1.School of Energy and Power Engineering, University of Shanghai for Science and Technology, Shanghai 200093, China 2.Shanghai Key Laboratory of Multiphase Flow and Heat Transfer in Power Engineering, Shanghai 200093, China
Fund Project:Project supported by the Shanghai Natural Science Foundation, China (Grant No. 19ZR1435700) and the National Natural Science Foundation of China (Grant No. 51736007)
Received Date:05 June 2019
Accepted Date:15 July 2019
Available Online:01 November 2019
Published Online:05 November 2019
Abstract:A new incompressible gas-liquid two-phase flow model for non-Newtonian power-law fluid is proposed based on an incompressible lattice Boltzmann model. And the fundamental physical mechanism of Newtonian fluid displacing non-Newtonian power-law fluid liquid in porous medium is studied by using the proposed model. The effects of capillary number Ca, dynamic viscosity ratio M, surface wettability θ, porous medium geometry, and power law index n on the displacement process are investigated. The comprehensive results show that with the increase of capillary number, the displacement process turns faster, the fingering phenomenon becomes more obvious and the displacement efficiency decreases. However, for different values of power-law index n, the effects of the Ca on the displacement process have some differences. Specially, the decrease rate of displacement efficiency becomes slow if the displaced fluid is shear thickening fluid as compared with that if the displaced fluid is shear thinning fluid. On the other hand, the displacement efficiency decreases as dynamic viscosity ratio M increases. And the effect of the viscosity ratio on the displacement process becomes more obvious for the low value of the power-law index n. Moreover, the effect of the surface wettability of the porous medium on the displacement process is also related to the size of the power-law index. With the increase of the contact angle of the porous medium, the fingering phenomenon turns less obvious, and the displacement efficiency increases. However, with the increase of power-law index n, the influence of the contact angle on the displacement process decreases. Besides, the displacement processes with different geometric types of the porous media are also studied in the work. The results show that comparing with the case of porous medium denoted by circle shape and square shape, the fingering phenomenon obtained by the case of triangular shape is most obvious, and the displacement efficiency is lowest. Keywords:power-law two-phase fluid/ lattice Boltzmann model/ immiscible displacement/ porous media
图 3 稳态接触角$\theta $与指标参数${\phi _{{\rm{wall}}}}$的线性关系 Figure3. Linear relationship between steady state contact angle $\theta $ and the order parameter of a solid wall ${\phi _{{\rm{wall}}}}$.
23.3.T型通道内液滴分离 -->
3.3.T型通道内液滴分离
本小节验证在不同$Ca$数下, T型微通道内形成液滴的大小. T型微管道结构如图4所示, 其中W0 = W1 = 30, L = 520, Y1 = 75, Y2 = 120. 初始分散相的子管道充满牛顿液相流体, 连续相主通道充满幂律流体, 其幂律指数n = 0.4. 边界条件设置为: 连续相与分散相的进口是速度进口边界, 连续相出口是对流边界条件[30], 管径的固壁面都采用无滑移边界条件. 图5给出了在不同的$Ca$数下, 分离的牛顿流体液滴(黄色). 图6给出了在不同的$Ca$数下, 分离的牛顿流体液滴尺寸, 并与数值结果[41]进行对比, 得到了一致的结果. 图 4 T型通道问题物理模型 Figure4. Physical model for the case of T shape channel.
图 5 不同Ca数对应的液滴形态 (a) Ca = 0.06370; (b) Ca = 0.06835; (c) Ca = 0.07300; (d) Ca = 0.07750; (e) Ca = 0.0820; (f) Ca = 0.08650; (g) Ca = 0.0910 Figure5. Droplet morphology obtained under various values of Ca: (a) Ca = 0.06370; (b) Ca = 0.06835; (c) Ca = 0.07300; (d) Ca = 0.07750; (e) Ca = 0.0820; (f) Ca = 0.08650; (g) Ca = 0.0910.
图 6 在剪切变稀幂律流体中, 不同的$Ca$数下形成液滴的无量纲直径(其中D是形成的液滴的直径, H是管径的直径) Figure6. Droplet dimensionless diameters at different values of $Ca$ in shear thinning power-law fluid. D is diameters of the droplet and H is width of the main channel.
-->
4.1.Ca数的影响
本小节研究Ca数对不混溶幂律流体驱替过程的影响. 在数值模型中M × N = 240 × 600, A = 30, B = 20, S = 6, S1 = 40, X = 60, Y = 60, 孔隙率$\xi = 0.8507$. 初始动力黏度比M = 5.0, 即被驱替液初始运动黏度${\nu _{\rm{l}}} = 1/6$, 动力黏度${\eta _{\,\rm{l}}} = 0.08333$, $\sigma = 0.0056$. 固体表面均为中性润湿($\theta = {90^{\rm{o}}}$). 图8给出了在不同Ca数下被驱替液为剪切变稀、牛顿以及剪切变稠三种流体时得到的驱替完成时指进形态图. 从图8可以看出, 不论被驱替液是剪切变稀(图8(a)—(c)n = 0.7)、牛顿(图8(d)—(f)n = 1.0)还是剪切变稠(图8(g)—(i)n = 1.3)流体, 都有Ca数越大, 指进现象越明显[17], 驱替完成时花费的时间越少. 具体而言, 当n = 0.7时(图8(a)—(c)), Ca = 0.0298对应的驱替完成时间t = 36.2, 当Ca数增加到0.0877时, 驱替时间减少到了t = 11.1, 减少了69.3%; 而当n = 1.0时(图8(d)—(f))和n = 1.3 (图8(g)—(i))时, Ca数从0.0298增加到0.0877, 驱替时间分别减少了68.8%和67.7%. 发生以上现象的原因是由于Ca数是表征黏性力与表面张力比值的无量纲参数, Ca数越大说明表面黏性力越大而表面张力越小, 而随着黏性力的增加, 驱替过程受到的阻力增加, 随着表面张力减小, 气液界面更容易发生变形, 因此Ca数越大指进现象越明显. Shiri等[48]以及Liu等[49]在研究多孔介质内的驱替问题时也得到了相似的结论. 另一方面, 从图8还可以发现当Ca数相同时, 当被驱替相的幂率指数n越大时, 指进现象越明显, 驱替完成所需时间越短[17]. 例如当Ca = 0.0298时(图8(a)、图8(d)、图8(g)), n = 0.7, 1.0, 1.3对应的驱替完成时间分别为36.2, 32.4, 29.1; Ca = 0.0595时(图8(b)、图8(e)、图8(h)), n = 0.7, 1.0, 1.3对应的驱替完成时间分别为16.9, 15.3, 14.2; 而Ca = 0.0877时(图8(c)、图8(f)、图8(i)), n = 0.7, 1.0, 1.3对应的驱替完成时间分别为11.1, 10.1, 9.4. 因此对应这三种Ca数的情况, 随着幂率指数n的增加驱替时间分别减小了19.6%, 16.0%和15.3%. 也就是说, 幂率指数n越大, Ca数的增加导致的驱替时间减少的速率越来越慢. 导致该现象的原因是对于剪切变稠流体随着驱替过程的进行其动力黏度会大于初始时刻的动力黏度, 剪切变稀流体的动力黏度会小于初始时刻的动力黏度, 而牛顿流体的动力黏度会保持不变. 为了说明这一现象, 图9给出了Ca = 0.0298时对应驱替完成时刻的两相流体的动力黏度分布. 从图9可以看出, 对于研究的所有n的值, 当驱替过程完成时, 驱替气相的动力黏度一直保持在初始值0.01667附近, 而对应幂律指数n = 0.7 (图9(a))、1.0 (图9(b))以及1.3 (图9(c))的被驱替液得到的动力黏度分别为0.04409, 0.08333与0.1579. 即剪切变稀、牛顿以及剪切变稠流体被驱替过程中, 其分别对应的两相动力黏度比M小于5、等于5以及大于5. 而两相动力黏度比越大, 两相流体间黏性力影响变大, 即所受黏性阻力越大, 因此指进现象越明显, 驱替越快[15,17]. 图 8 不同的$Ca$数下, 被驱替液为剪切变稀、牛顿与剪切变稠流体时得到的指进形态图 (a)?(c) n = 0.7; (d)?(f) n =1.0; (g)?(i) n = 1.3 Figure8. Final finger patterns obtained under different values of Ca for shear thinning, Newtonian and shear thickening fluids: (a)? (c) n = 0.7; (d)?(f) n =1.0; (g)?(i) n = 1.3.
图 9 驱替完成时, 不同幂律指数情况下得到的气液两相动力黏度示意图 $({\rm{a}})\;n = 0.7$; $({\rm{b}})\;n = 1.0$; $({\rm{c}})\;n = 1.3$ Figure9. Schematic diagram of gas-liquid two phase dynamics viscosity obtained under different values of power-law exponent: $({\rm{a}})\;n = 0.7$; $({\rm{b}})\;n = 1.0$; $({\rm{c}})\;n = 1.3$.
为了定量分析Ca数以及幂律指数n对驱替过程的影响, 图10给出了不同Ca数以及幂律指数n下得到的驱替效率. 从图10可以看出, 不论被驱替液是剪切变稀、牛顿还是剪切变稠流体, 都有Ca数越大, 驱替效率De越低. 具体而言, n = 0.7时, Ca数从0.0298到0.0877时驱替效率De的值从0.744减小到0.688, 减小了7.52%; n = 1.0时, Ca数从0.0298到0.0877时驱替效率De的值从0.663减小到0.617, 减小了6.93%; n = 1.3时, Ca数从0.0298到0.0877时驱替效率De的值从0.590减小到0.550, 减小了6.78%. 从以上分析以及图10中曲线变化可以发现, 不论被驱替相是牛顿流体还是幂律流体, 驱替效率随着Ca数的增加而减小, 然而驱替效率减小的速率随着n的增加而减小. 另外, Ca数一定时, 幂律指数n越大, 驱替效率De越低. 图 10$Ca$ 数和幂律指数n对幂律流体驱替效率的影响 Figure10. Effects of $Ca$and power-law exponent n on power-law fluid displacement efficiency.
24.2.动力黏度$M$的影响 -->
4.2.动力黏度$M$的影响
本小节研究动力黏度比M对不混溶幂律流体驱替过程的影响, 这里黏性比的增加是通过改变被驱替液的黏性实现的. 在本小节Ca数设置为0.0446, 其余参数与4.1节相同. 图11给出了被驱替液为剪切变稀、牛顿以及剪切变稠三种流体在不同初始动力黏度比M的情况下, 驱替完成时指进形态图. 从图11可以发现, 对于被驱替液是剪切变稀流体的情况(图11(a)—(c)n = 0.7), 黏度比从2.5增加到12.5, 驱替时间从26.5减少到19.7, 减少了25.7%. 对于被驱替液为牛顿流体(图11(d)—(f)n = 1.0)和剪切变稠(图11(g)—(i)n = 1.3)流体的情况, 动力黏度比从2.5增加到12.5时对应的驱替时间分别减少了19.6%和9.3%. 因此对于所有n的情况都有初始动力黏度比M越大, 指进现象越明显, 驱替完成时所花费的时间越少[15,17]. 这是因为黏性比越大说明被驱替液黏性越大, 而一个轻流体(黏性较小)驱替一个重流体(黏性较大)是非常困难的, 因此指进现象会更明显. 其他****[48—51]也发现了类似的现象. 从以上数据以及指进形态图还可以发现驱替相的幂律指数n越大, 黏性比的增加对驱替过程的影响越小. 另一方面从图11可以观察到当初始动力黏度比M相同时, 被驱替相的幂律指数n越大, 指进现象越明显, 对应的驱替完成所花费的时间越少. 这与4.1节流体越黏稠, 流体越难被驱替结论一致[15,17]. 如M = 2.5 (图11(a)、图11(d)、图11(g))时, 对应n = 0.7, 1.0和1.3的驱替完成时间分别为26.5, 23.5和20.3, 此时随着幂率指数n的增加, 驱替完成时间减小了23.4%. 图 11 不同的动力黏性比M下, 被驱替液为剪切变稀、牛顿与剪切变稠流体时得到的指进形态图 (a)?(c) n = 0.7; (d)?(f) n = 1.0; (g)?(i) n = 1.3 Figure11. Final finger patterns obtained under different values of viscosity ratios M for shear thinning, Newtonian and shear thickening fluids: (a)?(c) n = 0.7; (d)?(f) n = 1.0; (g)?(i) n = 1.3.
M = 7.5 (图11(b)、图11(e)、图11(h))时, 对应以上三个幂率指数的驱替完成时间分别为21.4, 19.6和18.5. M = 12.5 (图11(c)、图11(f)、图11(i))时对应的驱替完成时间分别为19.7, 18.9和18.4. 因此对应这两种黏性比的情况, 随着幂率指数增加驱替时间分别减小了13.6%和6.6%. 从以上数据分析可以发现随着黏性比的增加, 驱替速率减少, 且当黏性比较小时, 幂律指数n越大, 流体越难被驱替, 而随着黏性比的增加, 被驱替相是牛顿流体和幂律流体的驱替结果之间的差异越来越小. 为了进一步分析初始动力黏度比M以及幂律指数n对驱替过程的影响, 图12给出了不同初始动力黏度比M以及幂律指数n下得到的驱替效率. 从图12可以看出, 当不论被驱替液是剪切变稀、牛顿还是剪切变稠流体, 都有动力黏度比M越大, 驱替效率De越低, 而且随着n的增加得到的驱替效率曲线的斜率越来越平缓, 说明黏性比M对于驱替效率的影响随着n的增加而减小. 具体而言, n = 0.7时, M从2.5到12.5时驱替效率De的值从0.827减小到0.596, 减小了27.93%; n = 1.0时, M从2.5到12.5时驱替效率De的值从0.730减小到0.562, 减小了23.01%; n = 1.3时, M从2.5到12.5时驱替效率De的值从0.626减小到0.535, 减小了14.54%. 因此, 不论被驱替相是牛顿流体还是幂律流体, 驱替效率随着M的增加而减小, 然而驱替效率减小的速率随着n的增加而减小. 另外, M一定时, 幂律指数n越大, 驱替效率De越低. 图 12 动力黏度比M和幂律指数n对幂律流体驱替效率的影响 Figure12. Effects of viscosity ratio M and power-law exponent n on power-law fluid displacement efficiency.