Available online at www.sciencedirect.

com

ScienceDirect
Procedia Engineering 111 (2015) 8 – 13

XXIV R-S-P seminar, Theoretical Foundation of Civil Engineering (24RSP) (TFoCE 2015)

On the use of discrete-continual finite element method with

unstructured meshes
Pavel A. Akimova,b,c *, Oleg A. Negrozova
a
Moscow State University of Civil Engineering; 26, Yaroslavskoe Shosse, Moscow, 129337, Russia
b
Russian Academy of Architecture and Construction Sciences; 24-1, Ulitsa Bolshaya Dmitrovka, Moscow, 107031, Russia
c
Scientific Research Centre “StaDyO”, 18, 3-ya Ulitsa Yamskogo Polya, Moscow, 125040, Russia

Abstract

The field of application of discrete-continual finite element method comprises structures with regular (in particular, constant or
piecewise constant) physical and geometrical parameters in some dimension (so-called “basic” direction (dimension)).
Considering problems remain continual along “basic” direction while along other directions mesh approximation is performed.
Algorithm of computing of local stiffness matrix of discrete-continual finite element with quadrangular cross-section and
different integration schemes is presented. Constructing of global stiffness matrix with the use of adjacency matrix and several
numerical samples are considered as well.
© 2015 The Authors. Published by Elsevier B.V.
© 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
Peer-review under responsibility of organizing committee of the XXIV R-S-P seminar, Theoretical Foundation of Civil
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Engineering
Peer-review (24RSP)
under responsibility of organizing committee of the XXIV R-S-P seminar, Theoretical Foundation of Civil Engineering (24RSP)

Keywords: Discrete-continual finite element method; Structural analysis; Unstructured mesh; Stiffness matrix; Adjacency matrix; Integration
scheme; Discrete-continual finite element

1. Introduction

Development, research and verification of correct mathematical models and methods of structural mechanics are
the most important aspects of ensuring safety of buildings and complexes. The analysis and design of structures
normally require accurate computer-intensive calculations using numerical (discrete) methods [1-8]. The field of
application of semianalytical (discrete-continual) methods, that are now becoming available for computer

* Corresponding author. Tel.: +7-499-183-5994; fax: +7-499-183-5994.

E-mail address: akimov@raasn.ru

1877-7058 © 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of organizing committee of the XXIV R-S-P seminar, Theoretical Foundation of Civil Engineering (24RSP)
doi:10.1016/j.proeng.2015.07.028
 Pavel A. Akimov and Oleg A. Negrozov / Procedia Engineering 111 (2015) 8 – 13 9

realization, comprises structures with regular (in particular, constant or piecewise constant) physical and geometrical
parameters in some dimension (so-called “basic” direction (dimension)). Considering problems remain continual
along “basic” direction while along other directions discrete-continual methods presuppose mesh approximation.
Thus, we have so-called discrete-continual finite element method (DCFEM), proposed by P.A. Akimov and A.B.
Zolotov [9-13].
Piecewise constancy of physical and geometrical parameters in one dimension exists in various problems of
analysis of 4dstructures and corresponding mathematical models. We should mention here in particular such vital
objects as beams, strip foundations, thin-walled bars, deep beams, plates, shells, high-rise buildings, extensional
buildings, pipelines, rails, dams and others. Analytical solution is apparently preferable in all aspects for qualitative
analysis of calculation data. It allows investigator to consider boundary effects when some components of solution
are rapidly varying functions. Due to the abrupt decrease inside of mesh elements in many cases their rate of change
can’t be adequately considered by conventional numerical methods while analytics enables study. Another feature of
the proposing method is the absence of limitations on lengths of structures. Hence it appears that in this context
discrete-continual finite element method considering in this paper is peculiarly relevant. Generally, semianalytical
formulation are contemporary mathematical models which currently becoming available for computer realization.
Meshes topologically equivalent to rectangular [14] have been used mainly in previous versions of DCFEM for
approximation of cross-sections of structures. Nevertheless this limitation can be removed after implementation of
unstructured mesh (for instance without void discrete-continual finite elements in structures with embrasures) [15].
This realization of DCFEM is considered in the distinctive paper.

2. General Approach to Numbering of Nodes and Mesh Approximation

Arbitrary numbering of nodes can introduced within DCFEM. Let’s each discrete-continual finite element has list
of four nodes. Corresponding data about such mesh is stored as in matrix of 4 u N e size ( N e is the number of
elements) with lists of nodes for each element.
Besides, modification of some formulas of approximation is required (two indexes of nodes were used in
numbering algorithms in previous versions of DCFEM). In order to simplify these formulas transformation of cross-
section of discrete-continual finite element to square (Fig. 1).

t2
1
1
2

-1 1 t1

4
3
-1

Fig. 1. Transformation of cross-section of discrete-continual finite element.

Thus, we have t1  [1, 1] , t 2  [1, 1] , where t1 and t2 are corresponding local coordinates.
Transformation of coordinates take place according to the formula:

x 0.25 ˜ [(1  t1 )(1  t2 ) x1  (1  t1 )(1  t2 ) x 2  (1  t1 )(1  t2 ) x 3  (1  t1 )(1  t2 ) x 4 ] (1)

where xk is global coordinate vector of k-th node (nodal line) of considering discrete-continual finite element.
Let M be arbitrary function.
10 Pavel A. Akimov and Oleg A. Negrozov / Procedia Engineering 111 (2015) 8 – 13

Therefore we have the following differentiation formulas:

wM 2
wM wt s
wxk
¦ wt
s 1 s wxk
, s 1, 2 (2)

Derivatives of the local variables can be obtained from Jacobian of this transformation. Jacoby matrix has the
form:

Dt ª wt 1 wx1 wt 1 wx 2 º ªD 11 D 12 º
J ( x1 , x 2 ) «¬wt 2 wx1 (3)
Dx wt 2 wx 2 »¼ «¬D 21 D 22 »¼

Jacobian can be calculated after inverse of matrix E Dx Dt ,

Dx ª wx1 wt 1 wx1 wt 2 º ª E 11 E 12 º
«¬wx 2 wt 1 (4)
Dt wx 2 wt 2 »¼ «¬ E 21 E 22 »¼

Derivative of global variables should be obtained by formula:

wx 1
[(x 2  x1  x4  x 3 )t3  s  (1) s ( x1  x 3 )  x 2  x4 ] (5)
wts 4

3. Integration scheme.

Integration of corresponding functions is required for construction of local stiffness matrices of discrete-continual
finite element. Gauss method can be used in particular. Points of integration and their coordinates ( [ 1 / 3 ) are
showed on Fig. 2, weight of each point is equal to one. The following integration formula can be used:

³³
S
f ( x, y )dS ¦ det( J (x , y )) ˜ f ( x , y )
k 1
k k k k (6)

t2
1

ȟ
ȟ
-1 -ȟ 1 t1
-ȟ

-1

Fig. 2. Illustration of the Gauss method. Points of integration.

 Pavel A. Akimov and Oleg A. Negrozov / Procedia Engineering 111 (2015) 8 – 13 11

4. Program Implementation and Verification

Version of DCFEM, considering in the distinctive paper, has been realized in software DCFEM3D. Programming
environment is Microsoft Visual Studio 2013 and Intel Parallel Studio XE 2015 (Intel Visual Fortran Composer XE
2015) [16-19]. Mathematical library Intel MKL is used for solving of eigenvalue problem and linear equations.
Let’s consider three-dimensional beam with rectangular cross-section hinge-supported on two sides ( x3 0 and
x3 L with zero displacements u1 u2 u3 0 ). Length of structure ( L ) is equal to 300 cm. Width of structure
( h1 ) is equal to 50 cm. Height of structure ( h2 ) is equal to 50 cm. Elastic module ( E ) of material of structure is
equal to 3000 kN/cm2. Poisson's ratio is equal to 0.16. Value of the load is equal to 100 kN, dimension a is equal to
150 cm (Fig. 3).

Fig. 3. Design model of structure.

Abnormal numbering of nodes is used for verification of unstructured mesh (Fig. 4). Force is applied at node 7.
Graphical result (deformation and stress) will be showed through node 45. Corresponding nodes are shown on Fig.
4.

Fig. 4. Approximating Mesh and Loading Condition.

12 Pavel A. Akimov and Oleg A. Negrozov / Procedia Engineering 111 (2015) 8 – 13

Several comparisons of stresses and displacements, obtained by DCFEM3D and ANSYS Mechanical 15.0 (finite
element method) [20-24] for considering section are presented at fig. 5 and 6. Corresponding results agree well.

a) b)

Fig. 5. Comparison of results: a) Displacement u y ; b) normal stress V yy .

a) b)

Fig. 6. Comparison of results: a) normal stress V yy ; b) tangential stress V yz .

Acknowledgements

This work was financially supported by the Grants of the Russian Academy of Architecture and Construction
Sciences (7.1.7, 7.1.8).
 Pavel A. Akimov and Oleg A. Negrozov / Procedia Engineering 111 (2015) 8 – 13 13

References

[1] Bathe K J, Finite Element Procedures, Prentice Hall, 2007, 1050 pages.
[2] Fenner R.T. Finite Element Methods for Engineers, Imperial College Press, 2013, 272 pages.
[3] Fish J., Belytschko T. A First Course in Finite Elements, John Wiley & Sons Ltd, 2007, 344 pages.
[4] Godbole P.N. Introduction to Finite Element Methods, I K International Publishing House, 2013, 400 pages.
[5] Kaveh A. Computational Structural Analysis and Finite Element Methods, Springer, 2013, 432 pages.
[6] Li K., Huang A., Huai Q. Finite Element Method and Its' Applications, Alpha Science Intl Ltd, 2015, 400 pages.
[7] Zienkiewicz O.C., Taylor R.L., Fox D.D. The Finite Element Method for Solid and Structural Mechanics, Butterworth-Heinemann, 2013,
604 pages.
[8] Zienkiewicz O.C., Taylor R.L., Zhu J.Z. The Finite Element Method: Its Basis and Fundamentals, Butterworth-Heinemann, 2013, 756 pages.
[9] Akimov P.A. Correct Discrete-Continual Finite Element Method of Structural Analysis Based on Precise Analytical Solutions of Resulting
Multipoint Boundary Problems for Systems of Ordinary Differential Equations. // Applied Mechanics and Materials Vols. 204-208, 2012, pp.
4502-4505.
[10] Akimov P.A., Belostosky A.M., Mozgaleva M.L., Mojtaba Aslami, Negrozov O.A. Correct Multilevel Discrete-Continual Finite Element
Method of Structural Analysis. // Advanced Materials Research, Vol. 1040, 2014, pp. 664-669.
[11] Akimov P.A., Sidorov V.N. Correct Discrete-continual Finite Element Method for Analysis of Structures with Piecewise Constant Physical
and Geometrical Parameters in One Direction. // Abstract Volume, 14th International Conference on Computing in Civil and Building
Engineering (14th ICCCBE), Moscow, 27-29 June 2012, Moscow State University of Civil Engineering (National Research University),
Publishing House "ASV", pp. 152-153.
[12] Akimov P.A., Mozgaleva M.L., Negrozov O.A. About Verification of Discrete-Continual Finite Element Method for Two-Dimensional
Problems of Structural Analysis. Part 1: Deep Beam with Constant Physical and Geometrical Parameters Along Basic Direction. Advanced
Materials Research, Vols. 1025-1026, 2014, pp. 89-94.
[13] Akimov P.A., Mozgaleva M.L., Mojtaba Aslami, Negrozov O.A. About Verification of Discrete-Continual Finite Element Method for Two-
Dimensional Problems of Structural Analysis. Part 2: Deep Beam with Piecewise Constant Physical and Geometrical Parameters Along
Basic Direction. // Advanced Materials Research, Vols. 1025-1026, 2014, pp. 95-103.
[14] Akimov P.A., Mozgaleva M.L. Method of Extended Domain and General Principles of Mesh Approximation for Boundary Problems of
Structural Analysis. // Applied Mechanics and Materials, Vols. 580-583, 2014, pp. 2898-2902.
[15] Lo D.S.H. Finite Element Mesh Generation, CRC Press, 2014, 672 pages.
[16] Adams J.C., Brainerd W.S., Hendrickson R.A., Maine R.E., Martin J.T., Smith B.T. The Fortran 2003 Handbook: The Complete Syntax,
Features and Procedures, Springer, 2009, 713 pages.
[17] Brainerd W.S. Guide to Fortran 2003 Programming, Springer, 2009, 357 pages.
[18] Hanson R.J., Hopkins T. Numerical Computing With Modern Fortran (Applied Mathematics), SIAM-Society for Industrial and Applied
Mathematics, 2013, 260 pages.
[19] Smith I.M., Griffiths D.V., Margetts L. Programming the Finite Element Method, Wiley, 2013, 682 pages.
[20] Alawadhi E.M. Finite Element Simulations Using ANSYS, CRC Press, 2009, 416 pages.
[21] Chen X., Liu Y. Finite Element Modeling and Simulation with ANSYS Workbench, CRC Press, 2014, 411 pages.
[22] Madenci E., Guven I. The Finite Element Method and Applications in Engineering Using ANSYS, Springer, 2007, 686 pages.
[23] Moaveni S. Finite Element Analysis Theory and Application with ANSYS, Prentice Hall, 2007, 880 pages.
[24] Stolarski T., Nakasone Y., Yoshimoto S. Engineering Analysis with ANSYS Software, Butterworth-Heinemann, 2007, 480 pages.