MPEMathematical Problems in Engineering1563-51471024-123XHindawi10.1155/2020/43175384317538Research ArticleAnalyzing 3D Advection-Diffusion Problems by Using the Improved Element-Free Galerkin Methodhttps://orcid.org/0000-0003-4232-5631ChengHeng1ZhengGuodong2DaiYing1School of Applied ScienceTaiyuan University of Science and TechnologyTaiyuan 030024Chinatyust.edu.cn2Shanghai Key Laboratory of Mechanics in Energy EngineeringShanghai Institute of Applied Mathematics and MechanicsSchool of Mechanics and Engineering ScienceShanghai UniversityShanghai 200072Chinashu.edu.cn202014820202020140720202507202014820202020Copyright © 2020 Heng Cheng and Guodong Zheng.This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

In this paper, the improved element-free Galerkin (IEFG) method is used for solving 3D advection-diffusion problems. The improved moving least-squares (IMLS) approximation is used to form the trial function, the penalty method is applied to introduce the essential boundary conditions, the Galerkin weak form and the difference method are used to obtain the final discretized equations, and then the formulae of the IEFG method for 3D advection-diffusion problems are presented. The error and the convergence are analyzed by numerical examples, and the numerical results show that the IEFG method not only has a higher computational speed but also can avoid singular matrix of the element-free Galerkin (EFG) method.

National Natural Science Foundation of China11571223Science and Technology Innovation Project of Shanxi Colleges and Universities2020L0344
1. Introduction

Currently, meshless method has been applied successfully to deal with various kinds of problems in the fields of science and engineering. When solving large deformation problems and dynamic propagation of cracks, meshless method can obtain greater precision than finite element method .

As an important meshless method, the EFG method has been applied to different kinds of engineering problems. In the EFG method, the moving least-squares (MLS) approximation is employed to construct the shape function. Because the MLS approximation is based on the least-squares method [2, 3], the disadvantages of the least-squares method also exist in the MLS approximation, in which sometimes ill-conditional or singular matrices occur in the final equations.

Cheng and Chen studied the IMLS approximation  which can make up for the deficiency of the MLS approximation; then the IEFG method is applied for transient heat conduction , wave equation , fracture , elastoplasticity , and viscoelasticity  problems. The improved complex variable EFG method is presented for wave equation problem  and bending problem of thin plate on elastic foundations . The IEFG and the improved complex variable EFG method can enhance the computational speed of the EFG method.

Based on the MLS approximation with a singular weight function, the interpolating MLS method was proposed by Lancaster et al. , and the corresponding meshless method can be applied with the essential boundary condition directly. Based on the interpolating MLS method, Kaljevic et al. proposed the improved formulation of EFG method . Based on the concept of an inner production, Ren et al. improved the interpolating MLS method by using singular weight function in interpolating points and orthogonalizing some of basis functions . And the interpolating EFG method is presented for potential [15, 16], transient heat conduction , elasticity , viscoelasticity , elastoplasticity , and elastic large deformation  problems.

In order to overcome the difficulties caused by singular weight function in the interpolating MLS method, an improved interpolating least-squares (IIMLS) method based on nonsingular weight function was presented by Wang et al. Based on the IIMLS method, the improved interpolating EFG method is presented for potential , elasticity , and some complex mechanical problems .

By combining the dimension splitting method and meshless methods, the hybrid complex variable EFG method , the dimension split EFG method , and the dimension splitting reproducing kernel particle method  for 3D problems are proposed, respectively, and those new methods can improve the computational speed of traditional meshless methods for solving 3D problems greatly.

In this paper, combining the IMLS approximation and the Galerkin weak form, the IEFG method is used for solving 3D advection-diffusion problems. The IMLS approximation is used to form the trial function, the penalty method is applied to introduce the essential boundary conditions, the Galerkin weak form and the difference method are used to obtain the final discretized equations, and then the formulae of the IEFG method for 3D advection-diffusion problems are presented.

In the section of numerical examples, the weight functions, the scale parameter, the penalty factor, the node distribution, and the time step are discussed, respectively. The numerical results show that the IEFG method in this paper has a higher computational speed. And the advantage that the IEFG method can avoid the singular matrix is given.

2. The IMLS Approximation

The approximation of a function ux is defined as(1)uhx=i=1mpixaix=pTxax,xΩ,where pTx is the basis function vector, m is the basis function number, and(2)aTx=a1x,a2x,,amx,is the corresponding coefficient vector of pTx.

In general,(3)pTx=1,x1,x2,x3,pTx=1,x1,x2,x3,x12,x22,x32,x1x2,x2x3,x1x3.

The local approximation of equation (1) is(4)uhx,x^=i=1mpix^aix=pTx^ax.

Define(5)J=I=1nwxxIuhx,xIuI2=I=1nwxxIi=1mpixIaixuI2,where wxxI is a weighting function which contains compact support, and xII=1,2,,n are the nodes with influence domains covering the point x.

Equation (5) can be written as(6)J=PauTWxPau,where(7)uT=u1,u2,,un,P=p1x1p2x1pmx1p1x2p2x2pmx2p1xnp2xnpmxn,Wx=wxx1000wxx2000wxxn.

From(8)Ja=AxaxBxu=0,we have(9)Axax=Bxu,where(10)Ax=PTWxP,Bx=PTWx.

Equation (9) sometimes forms singular or ill-conditional matrix. In order to make up for this deficiency, for basis functions(11)q=qi=1,x1,x2,x3,x12,x22,x32,x1x2,x2x3,x3x1,,using Gram-Schmidt process, we can make the orthogonal basis functions as(12)pi=qik=1i1qi,pkpk,pkpk,i=1,2,3,,pi,pj=0,ij.

Then from equation (9), ax can be obtained directly as(13)ax=AxBxu,where(14)Ax=1p1,p10001p2,p200001pn,pn.

Substituting equation (13) into equation (4), we have(15)uhx=Φxu=I=1nΦIxuI,where(16)Φx=Φ1x,Φ2x,,Φnx=pTxAxBxis the shape function.

This is the IMLS approximation, in which ax can be obtained simply and directly because it is unnecessary to obtain the inverse matrix of Ax. Therefore, the IMLS approximation can avoid ill-conditional or singular matrix; then it can improve the computational precision and efficiency of the MLS approximation .

3. The IEFG Method for 3D Advection-Diffusion Problems

For 3D advection-diffusion problems, the governing equation is(17)utux1k1ux1ux2k2ux2ux3k3ux3+vx1ux1+vx2ux2+vx3ux3=fx,t,x=x1,x2,x3Ω,with the boundary conditions(18)ux,t=u¯x,t,xΓu,qx,t=k1ux,tx1n1+k2ux,tx2n2+k3ux,tx3n3=q¯x,t,xΓq,and the initial condition(19)ux,t=u0,where ux,t is the field function, u¯x,t is the given field function on the essential boundary Γu, q¯x,t is the given value on the natural boundary Γq, Γ=ΓuΓq is the boundary of the problem domain Ω, and ΓuΓq=; fx,t is the source term; ki is the diffusion efficient in the direction xi and vxi is the advection velocity in the direction xi; u0 is known function; ni is the unit outward normal to the boundary Γ in the direction xi.

The equivalent functional of 3D advection-diffusion problem can be obtained as(20)Π=ΩuutfdΩ+Ω12k1ux12+k2ux22+k3ux32dΩ+Ωuvx1ux1+vx2ux2+vx3ux3dΩΓquq¯dΓ.

Suppose that α is the penalty factor; by introducing the penalty method, we can obtain(21)Π=Π+α2Γuuu¯uu¯dΓ.

From(22)δΠ=0,we can obtain that the equivalent integral weak form is(23)δΠ=ΩδuutdΩΩδufdΩ+ΩδLuTk˜LudΩ+Ωδuvx1ux1dΩ+Ωδuvx2ux2dΩ+Ωδuvx3ux3dΩΓqδuq¯dΓ+αΓuδuudΓαΓuδuu¯dΓ,where(24)k˜=k1000k2000k3.(25)L=x1x2x3.

In the cubic domain Ω, we employ M nodes xII=1,2,,M. The function ux,t at the node xI is(26)uI=uxI,t.

From the IMLS approximation, at the time of t, the function at an arbitrary node x is(27)ux,t=Φxu=I=1nΦIxuI,where(28)u=u1,u2,,unT.

From equations (25) and (27), we have(29)ux,tt=tI=1nΦIxuI=I=1nΦIxuIt=Φxu˙,(30)Lux,t=I=1nx1x2x3ΦIxuI=I=1nBIxuI=Bxu,where(31)u˙=ux1,tt,ux2,tt,,uxn,ttT,Bx=B1x,B2x,,Bnx,BIx=ΦI,1xΦI,2xΦI,3x.

Substituting equations (27), (29), and (30) into equation (23) yields(32)ΩδΦxuΦxu˙dΩΩδΦxufdΩ+ΩδBxuTk˜BxudΩ+ΩδΦxuvx1x1ΦxudΩ+ΩδΦxuvx2x2ΦxudΩ+ΩδΦxuvx3x3ΦxudΩΓqδΦ(x)uq¯dΓ+αΓuδΦxuΦxudΓαΓuδΦxuu¯dΓ=0.

Analyzing the integral terms in equation (32), respectively, we have(33)ΩδΦxuΦxu˙dΩ=δuTΩΦTxΦxdΩu˙=δuTCu˙,(34)ΩδΦxufdΩ=δuTΩΦTxfdΩ=δuTF1,(35)ΩδBxuTk˜BxudΩ=δuTΩBTxk˜BxdΩu=δuTKu,(36)ΩδΦxuvx1x1ΦxudΩ=δuTΩΦTxvx1x1ΦxdΩu=δuTG1u,(37)ΩδΦxuvx2x2ΦxudΩ=δuTΩΦTxvx2x2ΦxdΩu=δuTG2u,(38)ΩδΦxuvx3x3ΦxudΩ=δuTΩΦTxvx3x3ΦxdΩu=δuTG3u,(39)ΓqδΦxuq¯dΓ=δuTΓqΦTxq¯dΓ=δuTF2,(40)αΓuδΦxuΦxudΓ=δuTαΓuΦTxΦxdΓu=δuTKαu,(41)αΓuδΦxuu¯dΓ=δuTαΓuΦTxu¯dΓ=δuTFα,where(42)C=ΩΦTxΦxdΩ,F1=ΩΦTxfdΩ,K=ΩBTxk˜BxdΩ,G1=ΩΦTxvx1x1ΦxdΩ,G2=ΩΦTxvx2x2ΦxdΩ,G3=ΩΦTxvx3x3ΦxdΩ,F2=ΓqΦTxq¯dΓ,Kα=αΓuΦTxΦxdΓ,Fα=αΓuΦTxu¯dΓ.

Substituting equations (33)–(41) into equation (32), we can obtain(43)δuTCu˙+Ku+Kαu+G1u+G2u+G3uF1F2Fα=0.

δuT is arbitrary in equation (43); then the second-order ordinary differential equations can be obtained as follows:(44)Cu˙+K^u=F^,where(45)K^=K+Kα+G1+G2+G3,F^=F1+F2+Fα.

Then, using the finite difference method to separate the time, the relation of ut+Δt and ut can be established as(46)θutt+Δt+1θutt=ut+ΔtutΔt,where Δt is the time step; solving equation (44) for u/tt+Δt and u/tt, respectively, and substituting the results into equation (44), we have(47)CΔt+θK^n+1un+1=CΔt1θK^nun+θF^n+1+1θF^n,where θ is a time weighed coefficient; we select θ=1 in numerical examples.

4. Numerical Examples

Three numerical examples of 3D advection-diffusion problems are solved with the EFG and the IEFG method, and the computational accuracy and efficiency are compared, respectively.

Define the relative error as(48)uuhL2Ωrel=uuhL2ΩuL2Ω,where(49)uuhL2Ω=Ωuuh2dΩ1/2.

In this section, the linear basis function is employed and the node distribution is regular. Moreover, 3×3×3 Gaussian points are used in an integral cell.

The first example we considered is a 3D advection-diffusion problem:(50)ux,tt=122ux,tx12+2ux,tx22+2ux,tx3212ux,tx1+ux,tx2+ux,tx3,xΩ,t0,T,with the initial condition(51)ux,0=ex1+ex2+ex3,and the boundary conditions(52)ux1=0=1+ex2+ex3et,ux1=1=e1+ex2+ex3et,ux2=0=ex1+1+ex3et,ux2=1=ex1+e1+ex3et,ux3=0=ex1+ex2+1et,ux3=1=ex1+ex2+e1et.

The problem domain is Ω=0,1×0,1×0,1, and T is total time.

The analytical solution is(53)ux,t=ex1+ex2+ex3et.

In this example, we discuss the effects of the weight functions, the scale parameter, the penalty factor, and the node distribution on the solution of the IEFG method, respectively.

4.1. Weight Functions

The time step Δt is selected as 0.01, 11×11×11 nodes are distributed regularly, and 10×10×10 integral cells are selected. Then the accuracy of the IEFG method is discussed.

When the cubic spline function is used, dmax=1.44, and α=8.7×106, then the smaller relative errors of the IEFG method are 1.2654%, 2.5893%, 2.7382%, 2.7478%, and 2.7483% at the times of 0.1 s, 0.3 s, 0.5 s, 0.7 s, and 0.9 s, respectively. When the quartic spline function is used, dmax=1.33, and α=1.1×107, the smaller relative errors of the IEFG method are 1.3092%, 2.7716%, 2.9469%, 2.9591%, and 2.9599% at the times of 0.1 s, 0.3 s, 0.5 s, 0.7 s, and 0.9 s, respectively.

Thus, when we use the cubic spline function as the weight function, higher accuracy can be obtained.

4.2. Scale Parameter

Δt=0.01, 11×11×11 regular nodes and 10×10×10 integral cells are selected, respectively, α=8.7×106, and the cubic spline function is used as the weight function. Figure 1 shows the relative errors of the IEFG method with the increase of dmax when T is 0.1 s. It is shown that the numerical solution has greater computational precision when dmax is 1.44.

The relative error of the IEFG method for different dmax.

4.3. Penalty Factor

The same time step, weight function, node distribution, and background integral grid are used, respectively; dmax=1.44. Figure 2 shows the relative errors of the IEFG method with the change of α when T is 0.1 s. It is shown that the numerical solution has greater computational precision when α is 5.0×1061.0×107.

The relative error of the IEFG method for different α.

4.4. Node Distribution

The same weight function and time step are used, respectively; dmax=1.44 and α=8.7×106. Figure 3 shows the relative errors of the IEFG method with the increase of nodes when T is 0.1 s. Therefore, the numerical solution of the IEFG method for 3D advection-diffusion problems is convergent.

The relative error of the IEFG method for different number of nodes.

4.5. Time Step

The same weight function is used; 11×11×11 regular nodes and 10×10×10 integral cells are selected, respectively; dmax=1.44 and α=8.7×106. Figure 4 shows the relative errors of the IEFG method with the change of Δt when T is 0.1 s. We can see that when Δt=0.01, not only is the greater precision obtained, but also the corresponding CPU time is saved.

The relative error of the IEFG method for different time step.

The EFG method is selected to solve this example, 11×11×11 regular nodes and 10×10×10 integral cells are selected, respectively, Δt=0.01, dmax=1.44, and α=8.7×106, and the cubic spline function is selected; then the great precision can be obtained. When the IEFG method is selected to solve it, the same parameters and the weight function are selected, respectively, then the great precision can also be obtained, and the relative errors are similar.

The comparison of the CPU time of the EFG and the IEFG methods under the same relative errors is shown in Table 1. And the numerical results are given to be compared with the analytical one when T are 0.1 s, 0.5 s, and 0.9 s, respectively (see Figures 57). Comparing with the EFG method, the IEFG method can obtain the solutions with similar computational accuracy, but it has greater computational efficiency.

Comparison of the CPU time of two methods under the same relative error.

ComparisonT
0.1 s0.3 s0.5 s0.7 s0.9 s
Relative error (%)1.26542.58932.73822.74782.7483
CPU time of EFG (s)101.6176250.5327.8395.8
CPU time of IEFG (s)97.2173.4240.5314.9386.3

The numerical results and the analytical solutions along x1-axis.

The numerical results and the analytical solutions along x2-axis.

The numerical results and the analytical solutions along x3-axis.

Moreover, the IEFG method which can avoid singular matrix is another advantage when constructing the shape functions. If dmax=1.0, when using the EFG method to solve it, the computational result cannot be obtained, and the error “Warning: Matrix is singular to working precision” appears in MATLAB code because of the singular matrix formed. But when the IEFG method is used, dmax=1.0, and other parameters are the same, then we can obtain the computational solution, and the relative error is 1.4187% when T is 0.1 s. The numerical solutions are compared with the analytical one which is shown in Figure 8; it is shown that the solutions of the IEFG method are in agreement with the analytic ones.

The numerical results and the analytical solutions along x1-axis when T is 0.1 s.

In the second example, a 3D advection-diffusion problem with source term is given as(54)ux,tt=2ux,tx12+2ux,tx22+2ux,tx32ux,tx1ux,tx2ux,tx3ex1+x2+x3t,xΩ,t0,T,with the initial condition(55)ux,0=ex1+x2+x3,and the boundary conditions(56)ux1=0=ex2+x3t,ux1=1=e1+x2+x3t,ux2=0=ex1+x3t,ux2=1=ex1+1+x3t,ux3=0=ex1+x2t,ux3=1=ex1+x2+1t.

The problem domain is Ω=0,1×0,1×0,1.

The analytical solution is(57)ux,t=ex1+x2+x3t.

The EFG method is used to solve this example, 11×11×11 regular nodes and 10×10×10 integral cells are selected, respectively, Δt=0.01, dmax=1.22, and α=8.4×107, and the cubic spline function is used; then the great precision can be obtained. When using the IEFG method to solve it, the same parameters and the weight function are selected, respectively, the great precision can also be obtained, and the relative errors of two methods are similar.

The comparison of the CPU time of the EFG and the IEFG methods under the same relative errors is shown in Table 2. And the numerical solutions are compared with the analytical one when T are 0.1 s, 0.5 s, and 0.9 s, respectively (see Figures 911). We can see again that the computational speed of the IEFG method is faster.

Comparison of the CPU time of two methods under the same relative error.

ComparisonT
0.1 s0.3 s0.5 s0.7 s0.9 s
Relative error (%)0.61560.80930.81090.81090.8109
CPU time of EFG (s)313.2816.71338.41822.82325.8
CPU time of IEFG (s)303.9785.31269.31763.82251.9

The numerical results and the analytical solutions along x1-axis.

The numerical results and the analytical solutions along x2-axis.

The numerical results and the analytical solutions along x3-axis.

Similarly, if dmax=1.0, when using the EFG method to solve it, the computational result cannot be obtained, and the same error appears in MATLAB code. But when the IEFG method is used, dmax=1.0, and other parameters are the same, then the relative error can be obtained which is 1.4096% when T is 0.1 s. The numerical solutions are compared with the analytical one which is shown in Figure 12; the solutions of the IEFG method are in agreement with the analytic ones which are shown.

The numerical results and the analytical solutions along x2-axis when T is 0.1 s.

In the third example, a 3D advection-diffusion equation is given as(58)ux,tt=k12ux,tx12+k22ux,tx22+k32ux,tx32v1ux,tx1v2ux,tx2v3ux,tx3,xΩ,t0,T,with the initial condition(59)ux,0=aec1x1+ec2x2+ec3x3,and the boundary conditions(60)u0,x2,x3,t=aebt1+ec2x2+ec3x3,u1,x2,x3,t=aebtec1+ec2x2+ec3x3,ux1,0,x3,t=aebtec1x1+1+ec3x3,ux1,1,x3,t=aebtec1x1+ec2+ec3x3,ux1,x2,0,t=aebtec1x1+ec2x2+1,ux1,x2,1,t=aebtec1x1+ec2x2+ec3.

The problem domain is Ω=0,1×0,1×0,1.

The analytical solution is(61)ux,t=aebtec1x1+ec2x2+ec3x3,where(62)c1=v1v12+4bk12k1,c2=v2v22+4bk22k2,c3=v3v32+4bk32k3.

We set k1=1.4, k2=1.7, k3=1.5, v1=v2=v3=1, and a=1, b=0.1, for simplicity. The EFG method is used to solve this example, 11×11×11 regular nodes and 10×10×10 integral cells are selected, respectively, Δt=0.01, dmax=2.15, and α=1.1×104, and the cubic spline function is selected; then the great precision can be obtained. When using the IEFG method to solve it, the same parameters and the weight function are selected, respectively, the numerical results can also be obtained with great precision, and the relative errors of two methods are similar.

The comparison of the CPU time of the EFG and the IEFG methods under the same relative errors is shown in Table 3. And the numerical solutions are compared with the analytical one when T are 0.1 s, 0.5 s, and 0.9 s, respectively (see Figures 1315).

Comparison of the CPU time of two methods under the same relative error.

ComparisonT
0.1 s0.3 s0.5 s0.7 s0.9 s
Relative error (%)0.12200.14020.14020.14020.1402
CPU time of EFG (s)366.2588.2781.4987.11184.2
CPU time of IEFG (s)336.1517.3702.4888.61064

The numerical results and the analytical solutions along x1-axis.

The numerical results and the analytical solutions along x2-axis.

The numerical results and the analytical solutions along x3-axis.

Similarly, if dmax=1.0, when using the EFG method to solve it, the computational result cannot be obtained, and the same error appears in MATLAB code. But when the IEFG method is used, dmax=1.0, and other parameters are the same, then the relative error can be obtained which is 0.1358% when T is 0.1 s. The numerical solutions are compared with the analytical one which is shown in Figure 16. Again, the solutions of the IEFG method are in agreement with the analytic ones which are shown.

The numerical results and the analytical solutions along x3-axis when T is 0.1 s.

For this example, we can select different parameters, k1=1.4, k2=1.7, k3=1.5, v1=v2=v3=5, and a=1, b=1. Using the EFG method to solve it, 11×11×11 regular nodes and 10×10×10 integral cells are selected, respectively, Δt=0.01, dmax=1.39, and α=4.7×103, and cubic spline function is selected; then the great precision can be obtained. When the IEFG method is selected to solve it, the same parameters and the weight function are selected, respectively, then the great precision can also be obtained, and the relative errors of two methods are similar.

The comparison of the CPU time of the EFG and the IEFG methods under the same relative errors is shown in Table 4. And the numerical solutions are compared with the analytical one when T are 0.1 s, 0.5 s, and 0.9 s, respectively (see Figures 1719).

Comparison of the CPU time of two methods under the same relative error.

ComparisonT
0.1 s0.3 s0.5 s0.7 s0.9 s
Relative error (%)0.58860.63300.63300.63300.6330
CPU time of EFG (s)100.4175.1249.1324.7398
CPU time of IEFG (s)98170246.6316393

The numerical results and the analytical solutions along x1-axis.

The numerical results and the analytical solutions along x2-axis.

The numerical results and the analytical solutions along x3-axis.

Similarly, if dmax=1.0, when using the EFG method to solve it, the computational result cannot be obtained, and the same error appears in MATLAB code. But when the IEFG method is used, dmax=1.0, and other parameters are the same, then the relative error can be obtained which is 0.6367% when T is 0.3 s. The numerical solutions are compared with the analytical one which is shown in Figure 20. The solutions of the IEFG method are in agreement with the analytic ones which are shown.

The numerical results and the analytical solutions along x3-axis when T is 0.3 s.

From this example, we can see that both diffusion dominated case and advection dominated one can be solved by using the EFG and the IEFG methods. Although two methods can obtain the solutions with similar computational accuracy, the computational speed of the IEFG is faster than the EFG method.

5. Conclusions

On the basis of the IMLS approximation, the IEFG method for 3D advection-diffusion problems is proposed in this paper.

In Section 4, the influences of the weight functions, the scale parameter, the penalty factor, the node distribution, and the time step on the computational precision of the solutions of the IEFG method are discussed, respectively. We can see that the IEFG method in this paper is convergent.

From three examples, we can conclude that, compared with the EFG method for 3D advection-diffusion problems, the IEFG method has a higher computational efficiency.

The numerical solutions show that the IEFG method can avoid singular matrix when constructing the shape functions; then the deficiency of the EFG method is overcome.

Data Availability

The data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest

The authors declare that there are no conflicts of interest regarding the publication of this paper.

Acknowledgments

This work was supported by the National Natural Science Foundation of China (no. 11571223) and the Science and Technology Innovation Project of Shanxi Colleges and Universities (Grant no. 2020L0344).

HuangR.ZhengS. J.LiuZ. S.NgT. Y.Recent advances of the constitutive models of smart materials- hydrogels and shape memory polymersInternational Journal of Applied Mechanics2020122205001410.1142/s1758825120500143ChengJ.Analyzing the factors influencing the choice of the government on leasing different types of land uses: evidence from Shanghai of ChinaLand Use Policy20209010430310.1016/j.landusepol.2019.104303ChengJ.Data analysis of the factors influencing the industrial land leasing in Shanghai based on mathematical modelsMathematical Problems in Engineering2020202011934686310.1155/2020/9346863ChengY. M.ChenM. J.A boundary element-free method for linear elasticityActa Mechanica Sinica2003352181186ZhangZ.WangJ. F.ChengY. M.LiewK. M.The improved element-free Galerkin method for three-dimensional transient heat conduction problemsScience China Physics, Mechanics and Astronomy20135681568158010.1007/s11433-013-5135-02-s2.0-84880303003ZhangZ.LiD. M.ChengY. M.LiewK. M.The improved element-free Galerkin method for three-dimensional wave equationActa Mechanica Sinica201228380881810.1007/s10409-012-0083-x2-s2.0-84865450119ZhangZ.LiewK. M.ChengY.LeeY. Y.Analyzing 2D fracture problems with the improved element-free Galerkin methodEngineering Analysis with Boundary Elements200832324125010.1016/j.enganabound.2007.08.0122-s2.0-38349104986YuS. Y.PengM. J.ChengH.ChengY. M.The improved element-free Galerkin method for three-dimensional elastoplasticity problemsEngineering Analysis with Boundary Elements201910421522410.1016/j.enganabound.2019.03.0402-s2.0-85063876398PengM. J.LiR. X.ChengY. M.Analyzing three-dimensional viscoelasticity problems via the improved element-free Galerkin (IEFG) methodEngineering Analysis with Boundary Elements20144010411310.1016/j.enganabound.2013.11.0182-s2.0-84891538167ChengH.PengM. J.ChengY. M.Analyzing wave propagation problems with the improved complex variable element-free Galerkin methodEngineering Analysis with Boundary Elements2019100808710.1016/j.enganabound.2018.02.0012-s2.0-85042564712WangB. H.MaY. Q.ChengY. M.The improved complex variable element-free Galerkin method for bending problem of thin plate on elastic foundationsInternational Journal of Applied Mechanics20191110195010510.1142/s1758825119501059LancasterP.SalkauskasK.Surfaces generated by moving least squares methodsMathematics of Computation19813715514115810.1090/s0025-5718-1981-0616367-12-s2.0-84966237423KaljevicI.SaigalS.An improved element free Galerkin formulationInternational Journal for Numerical Methods in Engineering199740162953297410.1002/(sici)1097-0207(19970830)40:16<2953::aid-nme201>3.0.co;2-sRenH. P.ChengY. M.ZhangW.Researches on the improved interpolating moving least-squares methodChinese Journal of Engineering Mathematics201027610211029RenH. P.ChengY. M.The interpolating element-free Galerkin (IEFG) method for two-dimensional potential problemsEngineering Analysis with Boundary Elements201236587388010.1016/j.enganabound.2011.09.0142-s2.0-84856211663LiuD.ChengY. M.The interpolating element-free Galerkin (IEFG) method for three-dimensional potential problemsEngineering Analysis with Boundary Elements201910811512310.1016/j.enganabound.2019.08.0212-s2.0-85071749916ZhaoN.RenH. P.The interpolating element-free Galerkin method for 2D transient heat conduction problemsMathematical Problems in Engineering20142014971283410.1155/2014/7128342-s2.0-84896923026RenH. P.ChengY. M.The interpolating element-free Galerkin (IEFG) method for two-dimensional elasticity problemsInternational Journal of Applied Mechanics20113473575810.1142/s17588251110012142-s2.0-84855952330ZhangP. X.PengM. J.Interpolating element-free Galerkin method for viscoelasticity problemsActa Physica Sinica2019681717020310.7498/aps.68.20191047WuQ.LiuF. B.ChengY. M.The interpolating element-free Galerkin method for three-dimensional elastoplasticity problemsEngineering Analysis with Boundary Elements202011515616710.1016/j.enganabound.2020.03.009WuQ.PengP. P.ChengY. M.The interpolating element-free Galerkin method for elastic large deformation problemsScience China Technological Sciences202063WangJ. F.SunF. X.ChengY. M.An improved interpolating element-free Galerkin method with nonsingular weight function for two-dimensional potential problemsChinese Physics B201221909020410.1088/1674-1056/21/9/0902042-s2.0-84866428955SunF. X.WangJ. F.ChengY. M.An improved interpolating element-free Galerkin method for elasticityChinese Physics B2013221212020310.1088/1674-1056/22/12/1202032-s2.0-84891600490LiuF. B.ChengY. M.The improved element-free Galerkin method based on the nonsingular weight functions for elastic large deformation problemsInternational Journal of Computational Materials Science and Engineering20187185002310.1142/S20476841185002392-s2.0-85062577918LiuF. B.WuQ.ChengY. M.A meshless method based on the nonsingular weight functions for elastoplastic large deformation problemsInternational Journal of Applied Mechanics2019111195000610.1142/S17588251195000662-s2.0-85061980556LiuF. B.ChengY. M.The improved element-free Galerkin method based on the nonsingular weight functions for inhomogeneous swelling of polymer gelsInternational Journal of Applied Mechanics2018104185004710.1142/s17588251185004732-s2.0-85048313513ChengH.PengM. J.ChengY. M.A hybrid improved complex variable element-free Galerkin method for three-dimensional potential problemsEngineering Analysis with Boundary Elements201784526210.1016/j.enganabound.2017.08.0012-s2.0-85027985412ChengH.PengM. J.ChengY. M.A fast complex variable element-free Galerkin method for three-dimensional wave propagation problemsInternational Journal of Applied Mechanics201796175009010.1142/S17588251175009092-s2.0-85029718650ChengH.PengM. J.ChengY. M.The dimension splitting and improved complex variable element-free Galerkin method for 3-dimensional transient heat conduction problemsInternational Journal for Numerical Methods in Engineering2018114332134510.1002/nme.57452-s2.0-85041178215ChengH.PengM. J.ChengY. M.A hybrid improved complex variable element-free Galerkin method for three-dimensional advection-diffusion problemsEngineering Analysis with Boundary Elements201897395410.1016/j.enganabound.2018.09.0072-s2.0-85054197411ChengH.PengM. J.ChengY. M.MengZ. J.A hybrid complex variable element-free Galerkin method for 3D elasticity problemsEngineering Structures2020219110835MengZ. J.ChengH.MaL. D.ChengY. M.The dimension split element-free Galerkin method for three-dimensional potential problemsActa Mechanica Sinica201834346247410.1007/s10409-017-0747-72-s2.0-85045053833MengZ. J.ChengH.MaL. D.ChengY. M.The dimension splitting element-free Galerkin method for 3D transient heat conduction problemsScience China Physics, Mechanics & Astronomy201962404071110.1007/s11433-018-9299-82-s2.0-85056886075MengZ. J.ChengH.MaL. D.ChengY. M.The hybrid element-free Galerkin method for three-dimensional wave propagation problemsInternational Journal for Numerical Methods in Engineering20191171153710.1002/nme.59442-s2.0-85053385732MaL. D.MengZ. J.ChaiJ. F.ChengY. M.Analyzing 3D advection-diffusion problems by using the dimension splitting element-free Galerkin methodEngineering Analysis with Boundary Elements202011116717710.1016/j.enganabound.2019.11.005PengP. P.WuQ.ChengY. M.The dimension splitting reproducing kernel particle method for three-dimensional potential problemsInternational Journal for Numerical Methods in Engineering2020121114616410.1002/nme.62032-s2.0-85071943445