Three-dimensional generalization and verification of structured bounding surface model for natural clay

As the proposed structured bounding surface model can only be used to solve planar strain problems of natural soft clay, a three-dimensional adaptive failure criterion is adopted to improve the model to capture the three-dimensional behaviors of natural soft clay. The three-dimensional adaptive fail...

Повний опис

Збережено в:
Бібліографічні деталі
Дата:2018
Автори: Cui Yunliang, Wang xinquan, Zhou lianying
Формат: Стаття
Мова:English
Опубліковано: НТК «Інститут монокристалів» НАН України 2018
Назва видання:Functional Materials
Теми:
Онлайн доступ:https://nasplib.isofts.kiev.ua/handle/123456789/154464
Теги: Додати тег
Немає тегів, Будьте першим, хто поставить тег для цього запису!
Назва журналу:Digital Library of Periodicals of National Academy of Sciences of Ukraine
Цитувати:Three-dimensional generalization and verification of structured bounding surface model for natural clay / Cui Yunliang, Wang xinquan, Zhou lianying // Functional Materials. — 2018. — Т. 25, № 1. — С. 138-143. — Бібліогр.: 13 назв. — англ.

Репозитарії

Digital Library of Periodicals of National Academy of Sciences of Ukraine
id nasplib_isofts_kiev_ua-123456789-154464
record_format dspace
spelling nasplib_isofts_kiev_ua-123456789-1544642025-02-23T18:07:36Z Three-dimensional generalization and verification of structured bounding surface model for natural clay Cui Yunliang Wang xinquan Zhou lianying Modeling and simulation As the proposed structured bounding surface model can only be used to solve planar strain problems of natural soft clay, a three-dimensional adaptive failure criterion is adopted to improve the model to capture the three-dimensional behaviors of natural soft clay. The three-dimensional adaptive failure criterion incorporated in this model can cover the Lade-Duncan criterion and the Matsuoka-Nakai criterion as its special ones. The structured bounding surface model is generalized into three-dimensional stress space by using the three-dimensional adaptive failure criterion. After improved with the three-dimensional adaptive failure criterion, the model can be seen as a modified bounding surface model which considers the destructuration and three-dimensional behaviors and neglects the anisotropy of natural soft clay. The simulations of undrained compression and extension tests of K0 consolidation state Bothkennar clay shows the unimportance of neglecting anisotropy in this model. It was validated on Pisa clay that the improved model can simulate well the three dimensional behaviors of natural soft clay under true triaxial conditions. The support of National Natural Science Foundation of China(Grant no. 51508507), Zhejiang Provincial Natural Science Foundation of China (Grant no. LQ16E080007), and Zhejiang Provincial Transportation Bureau Scientific Research Plan (Grant no. 2017004) are gratefully acknowledged. 2018 Article Three-dimensional generalization and verification of structured bounding surface model for natural clay / Cui Yunliang, Wang xinquan, Zhou lianying // Functional Materials. — 2018. — Т. 25, № 1. — С. 138-143. — Бібліогр.: 13 назв. — англ. 1027-5495 DOI:https://doi.org/10.15407/fm25.01.138 https://nasplib.isofts.kiev.ua/handle/123456789/154464 en Functional Materials application/pdf НТК «Інститут монокристалів» НАН України
institution Digital Library of Periodicals of National Academy of Sciences of Ukraine
collection DSpace DC
language English
topic Modeling and simulation
Modeling and simulation
spellingShingle Modeling and simulation
Modeling and simulation
Cui Yunliang
Wang xinquan
Zhou lianying
Three-dimensional generalization and verification of structured bounding surface model for natural clay
Functional Materials
description As the proposed structured bounding surface model can only be used to solve planar strain problems of natural soft clay, a three-dimensional adaptive failure criterion is adopted to improve the model to capture the three-dimensional behaviors of natural soft clay. The three-dimensional adaptive failure criterion incorporated in this model can cover the Lade-Duncan criterion and the Matsuoka-Nakai criterion as its special ones. The structured bounding surface model is generalized into three-dimensional stress space by using the three-dimensional adaptive failure criterion. After improved with the three-dimensional adaptive failure criterion, the model can be seen as a modified bounding surface model which considers the destructuration and three-dimensional behaviors and neglects the anisotropy of natural soft clay. The simulations of undrained compression and extension tests of K0 consolidation state Bothkennar clay shows the unimportance of neglecting anisotropy in this model. It was validated on Pisa clay that the improved model can simulate well the three dimensional behaviors of natural soft clay under true triaxial conditions.
format Article
author Cui Yunliang
Wang xinquan
Zhou lianying
author_facet Cui Yunliang
Wang xinquan
Zhou lianying
author_sort Cui Yunliang
title Three-dimensional generalization and verification of structured bounding surface model for natural clay
title_short Three-dimensional generalization and verification of structured bounding surface model for natural clay
title_full Three-dimensional generalization and verification of structured bounding surface model for natural clay
title_fullStr Three-dimensional generalization and verification of structured bounding surface model for natural clay
title_full_unstemmed Three-dimensional generalization and verification of structured bounding surface model for natural clay
title_sort three-dimensional generalization and verification of structured bounding surface model for natural clay
publisher НТК «Інститут монокристалів» НАН України
publishDate 2018
topic_facet Modeling and simulation
url https://nasplib.isofts.kiev.ua/handle/123456789/154464
citation_txt Three-dimensional generalization and verification of structured bounding surface model for natural clay / Cui Yunliang, Wang xinquan, Zhou lianying // Functional Materials. — 2018. — Т. 25, № 1. — С. 138-143. — Бібліогр.: 13 назв. — англ.
series Functional Materials
work_keys_str_mv AT cuiyunliang threedimensionalgeneralizationandverificationofstructuredboundingsurfacemodelfornaturalclay
AT wangxinquan threedimensionalgeneralizationandverificationofstructuredboundingsurfacemodelfornaturalclay
AT zhoulianying threedimensionalgeneralizationandverificationofstructuredboundingsurfacemodelfornaturalclay
first_indexed 2025-11-24T06:34:14Z
last_indexed 2025-11-24T06:34:14Z
_version_ 1849652469477408768
fulltext 138 Functional materials, 25, 1, 2018 ISSN 1027-5495. Functional Materials, 25, No.1 (2018), p. 138-143 doi:https://doi.org/10.15407/fm25.01.138 © 2018 — STC “Institute for Single Crystals” Three-dimensional generalization and verification of structured bounding surface model for natural clay Cui Yunliang, Wang xinquan, Zhou lianying School of Engineering, Zhejiang University City College, Hangzhou, Zhejiang, 310015, China Received December 1, 2017 As the proposed structured bounding surface model can only be used to solve planar strain problems of natural soft clay, a three-dimensional adaptive failure criterion is adopted to im- prove the model to capture the three-dimensional behaviors of natural soft clay. The three- dimensional adaptive failure criterion incorporated in this model can cover the Lade-Duncan criterion and the Matsuoka-Nakai criterion as its special ones. The structured bounding surface model is generalized into three-dimensional stress space by using the three-dimensional adap- tive failure criterion. After improved with the three-dimensional adaptive failure criterion, the model can be seen as a modified bounding surface model which considers the destructuration and three-dimensional behaviors and neglects the anisotropy of natural soft clay. The simula- tions of undrained compression and extension tests of K0 consolidation state Bothkennar clay shows the unimportance of neglecting anisotropy in this model. It was validated on Pisa clay that the improved model can simulate well the three dimensional behaviors of natural soft clay under true triaxial conditions. Keywords: natural soft clay, bounding surface, three-dimensional behavior, adaptive failure criterion Рассмотрен трехмерный адаптивный критерий отказа трехмерного поведения естественной мягкой глины, в котором использованы критерий Ладэ-Дункана и критерий Мацуока-Накаи. Структурная модель ограничивающей поверхности обобщается на трехмерное пространство напряжений с использованием трехмерного адаптивного критерия отказа. После его учета модель может рассматриваться как модифицированная модель ограниченной поверхности, которая учитывает разрушающее и трехмерное поведение и пренебрегает анизотропией естественной мягкой глины. Проведено моделирование недренированных грунтов на сжатие и растяжение, а также состояния консолидации. Подтверждено, что предложенная модель может хорошо имитировать трехмерное поведение естественной мягкой глины при истинных трехосных условиях воздействия нагрузок. Тривимірне узагальнення та перевірка структурованої моделі граничної поверхні для природної глини. Cui Yunliang, Wang xinquan, Zhou lianying. Розглянуто тривимірний адаптивний критерій відмови тривимірної поведінки природної м’якої глини, в якому використано критерій ладе-дункана і критерій мацуока-накаї. Структурна модель обмеженої поверхні узагальнюється на тривимірний простір напружень з використанням тривимірного адаптивного критерію відмови. Після його обліку модель може розглядатися як модифікована модель обмеженої поверхні, яка враховує руйнівну і тривимірну поведінку і нехтує анізотропією природної м’якої глини. Проведено моделювання недренованих грунтів на стиск і розтяг, а також стану консолідації. Підтверджено, що запропонована модель може добре імітувати тривимірну поведінку природною м’якої глини при істинних тривісних умовах впливу навантажень. Functional materials, 25, 1 2018 139 Cui Yunliang et al. / Three-dimensional generalization ... 1. Introduction Experimental results [1, 2] indicate that the stress-strain relationship curve of natural clay has a softening after peak stress in triaxial compression with low confining pres- sure, and the compression rate of one-dimen- sional compression becomes faster when the compression pressure exceeds the structure yielding stress. To capture the structured be- haviors of natural soft clay, many advanced constitutive models [3, 4] had been proposed. Based on the bounding surface concept initi- ated by Dafalias [5], AI-tabbaa and Wood [6] set a kinematic hardening yield surface, called ‘bubble surface’, inside the bounding surface to formulate a bubble model for soil. Rouainia and Wood [7] presented a structured bounding sur- face model, using the structure surface as the bounding surface and incorporating a structure measure of the bounding surface. The structure measure allows the size of the bounding sur- face to decay with plastic straining, so that the proposed model can describe the loss of struc- ture. Taiebat et al. [8] suggested a destructura- tion law to address both isotropic and frictional destructuration and applied it on SANICLAY model. The frictional destructuration is proved to have significant effect on the loss of structure [8]. Literature [9] proposed a simple bounding surface model incorporating the destructura- tion law with some modifications. The proposed model can be seen as a simplified model of the existing structured bounding surface models [6-7], because it neglected some complex prop- erties of soil, such as the kinematic hardening and anisotropy but considered the frictional de- structuration. However, the model proposed by work [9] is two-dimensional constitutive model. it can only be used to solve planar strain prob- lems of natural soft clay. It should be extended to be three-dimensional to capture the three- dimensional problems of natural soft clay. In order to describe the three-dimensional defor- mation of natural soft clay more precisely, the proposed model is generalized in three-dimen- sional principal stress space with a three-di- mensional adaptive failure criterion [10, 11]. The adopted criterion can cover the Lade crite- rion as well as the Matsuoka-Nakai criterion. The performance of the proposed model is veri- fied by typical experimental results on intact samples of natural soft clay. 2. Generalization of bounding surface model The yielding surface of bounding surface has the same elliptical shape with the Modified Cam-Clay model [7, 8]: F p p p q Mc= - +* *2 2 2/ (1) where, p and q are the mean effective stress and the generalized shear effective stress of the mapping point on the bounding surface of the current stress point, respectively; pc * is the intersection point of the bounding surface and the axial of p , which denotes the size of the bounding surface; M* is the critical state stress ratio that is the slope of the critical state line. By definition: p = + +σ σ σ1 2 3 3 (2) q = - + - + - 1 2 1 2 2 2 3 2 3 1 2( ) ( ) ( )σ σ σ σ σ σ (3) where σ σ σ1 2 3, , are principal stresses in three- dimensional stress. In the classical plastic theory, the basic con- stitutive equation can be expressed as follows: d C dp pε σ= × (4) wherein d pε is incremental plastic strain, dσ is incremental stress, and Cp is the plastic flex- ibility matrix. According to the associated flow rule, the plastic flexibility matrix can be presented as [10] C F F Kp T p = ¶ ¶ ¶ ¶ æ è ççç ö ø ÷÷÷÷σ σ (5) wherein Kp is the plastic hardening modulus. It is necessary to define the plastic hardening modulus in constitutive model. Work [9] pro- posed a structured bounding surface model. As presented in [9], K Kp p= +ζ ψP F p F q ba (( ) ( ) )( ) ¶ ¶ + ¶ ¶ -2 2 1 (6) The expression of Kp in eq. (6) can be seen in Ref. [9]. b is the measure of the distance between the loading surface which has been defined in work [9]. ζ andψ are interpolation parameters, reflecting the affect of the stress level on the modulus. Their values can be de- termined based on experimental curve fitting. Pa denotes the initial size of the bounding sur- face on axial of p in p q- stress space. The model proposed by work [9] is proved to be efficient and accurate for simulating the deg- radation of structure of natural clay. However, this model can only predict the two-dimension- al stress-strain relationship of natural for it is only a two-dimensional constitutive model. It can only be used to solve planar strain prob- lems. As we know, engineering problems are of- 140 Functional materials, 25, 1, 2018 Cui Yunliang et al. / Three-dimensional generalization ... ten three-dimensional problems, which should be solved with three-dimensional constitutive models. Thus, in order to simulate the three- dimensional behaviors of natural soft clay, a reasonable three dimensional adaptive failure criterion [10, 11] will be introduced to revise the proposed model in [9]. An adaptive crite- rion adopted here almost covers all the crite- ria between the Lade-Duncan criterion and the Matsuoka-Nakai criterion in principal stress space by adjusting the adaptive parameter m , including the Lade-Duncan criterion and the Matsuoka-Nakai criterion as its special cases. So, it is more adaptive compared with conven- tional criteria. The function of the three-dimen- sional adaptive failure criterion [10, 11] can be expressed as I I I I c c c c c 1 3 1 2 3 3 2 2 3 9 1 1 + = = - + - - - - m f m f f f f ( sin ) ( sin )( sin ) sin sin ++ sin3 fc (7) where m is a material constant which is deter- mined by true triaxial test; fc is the critical state internal friction angle for triaxial com- pression; I1, I2 and I3 are the first, second and third effective stress invariants, respectively. I I I 1 1 2 3 2 1 2 2 3 3 1 3 1 2 3 = + + = + + = ü ý ïïïï þ ïïïï s s s s s s s s s s s s (8) I1, I2 and I3 can be expressed as: I p I p q I p pq q 1 2 2 2 3 3 2 3 3 3 3 1 27 27 9 2 3 = = - = - + ü ý ïïïïïïïï þ ïïïïïï( cos )θ ïïï (9) wherein θ σ σ σ σ σ = - - - arctan ( )3 2 2 3 1 2 3 (10) Substituting (10) into (8), equation (8) can be written as p L q p L q3 2 2 3 3 0- + = (11) wherein L and L2 3 and L2 are as follows: L c c c c c 2 3 2 2 3 1 9 3 1 = -( ) + - -é ëê ù ûú - - - sin ( sin )( sin ) ( sin sin ϕ µ ϕ ϕ µ ϕ ϕ ++ ì í ïïï î ïïï ü ý ïïï þ ïïï -( ) + - - sin ) sin ( sin )( sin 3 3 23 3 1 9 ϕ ϕ µ ϕ ϕ c c c c)) ( )( sin sin sin ) é ëê ù ûú - + - - +é ëê ù ûú ì í ïïïï î ïïïï ü ý 27 3 1 2 3µ ϕ ϕ ϕc c c ïïïïï þ ïïïï = -( ) + - -é ëê ù ûúL c c c 3 3 22 3 1 9 3sin ( sin )( sin ) cos( )ϕ µ ϕ ϕ θ 227 3 1 9 729 243 1 3 2-( ) + - -é ëê ù ûú - + - sin ( sin )( sin ) ( )( sin ϕ µ ϕ ϕ µ ϕ c c c cc c c- + ì í ïïï î ïïï ü ý ïïï þ ïïï ü ý ïïïïïïïïïïïïïïïïï þ ïïï sin sin )2 3ϕ ϕ ïïïïïïïïïïïïïïï (12) It should be noted that L2 and L3 are only used to denote complex expressions, and they have no physical meanings. Derived from mathematical calibration, the shaping function of the criterion can be defined as [10,11] g L L ( ) cos arccos cos / q = - é ë ê ê ê ù û ú ú ú ì í ïïï î ïïï ü ý ïïï þ ïïï 1 3 3 3 2 3 2 3 2 11 3 3 3 3 2 3 2 3 2 arccos cos / - é ë ê ê ê ù û ú ú ú ì í ïïï î ïïï ü ý ïïï þ ïïï L L q (13) Bounding surface revised by the three-di- mensional adaptive failure criterion can be ex- pressed as Fig. 1. Comparison of prediction (solid lines) and experimental data (symbols) for undrained stress paths and stress-strain curves of triaxial compression and extension tests on Bothkennar clay following the K0 state at point A. (a) Comparison of stress-strain curves;(b) Comparison of stress paths. Functional materials, 25, 1 2018 141 Cui Yunliang et al. / Three-dimensional generalization ... F p p p q M gc= - +2 2 2 2 * * ( )q (14) where p S pc i c * = and M S Mf * = . Si is an iso- tropic destructuraion factor and Sf is a fric- tional destructuraion factor. Solving the partial derivatives of Eq. (15) with respect to p , q and q , respectively, it comes that ¶ ¶ = - F p p S pi c2 (15) ¶ ¶ = F q q S g Mf 2 2 2 2( )q (16) ¶ ¶ = - ¢F q g S M gfq q q 2 2 2 2 3 ( ) ( ) (17) In Eq. (17), ¢g ( )q is the derivative of g( )q with respect to q . Work [9] has already presented the expres- sion of plastic modulus at the bounding surface Kp . To generalize the expression of Kp into three-dimensional stress space, it can be re- vised by multiplying M by g( )q in the expres- sion of Kp in work [9]. M is the critical state stress ratio which is determined by the slope of the critical state line. Thus Kp can be written as, K S p e p F p p p S F p p i c c i mi = +( ) - ¶ ¶ - - -( ) - ( ) ¶ ¶ æ è çççç ö ø ÷÷÷÷ 1 1 0 λ κ λ κ β1- 22 2 + ¶ ¶ æ è çççç ö ø ÷÷÷÷ -β F q (18) - -( ) - - ¶ ¶ æ è çççç ö ø ÷÷÷÷ + ¶ ¶ æ è ççç 2 1 1 2 2 2 3 2 q g M S S F p F qf f mf ( ) ( ) q l k b b çç ö ø ÷÷÷÷ 2 where [9] S S i i m d p i = - - - ( )1 l k e (19) S S f i m d p f = - - - ( )1 l k e (20)   ε β ε βεd p v p q p= - +( )1 2 2 (21) In eq. (18)-(21), λ and κ are the slopes of the compression line and the swelling line in a volumetric strain-logarithmic mean stress plane, respectively. mi and mf are material constants which are determined by fitting the stress-strain curve of triaxial compression. εv p is volumetric plastic strain rate and εq p is de- viatoric plastic strain rate. β is a material con- stant distributing the effect of volumetric and deviatoric plastic strain rates to the value of εd p . β could be set to 0.5 as a default value. Si Sf The initial values of Si and Sf are needed in calculating with the proposed model. The initial isotropic destructuraion parameter Si0 is deter- mined by one-dimensional compression test and initial frictional destructuraion parameter Sf0 is determined by triaxial compression test. e0 is the initial void ratio. The plastic modulus at the current stress point, that is pK , can be derived by substitut- ing Eq. (18) into Eq. (6). 3. Model verification Natural soft clays are mainly in the state of K0 state which is an anisotropic state. To keep the proposed model simple, the anisotropy fea- ture of natural soft clay is neglected. However, this model is believed to be able to well capture the behaviors of natural anisotropic clay when structure is the main feature of the clay. In or- der to prove this statement and show the unim- portance of neglecting anisotropy, the proposed model is used to simulate the undrained triaxial compression and extension tests on intact sam- ples of Bothnennar clay following the in-situ K0 state. The mechanical response and stress- strain relationship of Bothkennar clay have been studied by Smith et al. [12] and Taiebat et al. [8], respectively. The samples were tak- en with Laval and Sherbrooke samplers from depths of 5.3-6.2 m and only the experimen- tal data of Laval samples is used in this work. The samples were reconsolidated to in-situ K0 state before shear test. The K0 state stresses are σ1 46= kPa , σ σ2 3 28= = kPa . Then, the initial mean effective stress p kPain = 34 . Fig. 2. Stress paths and failure points of true axial tests of Pisa soft clay 142 Functional materials, 25, 1, 2018 Cui Yunliang et al. / Three-dimensional generalization ... The stress-strain data of the undrained tri- axial compression and extension tests on intact samples of Bothnennar clay following the in- situ K0 state is shown in Figure 1 by discrete points. To simulate the experiments, the pro- posed constitutive model proposed in this work is coded in FORTRAN. The parameters for the proposed model are obtained according to the study of Smith et al. [12] and Taiebat et al. [8] and shown in Table 1. The lines in Figure 1 show the stress-strain curve obtained by the simulation. In Figure 1, pin is the initial mean effective stress which equal to 34 kPa, ε1 is the axial strain in compression or extension test. It can be seen in Figure 1 that the simulation stress- strain curves of the triaxial compression and extension tests both show good match to the ex- perimental data. Thus, The proposed model is accurate for simulating the destructuration of natural clay. From a practical point of view, it is acceptable to neglect the anisotropy feature in the proposed model for natural soft clay to keep the model simple. The samples of the true triaxial compression tests performed by Callisto [2, 13] were sampled under the Pisa tower. All of them were recon- solidated to the in-situ stress state at point O of Figure 2 (σ1 113 5= . kPa , σ σ2 3 75 5= = . kPa ). From point O, drained compression tests were performed along rectilinear stress paths with different orientations in the stress space. Tests are labeled with the prefix ‘T’, followed by a rel- evant value of angle α . In Figure 2, the rect- angular points are failure stress point normal- ized in plane π . Choosing an appropriate value for the adaptive parameter µ and fitting the failure points with the three-dimensional adap- tive failure criterion gives the optimized fitting curve in Figure 2 with µ =1.5. Adaptive crite- Table 1. Parameters of the model for Bothkennar clay l k n M Si0 Sf0 mi mf b V y 0.255 0.03 0.2 1.4 6.0 1.3 1.0 1.0 0.5 15.0 0.8 Fig. 3. Comparisons of experimental and simulation curves of the relationships of LSP versus following the T0, T60, T90 and T150 stress paths for Pisa clay. Functional materials, 25, 1 2018 143 Cui Yunliang et al. / Three-dimensional generalization ... rion in Figure 2 means the three-dimensional adaptive failure criterion expressed by eq.(7). It can be seen in Figure 2 that the adaptive cri- terion can fit the failure points better compared with the lade criterion, except of the failure point of test T150. The inaccuracy in fitting the failure point of test T150 is caused by the limit of the adopted three-dimensional adaptive fail- ure criterion that this criterion is isotropic but natural soils are always anisotropic. It seems that the anisotropy of soil should be considered in later research. To verify the proposed constitutive model generalized by the three-dimensinal failure criterion, the model is computed in FORTRAN and used to simulate the true triaxial compres- sion tests on natural Pisa clay. The optimized values of the parametersmi , mf , ς and ψ are obtained by adjusting the values to fit the ex- perimental curve of test T0 with the proposed model. The simulation of other tests takes the same model parameters with test T0. The model parameters are listed in Table 2. The compari- sons of simulation (heavy line) and observa- tion (thin line) of tests T0, T60, T90 and T150 are shown in Figure 3. LSP and εs denote the lengths of stress path and the deviatoric strain invariant, respectively. The expressions are LSP = - + - + -( ) ( ) ( )σ σ σ σ σ σ1 2 2 1 3 2 2 3 2 and ε ε ε ε ε ε εs = - + - + -2 3 1 2 2 1 3 2 2 3 2 1 2/ [( ) ( ) ( ) ] / where ε1,2,3 are axial strains in true triaxial compression respectively. It can be seen that the simulation curves can well fit the experi- mental curves. Therefore, the proposed model is demonstrated to be able to well capture the behaviors of natural soft clay under general stress conditions. 4. Conclusions The three dimensional adaptive failure criterion incorporated in this model can cover the Lade-Duncan criterion and the Matsuoka- Nakai criterion as its special ones. It is a good choice to use this failure criterion to generalize the bounding surface model into three-dimen- sional space. The proposed model improved with the three dimensional adaptive failure criterion can cap- ture the three-dimensional behaviors of natural clay, but neglect the anisotropy of natural soft clay. However, it can be proved by simulations of undrained compression and extension tests of K0 consolidation state Bothkennar clay, that it is not important to neglect anisotropy in this model. It can be verified by simulating the true tri- axial tests of Pisa clay, that improved model can simulate the three dimensional behaviors of natural soft clay under true triaxial condi- tions. Acknowledgements The support of National Natural Science Foundation of China(Grant no. 51508507), Zhe- jiang Provincial Natural Science Foundation of China (Grant no. LQ16E080007), and Zhejiang Provincial Transportation Bureau Scientific Research Plan (Grant no. 2017004) are grate- fully acknowledged. References 1. D. Wang and L. Korkiala-Tanttu, , Engin.Geol., 214, 20, 2016. 2. L. Callisto, G. Calabresi, Géotechn, 48, 495, 1998. 3. E. L. Liu, H. S. Yu, C. Zhou, et al, Intern. J.Geomech., 17, 1, 2017. 4. Amir Hamidi, Saeed Tourchi and Fatemeh Kar- dooni, J. Rock Mech. Geotechn.Eng., 9, 1094, 2017. 5. Y. F. Dafalias, L. R. Herrmann, J. Eng. Mech. ASCE, 112, 1263, 1986. 6. A. AI-Tabbaa, D. M. Wood, Proceeding of the 3rd International Conference on Numerical Mod- els in Geomechanics, Niagara Falls, pp.91-99, 1989. 7. M. Rouainia, D. M. Wood, Géotechn., 50, 153, 2000. 8. M. Taiebat, Y. F. Dafalias, R. Peek, Intern. J. Num. Anal. Meth. Geomech., 34. 1009, 2010. 9. Y. L. Cui, X. Q. Wang, S. M. Zhang, Functional Materials, 24. 628, 2017. 10. Y. Xiao, H. L. Liu and R. Y. Liang, Scie. China Ser.E, 54, 805, 2011. 11. Y. Xiao, H. L. Liu, J. G. Zhu, et al., Scie. China Ser.E, 55, 2877, 2012. 12. P. R. Smith, R. J. Jardine, D. W. Hight, Géo- techn.,42 , 257, 1992. 13. L. Callisto L, A. Gajo, D. M. Wood, Géotechn., 52, 649, 2002. Table 2. Parameters of the model for Pisa clay l k n M Si0 Sf0 mi mf b V y µ 0.110 0.024 0.2 0.983 3.0 1.4 0.9 0.9 0.5 12.0 1.1 1.5