0I xyp 0 , yym,0I yyp 0 ) at the interface between mineral I and the protein layer at d p 2d m , we can obtain the following relations, d m p f I 0 g I 0 2d m l 1 p2 d p 0 l 1 p d p 0 d p 0 0 s p p q EmCE 2d m 2 EmCE 2d m 2d m 2 (A1) l d p 0 0 p GmCG 2d m (A2) Em d p 0 0 0 f I I l 2d m (A3) f I 0 2 2 d p 0 d p 0 Em d p 0 0 0 f I I I s 2l 2d m 2d m 2d m (A4) m 0 p 0 m 0 p 0 m 0 p 0 m 0 p 0 Based on the continuity conditions ( vII v , u II u , xy, II xy , yy, II yy ) at the interface between mineral II and the protein layer at d p 1 2d m , we have, d p 0 0 f II g II d 2 m m l 1 p2 d p 0 l 1 p d p 0 d p 0 0 s p p q EmCE 2d m 2 EmCE 2dm 2d m 2 f II 0 2 l d p 0 0 p Gm CG 2d m (A5) (A6) Em d p 0 0 0 f II II l 2d m (A7) 2 d p 0 d p 0 Em d p 0 0 0 f II II II s . 2l 2d m 2d m 2d m (A8) m 0 0 Then, according to the symmetric boundary conditions on AB (see Fig.2), we set vI C and xym,0I 0 at d m d p 2d m . That is, d m d p 0 0 0 m f I g I C 2d m (A9) Em d m d p 0 0 f I I 0. l 2d m (A10) m 0 0 m 0 For the symmetric boundary conditions on GH (Fig. 2), we set vII C and xy, II 0 at d m d p 2d m , hence d m d p 0 0 0 f II g II C 2d m m (A11) Em d m d p 0 0 f II II 0. l 2d m (A12) Secondly, we solve the unknown functions with above boundary conditions. Combining Eq. (A7) and Eq. (A3), we obtain Em d p 0 Em d p 0 0 0 f I f II I II 0. l 2d m l 2d m 2 (A13) Similarly, subtracting Eq. (A12) from Eq. (A10), we have Em d m d p 0 Em d m d p 0 0 0 f I f II I II 0. l 2d m l 2d m (A14) By comparing Eq. (A13) with Eq. (A14), we get 0 0 f I f II . (A15) And substituting Eq. (A15) into Eq. (A10) and Eq. (A12) arrives I 0 II 0 Em d m d p 0 f I . l 2d m (A16) According to Eq. (A16) and Eq. (A3), we obtain 0 Em 0 f I . 2l (A17) Next, summation of Eq. (A2) with Eq. (A6) gives f I 0 f II 0 2 p 0 . (A18) Combining the second order derivatives of Eq. (A18) and Eq.(A15), we have p 0 0. (A19) which yields, p 0 C1 0 C2 0 . (A20) Substituting Eq. (A17) and Eq. (A20) into Eq. (A2), we have 1 2 f I 0 f I 0 C1 0 C2 0 . (A21) of which the solution is f I 0 C1 0 C2 0 C3 0 e C4 0 e where l 2 E p E d d 1 . m m p p (A22) Substituting Eq. (A22) into Eq. (A16) and Eq. (A17), respectively, we obtain, 3 I 0 II 0 2GmCG d m d p ld p C e 0 3 C4 0 e (A23) and 2GmCG d m 0 C3 e C4 0 e . ld p 0 (A24) Substituting Eq. (A20) and Eq. (A24) into Eq. (A6) yields f II 0 C1 0 C2 0 C3 0 e C4 0 e . (A25) Substituting Eq. (A22) into Eq. (A9), we have g I 0 d m d p 0 0 f I C 2d m d m d p 0 dm d p 0 m m C3 e C4 e 2d m 2d m m 0 0 C2 C (A26) and substituting Eq. (A25) into Eq. (A11), we obtain d m d p 0 0 0 g II m f II C 2d m d m d p 0 d m d p 0 0 0 m m C3 e C4 e C2 C . 2d m 2d m (A27) Then, subtracting Eq. (A5) from Eq. (A1), we obtain s 0 2 EmCE d m 0 EmCE m d m p d p 0 C C2 . l 1 p2 d p l 1 p2 d p (A28) and adding Eq. (A1) to Eq. (A5), we get q 0 m 2 1 d p 8d m p C3 0 e C4 0 e . Based on Eqs. (A4), (A17), (A22), (A23), (A28), we obtain 4 (A29) 2 d d E d I p 0 s 0 m p f I 0 p I 0 2l 2dm 2dm 2dm 0 GmCG d p 2ldm C3 0 e C4 0e 2EmCE dm 0 EmCE m dm p d p 0 C C2 . ld p 1 p2 ld p 1 p2 (A30) Based on Eqs. (A8), (A17), (A23), (A25), and (A28), we obtain 2 d d E d II p 0 s 0 m p f II 0 p II 0 2l 2dm 2dm 2dm 0 GmCG d p 2ldm 0 0 C3 e C4 e 2EmCE dm 0 EmCE mdm p d p 0 C C2 . ld p 1 p2 ld p 1 p2 (A31) Finally, according to the loading conditions (for instance, displacement loading u0 (0) 2 1 on boundary BD) and considering the equilibrium of the nanostructure and the anti-symmetry of displacement in mineral I respect to mineral II, we have uIm 0 uIIm 0 (A32) uIm 0 (A33) dm d p dp u0 0 2 1 4 2 dm 2 dm 14 1 4 xxm,0I d 0 where yym,0I d 0 where dm d p 2d m 1 4 (A34) . (A35) m 0 Substituting the expression for ui into Eq. (A32) and Eq. (A33), and also substituting the expressions for xx,i and yy,i into Eq. (A34) and Eq. (A35), respectively, we can obtain m 0 C 0 m 0 m d m p d p sh 4 u 0 d m sh 4ch 4 4 (A36) 0 C1 0 0 (A37) 5 C2 2 sh 0 4 sh 4ch 4 C3 C4 0 0 u0 0 (A38) 4 1 sh 4ch 4 u0 , 0 (A39) 4 Then we obtain the expressions of all the unknown function by substituting Eqs. (A36) to (A39) into Eqs. (A22), (A25), (A26), (A27), (A23), (A30), (A31), (A20), (A29), (A24) and (A28) fi gi 0 A 2 sh 0 i p q 0 0 2ch 4 (A40) dm d p m d p p d p m sh sh 4 dm dm A i 0 A 0 Em d m d p ld m A Em d p2 4ld 2 m 2 A sh A m 2 2 ch (A42) 3 sh (A43) (A41) (A44) 4 1 d p 8d m E 0 A m 2 ch l p 2 sh (A45) (A46) s 0 0. (A47) 6 Appendix B: The first order perturbation correction In this section, we solve the first order perturbation correction of the perturbation solutions of the governing equations. According to the perturbation expansions obtained in section 3.2, the governing equations of the first order perturbation (the second expansion term) for the mineral crystal are, m1 xx m1 yy m1 xy xym 0 1 1 u m1 m 1 m 0 xx m yy Em l 1 1 v m1 yym 0 m xxm1 Em l (B1) Gm u m1 v m 0 l (B2) 1 xym1 2Gm xym1 m1 xx 0 m1 m1 yy xy 0. (B3) and those for the protein are, p1 1 1 u p1 p 1 p 1 xx xx p yy EmCE l p 1 p1 1 p1 p1 1 v yy yy p xx EmCE l p 1 1 1 u v p 0 p 1 xyp1 xy 2GmCG 2l (B4) xyp1 p 0 xx 0 p1 yyp1 xy 0. (B5) Solving above governing equations we can obtain the displacement and stress fields of the mineral 7 and protein of the first order perturbation as below m1 1 ui f i Li , m1 1 1 vi m f i gi M i , Em 1 m1 f i N i , xx ,i l Em m1 1 1 xy ,i l f i i Oi , m1 Em 2 f 1 1 1 P , i i i i yy ,i 2l (B6) for mineral, where i I and II, standing for mineral I and mineral II, respectively, and l p 1 1 1 u G C p Q , m G 2 l 1 p2 1 l 1 p 2 1 1 p v s p p 1 q 1 R , Em C E 2 Em C E Em CE 1 p 1 1 1 p p s S , xx 2 p l xyp 1 1 T , p 1 1 s 1 U , yy (B7) for protein. The unknown functions Li , , M i , , N i , , Oi , , Pi , , Q , , 1 1 1 1 1 R , , S , , T , , U , ; and f i , gi , i , i , , p 1 , s 1 , q 1 can be determined by applying the boundary conditions. The unknown functions in Eq.(B6) are Li , 2 m 2 0 l fi gi 0 i 0 Gm 2 (B8) l 1 m 0 l 1 m 2 0 1 2m 3 0 M i , fi m 2 gi 0 i i 6 2 2Em Em 2 N i , 2 Em 2 0 E fi m gi 0 2 m i 0 m i 0 l l 8 (B9) (B10) Em 3 0 E 2 m 2 0 fi i m i 0 m 2 gi 0 3l 2l 2 E E 2 m 3 0 0 0 i m 2 i 0 Pi , m 4 f i m 3 gi 12l 6l 6 2 Oi , (B11) (B12) where i I and II denoting the mineral I and mineral II, respectively. And the unknown functions in Eq.(B7) are Q , R , l 1 p 3 p 6EmCE l 1 p 12EmCE S , 4 0 p 6 2 p 2 2 p 6 0 2 3 2 T , U , 3 3 0 2 p 2 2 0 p 1 2 p l p 1 p 2 s 0 q 0 p 1 2 p 2s0 EmCE q0 EmCE 2 0 p l l 2 (B15) 3 0 3EmCE 3s 0 (B13) (B14) 6 2EmCE 2 2q 0 p l 1 p 2 0 EmCE p 0 p s 0 l 3 0 EmCE 2 0 p p 2 s 0 . 2l 2 2 (B16) (B17) Substituting Eqs. (A40) to (A47) into above equations, we have Li , 2 m 2 0 l fi gi 0 i 0 Gm 2 Em dm d p 2 d d 2 m 2 l A 2 2ch A m p 2mch A ch Gm ldm 2 dm 2 m d m d p 2 2 m 2 A 2 2ch A ch dm 2 d dp 2 A 2 m 2 m ch dm 9 (B18) 2 l 1m2 0 l 1m 2 0 1 2m 3 0 m 2 0 Mi , fi gi i i Em 6 2 2Em d d 1 2m 3 A 2 3sh m 2 A m p 3msh 6 2 dm l 1m2 Emd p2 3 l 1m 2 Em dm d p 3 A sh A 2 sh 2Em 4ldm ldm Em 2 (B19) 2 2 1 2 dm d p 1 2m 2 1m d p 3 m 3 sh A 4dm2 3 2 dm Em 2 0 E fi m gi 0 2 m i 0 m i 0 l l d d E E m 2 A 2 3sh m A m p 3m sh l l dm Ni , Em dm d p Emd p2 3 2 m A sh m A 2 sh 4ldm ldm 2E E d2 2E dm d p A m 2 m m m2 p 3sh l 4ldm l dm (B20) 3 Em 3 0 E 2 m 2 0 fi m 2 gi 0 i mi 0 3l 2l 2 d dp 4 E E m 3 A 2 4ch m 2 A m mch 3l 2l dm Oi , Emd p2 4 2 m 2 Em dm d p 4 A ch m A 2 ch 2 4ldm ldm 2 2E E dm d p 2 Emm d p 4 A m 3 m ch 3l 4ldm2 l dm 10 (B21) Em 4 0 E 2 fi m 3 gi0 m 3i0 m 2 i0 12l 6l 6 2 dm d p 5 E E m 4 A 2 5sh m 3 A m sh 12l 6l dm Pi , E d2 2 m 3 Em dm d p 5 A sh m 2 A m 2p 5sh ldm 6 2 4ldm (B22) 2 E Em dm d p 3 Emmd p 2 5 m 4 A sh 6l dm 3l 8ldm2 Q, l 1p 3 p 6EmCE l 1p 3 p 6EmCE 2 p l 1 s 2 3 0 3 A p 2 2 0 p 2EmCE 2 0 q0 1p dp 2ch Em 4 ch Am l 4dm (B23) 1p 3 p 1p dp 2ch 3 2 m A 6CE 4dm R , l 1 p 2 4 0 12EmCE l 1 p 12EmCE 2 4 A 1 2 p p 6 3 0 l p 1 p 3EmCE 3s 0 p 2 2q 0 1p d p 3sh Em 5 sh p 2 Am l 2 4dm (B24) 1 2 1p d p 2 3sh p A 4 2 p m 12CE 2 4dm 1 2 p 2 s0 EmCE q0 EmCE 2 0 p 6 2 l l 3 2 p 3 A Em 5sh EmCE A 1 p d p 3sh m 6 4d m l l S , 3 2 p 3 0 3 2 p Em 1 p d p 3sh E C A - 3 2- m E m 6l 4dm l 11 (B25) T , 2 p 2 2 p 2 0 Em C E p 0 p s 0 l Em 4 ch l 2 2 p Em 2 4ch A 2l U , 2 p 6 2 p 2A 3 0 (B26) Em CE 2 0 p p 2 s 0 2l 2 Em 5 sh l 6 2 p Em 3 5 sh . A 6l 3A (B27) It can be found that LI , LII , (B28) M I , M II , (B29) N I , N II , (B30) OI , OII , (B31) PI , PII , (B32) Q , Q , (B33) R , R , (B34) S , S , (B35) T , T , (B36) U , U , . (B37) In the following, we will determine the coefficients in above unknown functions. Firstly, according to m1 p1 m1 p1 m1 p1 m1 p1 the continuity conditions ( u I u , vI v , xy, I xy , yy, I yy ) on the interface of the mineral I and the protein at d p 2d m , we have 12 d p 1 dp dp 1 m f I g I M I , R , 2d m 2d m 2d m (B38) l 1 p2 d p 1 l 1 p d p 1 d p 1 1 s p p q 2 EmCE 2d m EmCE 2d m 2d m 2 2 dp l dp 1 f I LI , 2d m GmCG 2d m 1 dp 1 p Q , 2d m (B39) dp dp Em d p 1 1 1 f I I OI , T , l 2d m 2d m 2d m (B40) 2 d p 1 dp Em d p 1 1 f I I I PI , 2l 2d m 2d m 2d m (B41) d d p 1 s 1 U , p . 2d m 2d m m 1 And according to the continuity condition ( u II u p 1 the interface of the mineral II and the protein at d p m1 , vII v p 1 , xy, II xy , yy, II yy ) on m1 p1 m1 p1 2d m , we obtain dp dp d p 1 1 f II g II M II , R , 2d m 2dm 2d m m l 1 p2 d p 1 l 1 p d p 1 d p 1 1 s p p q 2EmCE 2dm 2 EmCE 2dm d m 2 2 dp l 1 f II LII , 2d m GmCG dp d p 1 1 p Q , 2d m 2d m dp dp Em d p 1 1 1 f II II OII , T , 2d m 2d m l 2d m (B42) (B43) (B44) 2 dp d p 1 Em d p 1 1 f II II II PII , 2l 2d m 2d m 2d m d d p 1 s 1 U , p . 2d m 2d m (B45) m1 1 m1 And from the symmetric boundary conditions on AB (see Fig. 2), we set vI C and xy, I 0 at 13 d m d p 2d m , then we have d d p 1 dm d p 1 m m f I g I M I , 2d m 2d m Em l 1 C (B46) d m d p 1 dm d p 1 f I I OI , 0. 2d m 2d m (B47) And for the symmetric boundary conditions on GH (Fig. 2), we set vII C and xy, II 0 , at m1 1 m1 d m d p 2d m , then we have dm d p d m d p 1 1 f II g II M II , 2d m 2d m m 1 C (B48) dm d p Em d m d p 1 1 f II II OII , 0. 2d m l 2d m (B49) Secondly, we use above relations obtained from the boundary conditions and continuity conditions to solve the unknown functions. Subtracting Eq. (B40) from Eq. (B44), and applying OI , OII , , T , T , , we obtain Em l d p 1 Em d p 1 1 1 f I f II I II 0. l 2d m 2d m (B50) And subtracting Eq. (B47) from Eq. (B49) and applying OI , OII , , we obtain Em d m d p 1 Em d m d p 1 1 1 f I f II I II 0. l 2d m l 2d m (B51) Combining Eq. (B50) and (B51) leads to f I f II 0. 1 1 (B52) Then substituting Eq. (B52) into Eq. (B47) and Eq. (B49), and considering OI , OII , , we obtain I1 II1 dm d p Em d m d p 1 f I OI , 2d m l 2d m 14 . (B53) And substituting Eq. (B53) into Eq. (B40), we obtain 1 d d dp Em 1 f I OI , p OI , m 2l 2d m 2d m dp T , . 2d m (B54) Next, by adding Eq. (B39) to Eq. (B43), and considering the relationship LI , LII , , we obtain f I1 f II1 2 p 1 . (B55) According to Eq. (B52) and (B55), we have 2 p 0. 1 (B56) of which the general solution is p 1 C11 C21 . (B57) Substituting Eq. (B54) and Eq. (B57) into Eq. (B39), we obtain d d f I1 f I1 C11 C21 LI , p Q , p 2d m 2d m 1 2 dm d p dp 2l d p O O T , , , . I I Em 2 2d m 2d m 2d m (B58) of which the general solution is f I C1 C2 C3 e C4 e 1 1 1 1 AB 2 ch2 e ch sh ch2 e ch sh 3 48dm 2 0 where A u0 1 (B59) 2 2 2 3 sh 4ch , B 12m d p d m 3d m d p p d m d p 4d m . 4 4 From Eq. (B55) and Eq. (B57), we obtain f II1 C11 C21 C31 e C41 e AB 2 ch2 e ch sh ch 2 e ch sh . 3 48d m 2 15 (B60) From Eqs. (B46), (B48), (B53), (B54), (B57), (B41) and (B45), we obtain d d p 1 dm d p g I1 m m f I M I , 2d m 2d m 1 C (B61) dm d p d m d p 1 1 1 g II m f II M II , C 2d m 2d m (B62) dm d p Em d m d p 1 f I OI , l 2d m 2d m (B63) I1 II1 1 d d dp Em 1 f I OI , p OI , m 2l 2d m 2d m dp T , 2d m p 1 C11 C21 (B64) (B65) 2 d d E d I m p f I1 p 1 p I1 2l 2d m 2d m 2d m 1 d d s 1 U , p PI , p 2d m 2d m (B66) 2 d d E d II m p f II1 p 1 p II1 2l 2d m 2d m 2d m 1 d d s U , p PII , p . 2d m 2d m (B67) 1 Then subtracting Eq. (B38) from Eq. (B42), and considering the relationship M I , M II , and R , R , , we obtain s 1 EmCE d m l 1 p2 d p m d m p d p 1 1 C2 2C . dm (B68) By adding Eq. (B38) to Eq. (B42), and considering the relationship M I , M II , and R , R , , we obtain 16 q 1 f 4 m 1 I f II1 M I , dp dm d p M I , 2d m 2d m (B69) d l 1 p d p 1 R , p . 2 2 2 d E C d m m E m 2 2 Finally, the unknown coefficients C , C1 , C2 , C3 and C4 in above equations can be 1 1 1 1 1 obtained according to the loading conditions (For instance, displacement loading u0 2 0 on 1 boundary BD) and considering the equilibrium of the nanostructure and the anti-symmetry of displacement in mineral I respect to mineral II, which are uIm1 , uIIm1 , (B70) 1 d dm 1 uIm1 , p u0 2 4 2d m dm d p xxm,1I d 0 where 1 4 (B72) yym,1I d 0 where d m d p 2d m . (B73) 2 dm d p 2 dm 14 1 4 (B71) which can be further expressed as below, f I1 LI , f II1 LII , (B74) 1 d p d m u01 1 1 f I LI , 4 4 2d m 2 dm d p d p 2 dm 2 dm (B75) Em 1 1 1 l f I 4 N I 4 , d 0 (B76) E d p d m 2 1 d p d m 1 d p dm 1 1 4 2ml 2d m f I 2d m I I PI , 2d m 14 d 0. (B77) According to Eq. (B74) we have C1 0 , C3 C4 , therefore the original five unknown coefficients 1 1 1 can be reduced into three unknown coefficients C , C2 , C3 C4 . Then we have simplified 1 expression of Eq. (B59) as 17 1 1 1 f I1 C21 2C31 ch AB 2 3 1 ch sh 2 sh . 3 48dm 2 2 (B78) According to Eq. (B75), Eq. (B76) and Eq. (B78), together with Eqs. (B18) and (B20), we can obtain the unknown coefficients C2 and C3 as below, 1 1 2 B 2sh 2 ch 2 16dm dm 3d p sh 2 4 u1 C21 0 3 4ch sh 24dm 4ch sh 12 2 m dm dm2 2dm d p sh 4 4 4 4 2 2 sh A 3 (B79) 24 2m dmdp2sh B 2ch Bch 5B sh 12Bch u0 A 4 4 4 4 4 . C31 C41 2 3 4ch sh 192dm 4ch sh 4Bsh 32dm2 dm 3dp sh 96 2m dm dm dp ch 4 4 4 4 4 4 4 1 2 (B80) The unknown coefficients C can be obtained through Eq. (B77). By substituting Eq. (B78), Eq. (B63), 1 Eq. (B66) and Eq. (B22) into Eq. (B77), we can get the analytical solution of coefficient C as below 1 AB 2 d m 2d p m d m p d p 1 C2 1 p sh 3 96d m 2d m 4 2d m C 1 (B81) Finally, basing on the initial solution and the first order perturbation correction of the perturbation solution of governing equations, we can obtain the analytical expressions of the improved approximate perturbation solution of the displacement and stress fields in mineral and protein as below, m 0 uim ui m 1 2d ui vim d vim 0 2d vim1 (B82) (B83) xxm ,i xxm,i0 2d xxm,1i (B84) yym ,i 2d yym,0i 2d yym,1i xym ,i d m 0 xy ,i (B85) 2d xym ,1i (B86) u p u p 0 d2 u p1 v p d v p 0 2d v p1 (B87) (B88) 18 xxp 2d xxp 0 2d xxp1 yyp 2d xyp d p0 yy 2d yyp 1 p 0 xy 2d xyp 1 (B89) (B90) (B91) 19 1.0 (a) =40 Analytical =40 FEM =30 Analytical =30 FEM =20 Analytical =20 FEM u0 Em 0.6 l xxm ,I 0.8 0.4 0.2 0.0 -0.2 -0.1 0.0 0.1 0.2 -0.2 -0.1 0.0 0.1 0.2 -0.2 -0.1 0.0 0.1 0.2 -2 2.010 (b) 1.6 1.2 l xym , I u0 Em 0.8 0.4 0.0 0.50 (c) uIm u0 0.45 0.40 0.35 0.30 Figure S1 20 Figure S1: The comparison between our theoretical predictions and the FEM simulation results. (a) The dimensionless normal stress in the mineral crystal; (b) the dimensionless shear stress along the mineral-protein interface; (c) the dimensionless displacement in the mineral along the longitudinal direction. Em G p 2400 , and 50%. (Discussion: One reason for the discrepancy between our theoretical predictions and the FEM results at the mineral tips is that we assumed a traction free boundary condition at the mineral tips by neglecting the traction force acted by protein at the mineral tips during the model simplification from Fig. 1(b) to Fig. 2. The other reason is that we neglected the boundary condition at the end of protein layer. We note that there is non-zero shear stress at the end of protein layer, which violates the Newton’s third law for the mutual forces of action and reaction between the protein layer and its neighbors at lines CE and DF. This problem can be solved by constructing a boundary-layer type solution to satisfy the Newton’s third law. However, according to theSaint-Venant's principle, these simplifications will only affect the tip zone of the mineral of size about 2 d m d p . If the aspect ratio of the mineral is large, the ratio of this zone size to the length of mineral will be small. The consistency between the theoretical predictions and the FEM results just validated our assumptions) 21 b 45% 0.4 Eˆ Em 0.3 Our Analytical Solution TSC Model FEM Calculation 0.2 0.1 0.0 0 50 100 150 200 250 Aspect Ratio 300 350 Figure S2 Figure S2: The equivalent Young’s modulus of the nanocomposite structure versus the aspect ratio of mineral crystal predicted by our analytical solution, the TSC model and the FEM simulation ( Em G p 2400). The volume fraction of mineral 45% (as the case of bone). 22 1.0 (a) =30,Em/Gp=3000 =30,Em/Gp=2400 =30,Em/Gp=1800 =30,Em/Gp=1200 u0 Em 0.6 l xxm ,I 0.8 0.4 0.2 0.0 -0.2 0.025 -0.1 0.0 0.1 0.2 -0.1 0.0 0.1 0.2 -0.1 0.0 0.1 0.2 (b) u0 Em 0.015 l xym ,I 0.020 0.010 0.005 0.000 -0.2 0.50 (c) uIm u0 0.45 0.40 0.35 0.30 -0.2 Figure S3 23 Figure S3: The influence of the elastic modulus ratio of mineral to protein on the distribution of the dimensionless normal stress in mineral, the dimensionless shear stress at the mineral-protein interface and the dimensionless displacement in mineral. 50%. 7 =2.0 =4.0 =6.0 =8.0 =10.0 =20.0 6 l xxm , I u0 Em 5 4 3 2 1 0 -0.2 -0.1 0.0 0.1 0.2 Figure S4 Figure S4: The evolution of the distribution of the normal stress in the mineral along the longitudinal direction with the variation of the value. 24