pith. sign in

arxiv: 1906.10812 · v1 · pith:QP666DYDnew · submitted 2019-06-26 · 🧮 math.NA · cs.NA

Polyhedral Spline Finite Elements

Pith reviewed 2026-05-25 15:43 UTC · model grok-4.3

classification 🧮 math.NA cs.NA
keywords spline finite elementspolyhedral partitionsB-splinesmollificationapproximation powercollocation methodadaptive computationisogeometric analysis
0
0 comments X

The pith

Spline finite elements on general polyhedral partitions achieve high smoothness and strong approximation by mollifying low-order splines with B-splines.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper introduces multivariate spline finite elements defined on arbitrary polyhedral partitions. These elements are obtained by mollifying splines of low smoothness order with B-splines, yielding both high smoothness and strong approximation power. The base low-order splines can be chosen as simple polynomials on each cell of the partition. The resulting elements are easy to evaluate, making them suitable for collocation methods and adaptive computations in the numerical solution of differential equations.

Core claim

The central claim is that mollifying splines of low smooth order with B-splines produces multivariate spline finite elements on general partitions that possess both high smooth order and strong approximation power. The low smooth order splines can be chosen flexibly, for example as polynomials on individual cells of general polyhedral partitions. Evaluation of such splines is not difficult, so the obtained spline elements are suitable for the collocation method and adaptive computation.

What carries the argument

Mollification of low-smooth-order splines by B-splines, which raises the smoothness order without degrading approximation power on arbitrary polyhedral partitions.

If this is right

  • Finite element discretizations can achieve high continuity on general polyhedral domains without requiring tensor-product structure.
  • Collocation methods become practical because the spline values and derivatives are straightforward to compute.
  • Adaptive mesh refinement on polyhedral partitions is supported by the flexible choice of base splines per cell.
  • The same construction applies when the base functions are chosen differently on each cell, preserving approximation strength.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The method could serve as a drop-in replacement for NURBS in isogeometric analysis when the geometry is polyhedral rather than smooth.
  • Extension to three-dimensional or time-dependent problems would follow directly from the same mollification step.
  • Implementation cost remains low because evaluation reduces to standard B-spline operations plus the base spline values.

Load-bearing premise

Mollifying low-order splines with B-splines increases the smoothness order without degrading the approximation power of the underlying splines on arbitrary polyhedral partitions.

What would settle it

A numerical experiment on a non-simply-connected polyhedral partition where the mollified spline fails to recover the full approximation order of the base low-order spline would disprove the claim.

read the original abstract

Spline functions have long been used in numerically solving differential equations. Recently it revives as isogeometric analysis, which uses NURBS for both parametrization and element functions. In this paper, we introduce some multivariate spline finite elements on general partitions. These splines are constructed by mollifying splines of low smooth order with B-splines. They have both high smooth order and strong approximation power. The low smooth order splines can be chosen flexibly, for example polynomials on individual cells of general polyhedral partitions. Evaluation of such splines is not difficult, so the obtained spline elements are suitable for the collocation method and adaptive computation.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 1 minor

Summary. The paper introduces multivariate spline finite elements on general polyhedral partitions constructed by mollifying low-smoothness splines (e.g., piecewise polynomials on individual cells) with B-splines. The resulting splines achieve arbitrarily high smoothness order while retaining strong approximation power, and are positioned as suitable for collocation methods and adaptive computations in the context of isogeometric analysis and finite element methods.

Significance. If the central construction preserves approximation order on completely general (including non-shape-regular) polyhedral partitions, the work would offer a flexible route to high-smoothness spline spaces without tensor-product restrictions, which is potentially significant for numerical solution of PDEs on complex geometries.

major comments (1)
  1. [Abstract, paragraph 2] Abstract, paragraph 2: the claim that mollification with B-splines yields both C^k smoothness for arbitrary k and the same approximation power as the underlying low-order space on arbitrary polyhedral partitions is load-bearing for the central contribution. No argument, estimate, or reference is supplied in the abstract showing that the fixed-support B-spline kernel commutes with polynomial reproduction or preserves the Jackson constant when cell diameters vary by orders of magnitude; this is precisely the point raised by the stress-test note and must be addressed with a concrete proof or counter-example in the body of the paper.
minor comments (1)
  1. The abstract states that 'evaluation of such splines is not difficult' but provides no indication of the algorithmic complexity or data structures required for the mollification step on general partitions.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the careful reading and constructive comment on our manuscript. We address the major comment below and will revise the paper accordingly.

read point-by-point responses
  1. Referee: [Abstract, paragraph 2] Abstract, paragraph 2: the claim that mollification with B-splines yields both C^k smoothness for arbitrary k and the same approximation power as the underlying low-order space on arbitrary polyhedral partitions is load-bearing for the central contribution. No argument, estimate, or reference is supplied in the abstract showing that the fixed-support B-spline kernel commutes with polynomial reproduction or preserves the Jackson constant when cell diameters vary by orders of magnitude; this is precisely the point raised by the stress-test note and must be addressed with a concrete proof or counter-example in the body of the paper.

    Authors: We agree that the abstract, as a high-level summary, supplies no argument, estimate, or reference for the approximation-power claim, and that this must be addressed with a concrete proof (or counter-example) in the body. The manuscript establishes C^k smoothness via the standard properties of B-spline mollification. We will revise the abstract to include a brief pointer to the supporting analysis. In the body we will add a dedicated subsection that proves the mollification operator commutes with polynomial reproduction (by verifying that the kernel integrates to one and has vanishing moments up to the reproduction degree) and supplies local Jackson-type estimates showing that the approximation order is retained on completely general polyhedral partitions, including those whose cell diameters vary by orders of magnitude; the key step is to let the support radius of each B-spline kernel scale with the local cell diameter. Should a counter-example appear during the revision, it will be reported explicitly. revision: yes

Circularity Check

0 steps flagged

No circularity; construction presented as explicit definition without reduction to inputs

full rationale

The paper introduces the spline elements via an explicit construction: mollifying low-smoothness splines (e.g., piecewise polynomials on polyhedral cells) with B-splines to raise smoothness order. No equations, self-citations, or uniqueness theorems are quoted that would make a claimed approximation property equivalent to the input space by definition or by a fitted parameter renamed as prediction. The abstract states the properties as consequences of the construction rather than deriving them from a self-referential loop. No load-bearing self-citation chain or ansatz smuggling appears in the provided text. This is the common case of a self-contained constructive definition.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review supplies no explicit free parameters, axioms, or invented entities; all details of the mollification operator and approximation proofs are absent.

pith-pipeline@v0.9.0 · 5619 in / 1031 out tokens · 24731 ms · 2026-05-25T15:43:35.604282+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

1 extracted references · 1 canonical work pages

  1. [1]

    Ren-Hong Wang, Multivariate Spline Functions and Their Applications, Series: Mathematics and its applications, Vol

    [1]. Ren-Hong Wang, Multivariate Spline Functions and Their Applications, Series: Mathematics and its applications, Vol. 529, Springer, 2001. [2]. Wolfgang A. Dahmen & Charles A. Micchelli, Computation of inner products of multivariate B-splines, Numerical Functional Analysis and Optimization, Volume 3, Issue 3, pp367—375, 1981. [3]. T.N.T. Goodman, Polyh...