2012
DOI: 10.1002/fld.3710
|View full text |Cite
|
Sign up to set email alerts
|

Numerical approximation of parabolic problems by residual distribution schemes

Abstract: SUMMARYWe are interested in the numerical approximation of steady scalar convection–diffusion problems by means of high order schemes called Residual Distribution schemes. In the inviscid case, one can develop nonlinear Residual Distribution schemes that are nonoscillatory, even in the case of very strong discontinuities, while having the most possible compact stencil, on hybrid unstructured meshes. This paper proposes and compare extensions of these schemes for the convection–diffusion problem. This methodolo… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

1
9
1

Year Published

2014
2014
2017
2017

Publication Types

Select...
5
2

Relationship

3
4

Authors

Journals

citations
Cited by 9 publications
(11 citation statements)
references
References 26 publications
1
9
1
Order By: Relevance
“…The objective is to show that (i) the high order RD schemes previously proposed can be successfully used in the discretization of the advection-diffusion equation, (ii) the high order accuracy is preserved in all the range of the Peclet number. This is contrast with the method proposed in [5] where the region P e ≈ 1 was problematic.…”
Section: Strategy (A)contrasting
confidence: 45%
See 1 more Smart Citation
“…The objective is to show that (i) the high order RD schemes previously proposed can be successfully used in the discretization of the advection-diffusion equation, (ii) the high order accuracy is preserved in all the range of the Peclet number. This is contrast with the method proposed in [5] where the region P e ≈ 1 was problematic.…”
Section: Strategy (A)contrasting
confidence: 45%
“…A more general scheme consists in using distribution coefficients which are function of the local Peclet number in order to recover an isotropic scheme in the diffusion limit and an upwind scheme in the advection limit [19,11]. Another attempt in that direction is given by [5], the scheme give satisfactory results except in the region P e ≈ 1, which is typical of a boundary layer. Hence the present contribution can be viewed as an improvement over the previous references.…”
Section: Consistency and Accuracymentioning
confidence: 99%
“…On the other hand, this means that better solutions to extend RD to viscous flows have to be sought. This justifies the recent investment in genuinely residual based approximations of the viscous terms by H. Nishikawa [21,22], as well as by some of the authors of the present work [3,4,5]. While this new approach might allow a genuine improvement over the hybrid RD-Galerkin formulation.…”
Section: Discussionsupporting
confidence: 63%
“…Our results show that, while in the scalar case one can clearly show that this approach allows to recover a uniform second order convergence rate, for moderate/high Reynolds laminar flows its beneficial effects are much less pronounced even when looking at friction coefficients. This justifies the quest for new consistent approaches allowing to further reduce the discretization error in viscous flows [21,22,3,4,5].…”
mentioning
confidence: 99%
“…It has been shown [3] that a scheme resulting from the coupling of a RD scheme with a Galerkin discretization of the viscous term is still a residual method, but this is true only for P1 elements. Furthermore it is well know [26] that the scheme obtained as the sum of the RD scheme and the Galerkin scheme is second order accurate in the diffusion and advection limits, but it is only first order accurate when advection and diffusion are equally important.…”
Section: Extension To Viscous Termsmentioning
confidence: 99%