Abstract
In this work, we construct a theoretical framework to develop non Morley type nonconforming high-convergence elements for biharmonic problems. For each element domain, should be included in the space of shape functions. Besides the degrees of freedom of Morley elements, we add the integrals and first-order moments of the normal derivatives on edges. The choice of degrees of freedom and shape function space guarantees the possibility of improving the convergence order. As an application, we specifically construct a Morley type element on triangular meshes. Lastly, numerical experiments are carried out to verify the feasibility of the element.
MSC:
65N30
1. Introduction
In this work, we introduce a new non element method to handle biharmonic problems, which provides optimal convergence in norm. Biharmonic equations are closely related to practical physics problems. Nevertheless, the biharmonic problem still is challenging in mathematics. For a review of the biharmonic problem, see the literature [,].
When the conforming finite element method is employed for the elliptic problem, the error analysis only needs to consider the finite element space. Since biharmonic problems belong to fourth-order equations, typical conforming finite element methods require globally continuity. The practical application of conforming finite element methods is computationally difficult: either the local space dimension is quite large or the structure of the finite space is complicated. More precisely, conforming elements such as the Argyris element [] and Bell’s triangle [], the 16-dof Bogner–Fox–Schmit (BFS) element [] is computationally expensive due to the high-degree polynomials involved. For fourth-order problems, computationally simple but formally complex conforming elements contain Hsieh–Clough–Tocher (HCT) elements [], Zienkiewicz triangular elements [], FVS quadrilateral elements [] and so on.
To deal with the difficult problems caused by , the nonconforming finite element method is usually used. Following Lascaux and Lesaint [], two nonconforming finite elements that do not belong to were proposed. The first one is the famous Morley element [,]. For biharmonic problems, the Morley element has the least number of degrees of freedom on the triangular mesh. Simultaneously, this element is proved to have the optimal convergence order in the discrete and -seminorms. The second element, known as Fraeijs de Veubeke triangle [], selects averages as some degrees of freedom. The 12-dof Adini element [] is a commonly used network on a rectangular grid, which contains on the rectangle domain. On a rectangular grid, the incomplete biquadratic element [] is an analogy of the Morley element on a rectangular grid. Hu et al. [] developed Powell–Sabin type macro elements on two-dimensional and three-dimensional rectangles. For quadrilateral meshes, Veubeke [] proposed a novel biharmonic solution by splitting a convex quadrilateral into four non-overlapping triangles. Park and Sheen [] offered a quadrilateral Morley element to handle biharmonic problems. However, the appearance of discontinuities will lead to consistency errors, so to ensure the well-posedness and convergence of finite elements, some basic continuity of the element space is still necessary. In addition, the accuracy of nonconforming element methods is usually low. And achieving higher accuracy using nonconforming finite element methods remains difficult.
Subsequently, many papers [] conducted in-depth research on the superconvergence estimation for conforming finite elements. Nevertheless, the superconvergence of nonconforming finite element schemes is still at a relatively primitive stage. For biharmonic equations, Luo and Lin [] presented the asymptotic error expansion for the Adini element. Mao and Shi [] proved the superconvergence for the Morley element and incomplete biquadratic nonconforming element.
Using superconvergence methods, or by building an element where the shape function space includes or the internal penalty method, a second-order convergent numerical solution to the biharmonic equation can be obtained. Research on interior penalty methods can be found in []. Sun, Song, and Liu [] created a high nonconforming method, where the penalty parameter is estimated accurately. On simplicial triangulation, Gao et al. [] proposed a nonconforming element that converges at a rate of . They imposed interelement continuity on elements. Chen et al. [,] built a theoretical framework of nonconforming elements for the fourth-order elliptic problem and constructed a series of three-dimensional elements. They successfully constructed second-order convergent nonconforming elements for biharmonic equations. The virtual element method [,,] also proposed high-convergence elements for biharmonic problems, but they cannot express the finite element spaces explicitly.
This paper proposes a method to construct non nonconforming elements for biharmonic equations. We force the shape function space to include in order to get the approximation error. The Morley element method on triangular meshes we have mentioned above is first-order-convergent. We enrich the degrees of freedom of the Morley element. More precisely, the degrees of freedom are chosen as , where and P denote the outward unit normal vector to edge E and the vertex of E, respectively. The continuity requirement is relaxed, and we replace it with other constraints. The moments are required to be continuous along the edge E in order to obtain the consistency error.
This work is arranged as follows. We introduce the setting of the biharmonic problem, as well as some symbols and preparations in Section 2. In Section 3, theoretical error estimates are constructed for this problem. Then, in Section 4, we present our triangular Morley type element and prove the unisolvency. Section 5 presents two numerical experiments to verify the validity of the theory. Lastly, we draw some conclusions.
2. The Model Problem and Some Preliminaries
We first introduce some notations. Let be a bounded polygonal domain, and and denote the unit outer normal and tangential vectors on boundaries. The inner product on will be denoted by . For , denotes the Sobolev space, and the corresponding norm and seminorm are presented by and , respectively. The space is the closure in of . Specifically, the following important space is given
We denote by the boundary of . We discuss the biharmonic problem:
is a standard Laplace operator. The weak form of (1) is to search satisfying
where
Here, represents Poisson’s ratio with numerical range .
Let S be a polygonal domain. In addition to and ,
since
3. Convergence Theorems
Assume that is a family of regular convex polygonal subdivisions of . represents the set of edges. Let denote the finite element space with boundary conditions of (1) in some way.
Let and represent the norm and seminorm associated with partition:
For any integer , represents the set of polynomials with degree less than or equal to m defined on K. Define global interpolation operator , . If on K, , , by the Bramble–Hilbert lemma, one has
Here, C is independent of h.
On any edge E, , and denote the jump of one function, unit tangential and normal vectors, respectively. We begin the error analysis with the second Strang lemma.
For , invoking the definition of , one has
the consistency difference between and is
Based on the analysis of the above paragraphs, we are ready to give the following error estimate.
Theorem 1.
- (H1)
- , , ,
- (H2)
- , is continuous at the interior vertices and zero at the vertices on ,
- (H3)
- , ,
- (H4)
- , , ,
then,
Proof.
Denote by the -projection from to and is the -projection from to . Denote by the lowest-order -conforming interpolation of . We first consider the consistency difference . The first two terms in (11) are bounded as follows:
Abbreviating for each edge , we have
Since , , similarly, one has
Collecting the above estimates, here , we have
Since preserves functions in , (7) implies the following result
Next, we derive the error estimates in the - and -norm. We shall consider the duality problem
Since is a convex polygon, it holds (see [])
and we assume that
Remark 1.
For the regularity estimate (19) is always correct. For , there exists q that satisfies
The maximum angle in Ω determines the value of q. Moreover, there exists a satisfying , and thus holds.
Theorem 2.
Proof.
Let and denote by the lowest-order -conforming interpolation of . Setting in (18), satisfies
then, (19) implies that
Then, it yields that
First, we use the estimate (7) to obtain
From the above Theorem 1, we have
Together with (24), one has
Observing (11), we have
Recalling the proof of Theorem 1, we can bound the last two terms in the above equation
For , the interpolation serves to get
The first two terms are estimated as follows:
The estimate of the last term in (29) is:
Finally, applying the triangle inequality leads to
which completes the proof. □
Theorem 3.
Proof.
For convenience, the same notations as in the proof of Theorem 2 are adopted. Setting in (18), satisfies
and then (20) implies
Observing (11) again, we have
For the last two terms, arguments similar to the proof of Theorem 1 yield
For , we proceed as in the proof of Theorem 2 and obtain
where
and
4. A New Morley Type Element with High Convergence
In this section, we introduce our new nonconforming element and prove the unisolvency. Let K be a triangle where denote the vertices, designates the edge between and modulo 3, , and denotes the unit outward normal to , .
Definition 1.
The triangular finite element is defined as follows:
- K is a triangle.
- .
- The degrees of freedom are given by ,
where is taken such that and .
Theorem 4.
The degrees of freedom are unisolvent for .
Proof.
For , we have
since
We prove the unisolvency with the help of a reference triangle whose vertices are denoted by . Correspondingly, we denote the edges of the reference triangle by and is the unit normal to the edge . We only need to consider the unisolvency on the reference triangle since there exists a unique map F from the reference triangle to an arbitrary triangle and , where represents the Jacobian matrix of F. Define . Define a matrix
Precisely,
and here, for convenience, we ignore the positive and negative sign for calculus since it does not affect the final determinant. By a simple computation, we obtain
This completes the proof. □
5. Numerical Experiments
Now, we will use the proposed element method to calculate two examples. Here, we take . Set . We first divide the unit into uniform squares, and then connect the secondary diagonals of each square. Thus, two types of triangulation are used: uniform triangular meshes (see Figure 1 as an example) and randomly perturbed triangular meshes (see Figure 2 as an example).
Example 1.
Consider problem (1) with , where .

Figure 1.
A mesh with uniform triangular partition.

Figure 2.
A mesh with nonuniform randomly perturbed triangular partition.
The errors measured by the seminorms and norm are shown in Table 1 and Table 2, respectively. The optimal order of convergence in discrete , seminorms and norm is indicated in Theorems 1–3. Obviously, the convergence orders in norm and and seminorms mathematically approximate 4, 3, and 2, respectively. In short, from the following two tables, we can find that theoretical analysis and numerical results are consistent.
Example 2.
Assume that satisfies (1).

Table 1.
Error table for uniform triangular meshes applied to Example 1.

Table 2.
Error table for randomly perturbed triangular meshes applied to Example 1.
Similarly, the errors measured by the seminorms and norm are shown in Table 3 and Table 4, respectively. And the numerical experimental results are consistent with the theoretical analysis.

Table 3.
Error table for uniform triangular meshes applied to Example 2.

Table 4.
Error table for randomly perturbed triangular meshes applied to Example 2.
6. Conclusions
In this paper, an error analysis for the biharmonic problem is given. And the error analysis provides a theoretical basis for constructing nonconforming high-order Morley type elements for biharmonic problems. Subsequently, a Morley type triangular element is presented. Numerical experiments are carried out on two types of triangular meshes. Furthermore, based on this error analysis, we can also study the Morley type element on quadrilateral meshes in the future.
Author Contributions
Methodology, Y.B.; validation, Y.B.; formal analysis, Y.B.; writing—review and editing, Y.B. and S.Y.; visualization, S.Y.; supervision, Y.B. and S.Y.; project administration, Y.B. and S.Y. All authors have read and agreed to the published version of the manuscript.
Funding
This research received no external funding.
Data Availability Statement
Data are contained within the article.
Conflicts of Interest
The authors declare no conflicts of interest.
References
- Meleshko, V. Selected topics in the history of the two-dimensional biharmonic problem. Appl. Mech. Rev. 2003, 56, 33–85. [Google Scholar] [CrossRef]
- Ushakov, A.L. Analysis of the problem for the biharmonic equation. J. Comput. Eng. Math. 2022, 9, 43–58. [Google Scholar] [CrossRef]
- Argyris, J.; Balmer, H.; Doltsinis, J.S.; Dunne, P.; Haase, M.; Kleiber, M.; Malejannakis, G.; Mlejnek, H.P.; Müller, M.; Scharpf, D. Finite element method—The natural approach. Comput. Methods Appl. Mech. Eng. 1979, 17, 1–106. [Google Scholar] [CrossRef]
- Bell, K. A refined triangular plate bending finite element. Int. J. Numer. Methods Eng. 1969, 1, 101–122. [Google Scholar] [CrossRef]
- Bogner, F. The generation of interelement compatible stiffness and mass matrices by the use of interpolation formulae. In Proceedings of the Conference on Matrix Methods in Structural Mechanics, Wright Patterson Air Force Base, OH, USA, 26–28 October 1965. [Google Scholar]
- Clough, R. Finite element stiffness matrices for analysis of plates in bending. In Proceedings of the 1st Conference on Matrix Methods in Structural Mechanics, Wright-Patterson Air Force Base, OH, USA, 26–28 October 1965. [Google Scholar]
- Zienkiewicz, O.; Taylor, R.; Too, J. Reduced integration technique in general analysis of plates and shells. Int. J. Numer. Methods Eng. 1971, 3, 275–290. [Google Scholar] [CrossRef]
- De Veubeke, B.F. A conforming finite element for plate bending. Int. J. Solids Struct. 1968, 4, 95–108. [Google Scholar] [CrossRef]
- Lascaux, P.; Lesaint, P. Some nonconforming finite elements for the plate bending problem. Rev. Fr. d’Autom. Inform. Rech. Opér. Anal. Numér. 1975, 9, 9–53. [Google Scholar] [CrossRef]
- Morley, L. The triangular equilibrium element in the solution of plate bending problems. Aeronaut. Q. 1968, 19, 149–169. [Google Scholar] [CrossRef]
- Adak, D.; Mora, D.; Silgado, A. A Morley-type virtual element approximation for a wind-driven ocean circulation model on polygonal meshes. J. Comput. Appl. Math. 2023, 425, 115026. [Google Scholar] [CrossRef]
- De Veubeke, B.F. Variational principles and the patch test. Int. J. Numer. Methods Eng. 1974, 8, 783–801. [Google Scholar] [CrossRef]
- Adini, A.; Clough, R.W. Analysis of Plate Bending by the Finite Element Method; University of California: Berkeley, CA, USA, 1960. [Google Scholar]
- Shi, Z. On the convergence of the incomplete biquadratic nonconforming plate element. Math. Numer. Sin. 1986, 8, 53–62. [Google Scholar]
- Hu, J.; Huang, Y.; Zhang, S. The lowest order differentiable finite element on rectangular grids. SIAM J. Numer. Anal. 2011, 49, 1350–1368. [Google Scholar] [CrossRef]
- Park, C.; Sheen, D. A quadrilateral Morley element for biharmonic equations. Numer. Math. 2013, 124, 395–413. [Google Scholar] [CrossRef]
- Shi, D.; Yang, H. Superconvergent estimates of conforming finite element method for nonlinear time-dependent Joule heating equations. Numer. Methods Partial. Differ. Equ. 2018, 34, 336–356. [Google Scholar] [CrossRef]
- Luo, P.; Lin, Q. Accuracy analysis of the Adini element for biharmonic equation. Acta Math. Sin. 2004, 20, 135–146. [Google Scholar] [CrossRef]
- Mao, S.; Shi, Z.c. High accuracy analysis of two nonconforming plate elements. Numer. Math. 2009, 111, 407–443. [Google Scholar] [CrossRef]
- Zhao, J.; Mao, S.; Zhang, B.; Wang, F. The interior penalty virtual element method for the biharmonic problem. Math. Comput. 2023, 92, 1543–1574. [Google Scholar] [CrossRef]
- Sun, M.; Song, S.; Liu, Z. A nonconforming scheme with high accuracy for the plate bending problem. Comput. Math. Appl. 2014, 68, 1083–1092. [Google Scholar] [CrossRef]
- Gao, B.; Zhang, S.; Wang, M. A note on the nonconforming finite elements for elliptic problems. J. Comput. Math. 2011, 29, 215–226. [Google Scholar]
- Chen, H.; Chen, S.; Qiao, Z. C0-nonconforming tetrahedral and cuboid elements for the three-dimensional fourth order elliptic problem. Numer. Math. 2013, 124, 99–119. [Google Scholar] [CrossRef]
- Chen, H.; Chen, S.; Qiao, Z.H. C0-nonconforming triangular prism elements for the three-dimensional fourth order elliptic problem. J. Sci. Comput. 2013, 55, 645–658. [Google Scholar] [CrossRef]
- Zhao, J.; Chen, S.; Zhang, B. The nonconforming virtual element method for plate bending problems. Math. Models Methods Appl. Sci. 2016, 26, 1671–1687. [Google Scholar] [CrossRef]
- Antonietti, P.F.; Manzini, G.; Verani, M. The fully nonconforming virtual element method for biharmonic problems. Math. Models Methods Appl. Sci. 2018, 28, 387–407. [Google Scholar] [CrossRef]
- Carstensen, C.; Khot, R.; Pani, A.K. Nonconforming virtual elements for the biharmonic equation with Morley degrees of freedom on polygonal meshes. SIAM J. Numer. Anal. 2023, 61, 2460–2484. [Google Scholar] [CrossRef]
- Grisvard, P. Singularities in Boundary Value Problems; Recherches en Mathématiques Appliquées [Research in Applied Mathematics]; Masson: Paris, France; Springer: Berlin/Heidelberg, Germany, 1992; Volume 22, pp. xiv+199. [Google Scholar]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2024 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).