9
Crack Growth and Pattern Formation Induced by Desiccation in Paste So Kitsunezaki * The method of using a spring network model with a breaking threshold is suc- cessful for the pattern formation of cracks in the drying process, in which paste is assumed as an elastic material. However recent researches reveal that the porous and rheological properties of paste influence the fracture process. After reviewing the formation of a crack pattern both in a thin layer of uniformly dried paste and in a directionally dried system, we describe the memory effect found by Nakahara and the drying-rate dependence of crack growth, which are distinctive features different from those in normal solids. Key Words: Mud crack pattern, Drying fracture, Paste rheology, Wet granular 1 *1) (mud crack pattern) Griffith * ( 630-8506 ) Tel 0742-20-3988. (Graduate School of Human Culture, Nara Women’s University, Nara 630-8506, Japan) ( ): 1992 , 1997 1997 : *1) colloid, slurry, wet granular, granular paste 1

Crack Growth and Pattern Formation Induced by Desiccation ...kitsune/paper/kitsune09b.pdf · Crack Growth and Pattern Formation Induced by Desiccation in Paste So Kitsunezaki The

  • Upload
    others

  • View
    8

  • Download
    0

Embed Size (px)

Citation preview

  • � � � ����� � �� � � � ��� � � ��� �

    Crack Growth and Pattern Formation

    Induced by Desiccation in Paste

    ��� �So Kitsunezaki∗

    ���The method of using a spring network model with a breaking threshold is suc-

    cessful for the pattern formation of cracks in the drying process, in which paste isassumed as an elastic material. However recent researches reveal that the porousand rheological properties of paste influence the fracture process. After reviewingthe formation of a crack pattern both in a thin layer of uniformly dried paste and ina directionally dried system, we describe the memory effect found by Nakahara andthe drying-rate dependence of crack growth, which are distinctive features differentfrom those in normal solids.

    Key Words: Mud crack pattern, Drying fracture, Paste rheology, Wet granular

    1 � �"!$#%'&)(+*-,/.1032547698;:@?BA* 1 ) C9D;EGFIH;JLKNMPO'QSRUT >

  • Fig. 1: Typical mud crack patternin a thin layer of calcium carbonate(CaCO3) paste. The thickness wasabout 4 mm after drying. The width ofthis photograph corresponds to 10 cm.

    i-1 i

    H

    a1

    2wi

    (u ,v )i i

    Fig. 2: 1-dimensional springnetwork model.

    H

    2L

    H

    L

    H

    L

    Fig. 3: Schematicpicture of the di-vision process of athin layer of paste.

    dl 4���� J�����5l� >�� @FnH���"FXe 0 V���� ß���� 0 SdJSs+MPz��@Ñ'i��1¯'i8N:h6L����50 ¯ ¬ 0�� x � \ f�! ¬ p#"%$ 2S4 '&)M)( )x�*)ª i,+1lXå Ù=æ=Å�- p. FIe�/10�23 � 032 0�4�5�6 ¤]@FdH9N�7�85C�9�:=@?S0GABA%C 0 ? /FdH ? T > A FIeID@E@@ 0 Ê1Ë)J V ONQ x2 MnRÉT >h, ê1j 0 D�E p)"F$9i�G�H 6 ! ¬JI�K�L J+« p#M@J&hMS;uC)µ f��N 1; ezPOÜ 0 Ê@ËSJ *+>9? A p�Q1R : 6 FHSU 6 =3T�UÌp#V5W È@º M9l È3X MY�Z�[ I�KPL C�\ È iF]_^ 0 T3USJ 0 ONQ'R5, p �P� ß 2 `MNhJÉK/M;SuC È[) ezh« 0a SJXÊNËSC�bdc x ? Hi;D@E@= x=V *S>/?½A0 O�e�f (hgji 23k > FH 0 R�fSClFmUF¼i�n�o 0PpP:50 = x c\ 6 l D�E C+s'M+S ` ¶ \ [ x 6) HÉKH@lSMnzNRÉT >h,I0=¨ q p#r½1s x FHi *S>9?ÜA¾0 ¯3R"=�7�8 x ? l@H 0)2 ý�fß 61¨�q C c ª H1l)M5l 4 Å JSs;MXz@SJ V 2 t J Y�Z�[ I�K�L pd¤)l'enD'ENN 0 RLT >5, ê1j 0 ¬1Ô Ê1ËÌpl³ ?'[p'u . F_i 3 t J V *+>9?ÜA0vg�i 2�k > x3w#x &hMh´ È MN@1 0 D3E pByF$5MIz

    2z|{~} #h~

    2.1 d,%�� ¡¢J£1¤F¥P¦¨§ª©¬«�®#¯3°d±²�³�´ &dµ�¶ ¯3°¬·¹¸»º¹¼%½�¾�¿�À3Á�Â�Ã�°v«'ÄdÅ1Æ�ÇdÈÉ_ÊÌËFig. 1 Í�Î%Ï ·¹ÐPÑ«�Ò�Ó�Ô,· ÕPÖ &Jµ'× 2 Ø3Ù�Ú ÁÛ�Ü�¥JÝ (mud crack pattern)È�Þ�ß µ àá�á Þ âã¢J£�ä ¶)å ¤F¥1¦æ§çÈ%½�¾�èPé·¹³�´�ê�ë &dµ%á�¶íìî º ˬï é Í ÆÇ%ÛÜ1¥�Ý å�ðdñ È,ò�ó�ô Í�õöÍ�÷øúù�û cm Í�ü�ý ©d±² ù µm Í�þ3ÿ��öÍ monolayer� Þ��d£3¦��F¥�Ó%Þ��� &�µ�áJ¶ È�Þöß µ 1) à Groisman & Kaplan Í�

    � å�� ¥��¥��¶�� Í ¤J¥P¦ § «��J£�ó���� Í���� Á

    � Þ�� µ È�¢�© Í ³�´�Â��·1Å%£3¼ (1) Æ3Ç å!

    2

  • 0 1 20

    10

    20

    s/sb

    ss/s

    b

    ~

    breaking

    start of slip

    sb

    ss

    qL

    fixed b.c.

    slip bondary condition

    Fig. 4: The shrinking rate at the first break-ing is plotted for the system size.3) Arrows in-dicate the change of the system size in a divi-sion process as in Fig. 3. • and ◦ are the break-ing points for the fixed bottom and a slipperybottom, respectively.

    Fig. 5: Schematic picture of a 2-d springnetwork model. The 2-d elastic layer is dis-cretized to a random lattice composed of tri-angular tiles, which are removed if their elas-tic energy exceeds a breaking threshold. Weintroduced the viscous relaxation equivalentto adopting Voigt elements (springs and dash-pots), as shown in the bottom of the figure.

    ØöÚ ·���� ø���� · T � ·���ò�¼�ÒPÓö«� µ Ë (2) ���Ú Á�Ò�Ó������ å © ÷ ·���� º¤J¥P¦æ§ © õ °dÞ Í���� ¶ Í���� È�� ø £%½! #"ß�£ Ë (3) © ÷ È�� ø%$ Á µ�¶ Y &��Í('JÈ�)d£�Û¬Ü1¥,Ý�·�*�+ &,µ Ë (4) Æ�Ç-, ¶ Í�.(/öÍ�0 ²BÅß å "ߣ3È2143 Í Æ�Ç å5(6�728 Í Æ�Ç Í�9�: «= Í�? £ .(/ Þ���@ &,µ Ë ¶ £ ÏA��B Á�C�D¬«FE�Gº¹¼�£ µ 2) àá Í\ Á4] Ø�Ù O^P�Q Ê_R�T Ó å Fig. 2 ÍdÎJÏ ·�¤�¥P¦ § ©L^ P ÓW�¥ å E = (1/2) ∑i [K1(ui+1 − ui + sa)2 + K2(ui − wi)2] ¶ Áµ È * 2 ) Ë h>�K

  • Fig. 6: Crack pattern obtained froma numerical simulation.3) Broken tilesare indicated with black color, and slipoccurs in the dotted region of each cell.

    Fig. 7: Columnar structure of cracksdeveloped in cornstarch paste. Thissample was dried from the top surfacein a cylindrical container, and then bro-ken by hand in the middle of drying.

    � µ������ Ú�B� å ¤�¥�¦æ§ © Í Ã�° Í ��Ô L̃ ≡ L/H Í�� Þ�K>L�*�� Í C(D Ú Á�@ ø åH/q

    ·PÁ µ�àFig. 3 Í,ÎdÏ Á�Æ�Ç· Î µ Ò�Ó Í Õ�Ö å Æ3Ç�È��>º ¼1Ð�Ñ«��Jà &öµ��� Þ � Ê Ø

    | Þ� µ ÎFÏ ·����JV Í�} È2)J£ à º#±º Æ�Ç�ÛöÜ�¥,Ý�� Í Y Í3Í �(�ö·�����¼�1 3 Í ÆÇ���@ö·�±�± µ I�d�È���dÞ�ß µ ½% ���£�����È�) $ Ë��! Á�Ò1Ó1Õ�Ö�( ¶ õ Þ Í�~ ÊÍ (¬« N = &dµFá%¶ Þ3ÒPÓ������ Í Id#"�$

  • 0

    2.5

    0

    2.50

    2.5

    Fig. 8: Typical crack pattern obtainedfrom a numerical simulation by Nishimotousing a spring network model with non-linear water transportation.11) A columnarstructure of cracks develops with dryingfrom the front surface.

    Fig. 9: Nakahara effect for a thin layerof CaCO3 paste. Before drying, thesample was shaken for about 2 min at140 rpm in the direction indicated bythe arrow. The size and the drying con-dition were approximately the same asthose in Fig. 1.

    w Í�������� U ≡ (1 + τ∂/∂t)u, W ≡ (1 + τ∂/∂t)w ���

    ������������������� �"!#�$�% �'&"(*)�+',�-/.�021436547�8"9�:�� W ����;� U )��?��"

    ��@'A�B20�CD 5FE�GIH�JIK�LM)NPO�� u ) U �RQ�SUTF�FVIWYXZ�[�\^]`_Iab�dce�)2X�Z f�.�gF�h2i [�j�k6�R�R�*54l��PmR[/nRo�p�q*r�B�s�t�uU�0�vw���Ixyz��{|f�}/OF~�IRU[nRo�z/('���U[F�);�'5P"f4."- 2 ���I! #� �w��`! #I jY)�=YX `Uf�B�b]SIOR�(R524n�o6\�&�P&�U��[^�� 4¡ # \/¢f�g��;\��*S'f4.� (3) ��|w(U���£I¤P¥§¦�¨ ��©)ª*«zf¬2"®'5|¯±°'S�AY]³²2��´��µB6©�¶z��·R¸�\�¹�ºP�4»I¼��/½��PmI�¾Y [M�À¿�Á2�·R¸2��Â"Ã2Ä # \FÅ�Æ�Uf h6iÇ [6T/-ÈIÉPÊ �;ËI,Y]³²Uf 4,5) .�Â"Ã�Ä # Ç/Ì O��*FÍ %�Î�Ï ���R�*5�O�S2��k�Ðzm�ÑIÒ'\2Óf4\�B4©R¶'��2·w¸4H�Ô*[��;�IÕ )FÖ�° Ç× �-Ñ�ÒU�Ø�ÙU��ÚfFÛ Ï 5�I Ê [�O/.g�²� 3 Ü'5FÝ i I (4) �`Þ Ç 6("xR��n�o�q�ßz��à�á Ç Õ�â'Uf"ã4ä Ê2å i .

    2.2 æç2è�éê|ë2ìPí�îzé�ï ð^ñMòPó^ô2(wb±õ±öP��s�t'\wÄ"�'5I[�O�dn�o�q�ßU\/÷�øõöz�wùUTz Ï �-�úIû�ü6ýFþP�4£¤ Ç�ÿ�� X�²Pf����6\*Ó2f�.Yg�� �w�n�oP�à��zm Ç Griffith 9I:'\Fq Ï� Ì|h�iÇ�

    �m Ç q�ß''f/\ * 4 ) B��z- Ç 6("õ/��ûU�������³n�o ����½�����6\�ã�ä �[6f4.�� m�[Rn*o2q*ß�)���������B*p�q�X�²|fFY�(|w�� �|mI5�!#"Ó'fU¯"�I\$zO�.Allain & Limat &%UO�'(R� ¹Y) ¥(' � 5�)�*�5�+�, �-�.�S^]0/21^X�Z f ¥`��6mR[&3�4�5�¬� )65'O�B*s�t Ç Þ|T��&7Rjz�4n�o6\�8�9&: *I5=�?@����O�f 6) .����U�¬I62AR�4u2$B5�C�²"B6nRo�D0E2\F/1|)�G�H��6�wú�û�üýwþ)�p�q��[�\Y]`q�ß''f�I�J # (self-driven crack) ¯LK�MMX²���O�f 1) .* 4 ) NPOPQ0RTSVUXWLY[Z]\6^T_]`0aLbV^dc]e]fPg6hPidj�kVlPY[mPn&oqpXrPst>u&vBs Griffith wLx ozy�{B|

    5

  • gF� 3 ����� Ç���� Pf��R\"B Müller Ç'h T&��=�?MX�²6-�©I¿IÁ �"'(R�`��ºPO÷Rø)���6S] sItYXZPf¬*25�Ó2f 7) .����OUg�� Ç�� ���"g å �©�¿�Á �/6(wÀ5��S�� Ì SUTF��´^]± A���k���¯���N Ö Ê \"B gF��34'5��MXJUQ Ç Fig. 7 � h�i [�� >P��nRoIzw(U\�1��6Uf.�g4�! '�"6,�-�#�$'\&%�²���5PRf � >�Ü�k!'�pz��(õ¬�)� ���P¯+*-,/X²wBIØ�ÙY¯ú/.�O^ Ì S���021�(�3F5FÑ�ÒMX²��IO2f 8–10) .'g;²^]��n�oU� � �Ä"�6[4s�tU�4u Ç �(R ����I~�4 Ç p�qX�²Pf;s�t65�76|\98�:��VH.�� Ç ÞT���p�qX²Pf2g/�;\�C*SUT���´ Ï B*j�;Um Ç ¯=�?U5�@#CI²���O�fX$�A!B|x��û2´ hC û�/�DP������LE�³Í %�ÎeÏ+F!G 1|)IHKJ�� C�Z f2gL�;5 Fig. 8 � h�i�Ç sIt5�7�|2�M� � >�nRoU��N�J)��66f�g���\�5zwf 9,11) .Ug4��ãä Ç2OKP \�Ó�f;JU9Q�SR ûUTVRXW/Y Ç6h f�N�Z|)I[�\�]M/-RO 12) .

    3 ^S_a` b�cedgfhcjilkm/n b§�/1 Ç ÞT���o��UmR[�>�?6SM]�pF¶UmR[q�r!s!t�:u�wv�xLy{z�|�}�~�� p&K��L�/z��{ n9 ��{� n �L�!-w� �p&¡�¢�£)¤¥V¦¨§�©6ª�« y¬/¡®�¯±° ¢ £{²�³´&³!µ z&|�¶�· �¸2¹ ° »º ²³¼!½�¾2¿ÀÁ ÂKÃ�° ¿�³!µ z�Ä!Å�Æ�Ç ( È/É ¸�¹ Æ!ÇV¡®Ê�Ë ) ª®Ì!Í 13) !ÎÏK�Ð�Ñ{¶!Ò�Ó ²³�µ zIÔ!ÕÖ�×&Ø�Ù!Ú!Û ª�Ü�Ý{Þ�ß/à á µ!³â{Ý z&|

    3.1 ã)ä6åçæ�èéLê�ëì ( í»î�ëì )ï�ð�ñ �/ò �ó�ôõ �ö/÷ ª�øKù�ú�ûKü zK¡= ï�ð Î ´ zIÖ×��ýþ ª®ÿ�� ´ z�|6¬ ¿ û{ü!Þ ö�÷L����{�/ò e Ä�Å ¾I¿�³��u´ z��»�� ª ³����Kú�

    1 �������± ³!¢ ° ï!ð ¾���³������ ´ z|� ü�! Fig. 1 ¡"$#� ß�à �·&% �' ýþ��(�) ó!ô ª*Lü z{¡ Fig. 9 º ¦ ¬& ó�ôK,+.- £ ý{þ��Ö×L¶�· ,/ Ø ¾2¿z * 5 ) |¸I¹ °®{¬& º ¦�£ Ä�ÅÆ�ÇL�� ¿ z�& ó�ô ö�÷ »º z10&243/÷�Lò e 65673�÷ ª�8Lü z�9�:»�/z/¬/¡ ª®½�@ �9»ò e 1A�L ï�ð Î}~�1B�ý �Ly{z»¬/¡ ª�« Þ |DC Þ E" L�R.S�T Kº z ¥�µ

  • flat

    paste

    displacement sensors

    balance

    springs

    Fig. 10: Measure-ment of stress.

    0.5 0.55 0.6volume fraction

    0.000200.00131

    0 500 1000time (min)

    0

    50

    100

    aver

    age

    stre

    ss (

    kPa)

    drying rateat the cracking time

    (1/min)

    (a) (b)

    Fig. 11: Stresses arising in alayer of paste for two drying con-ditions are plotted with respectto (a) time and (b) the volumefraction of water.16) Cracks firstappear near the peak of stress.

    10-4

    10-3

    drying rate(1/min)

    10

    100

    crac

    k sp

    eed

    (mm

    /min

    )

    10-4

    10-3

    drying rate(1/min)

    expo

    nent

    1.0

    0.587

    Fig. 12: Dependence of crack speed onthe drying rate. A thin layer of CaCO3paste was dried at 40 ± 1◦C for variousdrying conditions.16)

    £ 2�� (plumose structure) ª�ND ¬�¡,��E° ¿!³�µ!Þ * 6 ) �Ö�×�����ò�& ß!à Ý�³ Ö�×�Ø�Ù � ª�Ü/Ý

  • ��������� ��������� |�� � ���L� ��������������� ��! �"$#�%'&)(f*"�� |� ,+ � ��� �.-/10 ��2435�6879�:@?BA � P@~� L*r f�� 2/' W$ ���.����U D�� X v� Y p* "� � F�G �8VLL� ��F�GP$� K"�X,�E �.� � � 2 � "�X | D@ef�g�h 2*�� P� ��*� ?  >87��8¡ / KW�X��2�¢�£ D ] 6 ^_ ; � (�¤�¥¦*7 �=@?�ACB 8 Y � � ��~;�'�D� Ç Y;+�� W¦�>1º�E \GFHJI � [¶ ` KK� | "�E 3.2 L ���� ��GM�� (19740240) � H ú��ON .�� I$� � 0 |PRQTSRU

    1) Boeck, T., H. -A. Bahr, S. Lampenscherf and U. Bahr. : “Self-Driven Propagation of

    Crack Arrays: A Stationary Two-Dimensional Model”, Phys. Rev. E, 59, 1408–1416

    (1999).

    2) Groisman, A., and E. Kaplan : “An Experimental Study of Cracking Induced by

    Desiccation”, Europhys. Lett., 25, 415–420 (1994).

    3) Kitsunezaki, S. : “Fracture Patterns Induced by Desiccation in a Thin Layer”, Phys.

    Rev. E, 60, 6449–6464 (1999).

    4) Ito, H. and Y. Miyata : “Experimental Study on Mud Crack Patterns (in Japanese)”,

    J. Geol. Soc. Japan, 104, 90–98 (1998).

    * 8 ) VXWZY\[\]O^@_a`Gb7ced\fGgihK_DjDkilZmKkonGpaq*r7stY\u�_trvlehDwvr\wKxzyD{Zsvq\r"|�}t~Kxu�_

    8

  • 5) Bohn, S., J. Platkiewicz, B. Andreotti, M. Adda-Bedia and Y. Couder : “Hierarchi-

    cal Crack Pattern as Formed by Successive Domain Divisions. II. From Disordered

    to Deterministic Behavior” Phys. Rev. E, Vol. 71, No. 4, p. 046215, 2005.

    6) Allain, C. and L. Limat. : “Regular Pattern of Cracks Formed by Directional Drying

    of a Colloidal Suspension”, Phys. Rev. Lett., 74, 2981–2984 (1995).

    7) Müller, G. : “Starch Columns: Analog Model for Basalt Columns”, J. Geophys.

    Res., 103, 15239–15253 (1998).

    8) Toramaru, A. and T. Matsumoto : “Columnar Joint Morphology and Cooling Rate:

    a Starch-Water Mixture Experiment”, J. Geophys. Res., 109, B02205 (2004).

    9) Mizuguchi, T., A. Nishimoto, S. Kitsunezaki, Y. Yamazaki and I. Aoki : “Directional

    Crack Propagation of Granular Water Systems”, Phys. Rev. E, 71, 056122 (2005).

    10) Goehring, L., S. W. Morris and Z. Lin : “Experimental Investigation of the Scaling

    of Columnar Joints”, Phys. Rev. E, 74, 036115 (2006).

    11) Nishimoto, A., T. Mizuguchi and S. Kitsunezaki : “Numerical Study of Drying

    Process and Columnar Fracture Process in Granule-Water Mixtures”, Phys. Rev. E,

    76, 016102 (2007).

    12) Nishimoto, A., T. Mizuguchi and S. Kitsunezaki : “Columnar Structure in Desicca-

    tion Cracks (in Japanese)”, J. Geol. Soc. Japan (in press); “Starch Columnar Joints

    (in Japanese)”, BUTSURI (in press).

    13) Nakahara, A. and Y. Matsuo : “Imprinting Memory into Paste and Its Visualization

    as Crack Patterns in Drying Process”, J. Phys. Soc. Jpn., 74, 1362–1365 (2005);

    “Transition in the Pattern of Cracks Resulting from Memory Effects in Paste”, Phys.

    Rev. E, 74, 045102R (2006).

    14) Ooshida, T. : “Continuum Theory of Memory Effect in Crack Patterns of Drying

    Pastes”, Phys. Rev. E, 77, 061501 (2008).

    15) Müller, G. and T. Dahm : “Fracture Morphology of Tensile Cracks and Rupture

    Velocity”, J. Geophys. Res., 105, 723–738 (2000).

    16) Kitsunezaki, S. : “Crack Propagation Speed in the Drying Process of Paste”, J.

    Phys. Soc. Jpn., 78 (2009) (in press).

    9