Asymptotic justification of the Reynolds equation for a spherical bearing
Guy Bayada, José M. Rodríguez, Raquel Taboada-Vázquez · 2026
Read Asymptotic justification of the Reynolds equation for a spherical bearing by Guy Bayada, José M. Rodríguez, Raquel Taboada-Vázquez (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2606-30148.txt.
Introduction
Guy Bayada, José Rodríguez, and Raquel Taboada-Vázquez supply a missing proof in lubrication theory. The Reynolds equation models thin fluid films in bearings and joints. Reynolds originally assumed one confining surface was flat. For a spherical bearing, neither surface is flat at all. No rigorous justification existed for that curved case before their 2026 proof.
The authors study the Stokes problem between two closely spaced spheres. As the gap between the spheres shrinks toward zero, solutions converge. The limit is exactly the solution of the corresponding Reynolds equation. This is an asymptotic justification, not a formal derivation. Formal versions for general surfaces existed in earlier joint work. Rigorous ones covered flat or journal bearing geometries only. Their proof closes the spherical gap in that literature. Engineers gain firmer ground for spherical bearing design formulas.
Catalog record
| CREATOR | Guy Bayada, José M. Rodríguez, Raquel Taboada-Vázquez |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 65 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2606.30148
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Asymptotic justification of the Reynolds equation for a spherical bearing Guy Bayada, José M. Rodríguez, Raquel Taboada-Vázquez Abstract To our knowledge, there is no rigorous mathematical justification of the Reynolds equation for a spherical bearing. In this article, we demonstrate that the solution of the Stokes problem in a domain between two closely spaced spheres converges, as the distance between the spheres approaches zero, to the solution of a “Reynolds equation”. 1. INSA Lyon, Centrale Lyon, Université Lyon 1, Université Jean Monnet, CNRS, ICJ UMR5208, 69621 Villeurbanne, France. E-mail: [email protected]. 2. CITMAga, Santiago de Compostela, Spain. 3. Department of Mathematics, Higher Technical University College of Architecture, Universidade da Coruña, A Coruña, Spain. E-mail: [email protected]. 4. Department of Mathematics, School of Civil Engineering, Universidade da Coruña, A Coruña, Spain. E-mail: [email protected]. 1 Introduction In 1886, Reynolds introduced a model describing the behavior of a thin fluid layer in a lubrication problem (see [ 22 ] ). In deriving his model, Reynolds assumed that one of the two surfaces confining the fluid was flat. Although this assumption is locally acceptable, it may not be valid in many practical cases, leading several authors to propose alternative versions of the Reynolds equation, in which neither of the bounding surfaces is flat (see, for example, [ 3 ] , [ 6 ] , [ 9 ] , [ 15 ] , [ 18 ] , [ 19 ] , [ 20 ] , [ 21 ] , [ 26 ] ). In [ 23 ] - [ 24 ] , two of the authors of this paper presented a new lubrication model applicable to any two boundary surfaces, using the method of asymptotic expansions to derive it formally. In [ 7 ] , the authors of this paper apply the results of [ 23 ] - [ 24 ] to the case of cylindrical, spherical, and conical bearings, comparing the resulting models with others that can be found in the literature (for example, those already mentioned [ 6 ] , [ 18 ] , [ 19 ] , [ 20 ] , [ 21 ] ). Since most of the methods used in the literature to derive lubrication models are formal, several authors have attempted to justify these models more rigorously (for example, by proving that the solution of the Stokes problem converges, in a certain sense, to the solution of the lubrication model). Several justifications of the classical Reynolds equation can be found in [ 4 ] , [ 10 ] and [ 11 ] , while [ 12 ] establishes the validity of the compressible Reynolds equation. Additional convergence results are mentioned in [ 8 ] . However, rigorous justifications of the Reynolds equation when neither of the confining surfaces is flat are scarce in the literature. A justification for a journal bearing is presented in [ 6 ] , and [ 17 ] provides a justification for a flow through a curved pipe. To our knowledge, no rigorous justification of the Reynolds equation between two arbitrary surfaces is currently available. In this article, we present a rigorous proof that, in the case of a viscous fluid confined between two closely spaced spheres, the solution of the Stokes problem converges, in a sense that we will specify, to the solution of the Reynolds equation ( 168 )–( 171 ). In section 2 , we define the domain where the Stokes problem is posed, namely the region between two closely spaced spheres. This domain depends on a small dimensionless parameter \varepsilon related to the distance between the two spheres, which is assumed to tend to zero. We map the problem onto a reference domain independent of \varepsilon by means of a change of variable. In section 3 we formulate the Stokes problem in the domain defined in section 2 . In section 4 we define the function spaces, and their norms, that will be used to study the convergence of the solution of the Stokes problem. Section 5 is dedicated to proving the existence of a divergence-free velocity field, which satisfies the boundary condition ( 21 ) and, also, the estimate ( 54 ). In sections 6 and 7 we demonstrate the convergence of \vec{u}^{\varepsilon} and p^{\varepsilon} , where \vec{u}^{\varepsilon} and p^{\varepsilon} denote, respectively, the velocity field and pressure of the Stokes problem posed in section 3 . Finally, in section 8 , we determine their respective limits. 2 Working domain and a reference domain We wish to point out that, in this article, the word “sphere” should be understood as a simplification, and that the proposed procedure can be extended without major modifications to any two close surfaces defined in spherical coordinates by an expression of the type \rho=\rho(\varphi,\theta) . In most of the lubricated devices, the outer ring (the spherical housing) must be able to assemble over the inner spherical ball, and standard spherical bearings do not reach 180^{\circ} coverage, typically covering around 120^{\circ} - 150^{\circ} of the inner spherical surface (see chapter 12 in [ 16 ] , [ 19 ] , [ 20 ] ). Therefore, in the study that follows, we will only consider a sector of the inner sphere, which is assumed to have radius R , and which can be parameterized in cartesian coordinates, using spherical coordinates (\varphi,\theta,\rho) , as follows: \vec{X}(\varphi,\theta)=R(\sin\varphi\cos\theta,\sin\varphi\sin\theta,\cos\varphi),\quad(\varphi,\theta)\in D, (1) where D=\left\{(\varphi,\theta)\,/\,\varphi\in(\varphi_{0},\varphi_{1}),\ \theta\in(0,2\pi)\right\},\quad 0<\varphi_{0}<\varphi_{1}<\pi. (2) Remark 1 . Assumption ( 2 ) about the angle \varphi implies that there exists c_{0}>0 such that \sin\varphi\geq c_{0} for all \varphi\in(\varphi_{0},\varphi_{1}) . Remark 2 . As in the case of flat geometries, various boundary conditions can be considered, involving both pressure and velocities (see [ 7 ] ), and they are mainly related to operational conditions, the geometry of the bearing and the location of the supply holes. From a physical point of view, the regions \varphi=\varphi_{0} and \varphi=\varphi_{1} define holes through which lubricant is supplied to the device (see ( 59 )–( 60 ), Figure 1 , and [ 16 ] or [ 20 ] ). Figure 1: Example of spherical bearing The gap between the inner sphere and the outer sphere is given by \displaystyle h^{\varepsilon}(\varphi,\theta)=\varepsilon h(\varphi,\theta),\quad 0<h_{0}\leq h(\varphi,\theta)\leq h_{1}, (3) \displaystyle h\textrm{ is smooth and }2\pi\textrm{-periodic in the variable }\theta (4) Let us define the unit normal to the sphere ( 1 ) as \vec{N}(\varphi,\theta)=(\sin\varphi\cos\theta,\sin\varphi\sin\theta,\cos\varphi) (5) then, the domain between the two spheres is defined by \displaystyle\Omega^{\varepsilon} \displaystyle=\{\vec{x}=(x_{1},x_{2},x_{3})\in{\mathbb{R}}^{3}/ \displaystyle\quad X_{i}(\varphi,\theta)\leq x_{i}\leq X_{i}(\varphi,\theta)+\xi_{3}h^{\varepsilon}(\varphi,\theta)N_{i}(\varphi,\theta),\quad(i=1,2,3), \displaystyle\quad(\varphi,\theta)\in D,\xi_{3}\in(0,1)\} (6) In the above equations \varepsilon>0 denotes a small dimensionless parameter that we will make tend to zero, indicating that the distance between the two spheres is very small. Figure 2: Parameters \xi_{1},\xi_{2} Remark 3 . Denoting \vec{\xi}=(\xi_{1},\xi_{2},\xi_{3}) , with \xi_{1}=\varphi and \xi_{2}=\theta , the domain ( 6 ) can also be defined by means of a change of variables \vec{x}=\Phi^{\varepsilon}(\vec{\xi})=\vec{X}(\xi_{1},\xi_{2})+\varepsilon\xi_{3}h(\xi_{1},\xi_{2})\vec{N}(\xi_{1},\xi_{2}) (7) so we have that \Omega^{\varepsilon}=\Phi^{\varepsilon}(\Omega),\quad\Omega=D\times(0,1) (8) For the steps that follow, it will be useful to decompose the previous change of variables into the following simpler ones: \displaystyle\vec{\zeta}=(\zeta_{1},\zeta_{2},\zeta_{3})=\Phi_{1}^{\varepsilon}(\vec{\xi})=(\xi_{1},\xi_{2},\varepsilon\xi_{3}), (9) \displaystyle(\varphi,\theta,\rho)=\Phi_{2}^{\varepsilon}(\vec{\zeta})=(\zeta_{1},\zeta_{2},R+\zeta_{3}h(\zeta_{1},\zeta_{2})), (10) \displaystyle\vec{x}=\Phi_{3}^{\varepsilon}(\varphi,\theta,\rho)=(\rho\sin\varphi\cos\theta,\rho\sin\varphi\sin\theta,\rho\cos\varphi), (11) Thus, the change of variables ( 7 ) can be written \Phi^{\varepsilon}=\Phi_{3}^{\varepsilon}\circ\Phi_{2}^{\varepsilon}\circ\Phi_{1}^{\varepsilon} (12) and the domain ( 6 ) is expressed, for each of the different variables, \displaystyle\Omega_{i}^{\varepsilon}=\Phi_{i}^{\varepsilon}(\Omega_{i-1}^{\varepsilon}),\qquad(i=1,2,3), (13) \displaystyle\Omega_{0}^{\varepsilon}=\Omega, (14) so \Omega_{0}^{\varepsilon}=\Omega does