C 1 , α \displaystyle C^{1,\alpha} regularity for a class of singular/degenerate fully nonlinear elliptic equations with oblique boundary conditions

Sun-Sig Byun, Hongsoo Kim, SeungHyun Kim · 2026

Read C 1 , α \displaystyle C^{1,\alpha} regularity for a class of singular/degenerate fully nonlinear elliptic equations with oblique boundary conditions by Sun-Sig Byun, Hongsoo Kim, SeungHyun Kim (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-2604-06030.txt.

Introduction

Byun, Hongsoo Kim, and Seunghyun Kim prove global C^{1,alpha} regularity for viscosity solutions of a class of singular and degenerate fully nonlinear elliptic equations. That is a smoothness result for weak solutions. The result holds across the whole domain. They carry oblique boundary conditions, where the boundary rule at each edge point uses a directional derivative at an angle. Their 2026 paper extends earlier findings to a broader class of equations, with the singular case too.

Catalog record

CREATORSun-Sig Byun, Hongsoo Kim, SeungHyun Kim
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE52882
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2604.06030

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

C 1 , α \displaystyle C^{1,\alpha} regularity for a class of singular/degenerate fully nonlinear elliptic equations with oblique boundary conditions Sun-Sig Byun, Hongsoo Kim, SeungHyun Kim Abstract . In this paper, we establish global C 1 , α \displaystyle C^{1,\alpha} regularity for viscosity solutions to a class of singular and degenerate fully nonlinear elliptic equations subject to oblique boundary conditions. Our work extends the findings in [ 6 ] to a broader class of equations, notably encompassing the singular case. 2020 Mathematics Subject Classification Primary 35B65 35D40 35J25; Secondary 35J60, 35J70 1. Introduction In this paper, we consider global regularity results for viscosity solutions to a class of fully nonlinear elliptic equations of the form (1.1) \left\{\begin{aligned} \Phi(x,|Du|)F(D^{2}u)&=f&&\text{in }\Omega,\\ \beta\cdot Du&=g&&\text{on }\partial\Omega,\end{aligned}\right. where \displaystyle\Omega\subset\mathbb{R}^{n} is a bounded \displaystyle C^{1} domain with boundary \displaystyle\partial\Omega , \displaystyle F:\mathcal{S}^{n}\rightarrow\mathbb{R} a uniformly elliptic operator in the sense of (A1): and satisfies \displaystyle F(0)=0 , and \displaystyle\Phi:\Omega\times(0,\infty)\rightarrow[0,\infty) is a continuous function that exhibits singular and degenerate behavior in the gradient, as described in (A2): . Here, \displaystyle f\in C(\Omega) , \displaystyle g\in C^{\alpha}(\partial\Omega) , and \displaystyle\beta\in C^{\alpha}(\partial\Omega;\mathbb{R}^{n}) for some \displaystyle\alpha\in(0,1) , with \displaystyle\beta satisfying (A3): . We begin by discussing condition (A2): . This condition was introduced in [ 1 ] , where the interior Hölder continuity of the gradient for viscosity solutions to (1.2) \displaystyle\displaystyle\Phi(x,|Du|)F(D^{2}u)=f\quad\text{ in }B_{1}, was established under the assumptions that \displaystyle F is a fully nonlinear uniformly elliptic operator and \displaystyle f\in C(B_{1})\cap L^{\infty}(B_{1}) . We now present some examples satisfying (A2): as follows: (1) \displaystyle\Phi(x,t)=t^{p} with \displaystyle i(\Phi)=s(\Phi)=p , where \displaystyle p>-1 . (2) \displaystyle\Phi(x,t)=t^{p(x)} with \displaystyle i(\Phi)=\inf_{x\in\Omega}p(x) and \displaystyle s(\Phi)=\sup_{x\in\Omega}p(x) , where \displaystyle p(\cdot)\in C(\Omega) and \displaystyle\inf_{x\in\Omega}p(x)>-1 . (3) \displaystyle\Phi(x,t)=t^{p(x)}+a(x)t^{q(x)} with \displaystyle i(\Phi)=\inf_{x\in\Omega}\left\{p(x),q(x)\right\} and \displaystyle s(\Phi)=\sup_{x\in\Omega}\left\{p(x),q(x)\right\} , where \displaystyle p(\cdot),q(\cdot)\in C(\Omega) , \displaystyle\inf_{x\in\Omega}\left\{p(x),q(x)\right\}>-1 and \displaystyle 0\leq a(\cdot)\in C(\Omega) . (4) \displaystyle\Phi(x,t)=\phi(t)+a(x)\psi(t) where \displaystyle\phi and \displaystyle\psi are suitable \displaystyle N -functions and \displaystyle 0\leq a(\cdot)\in C(\Omega) . We next turn to boundary regularity results for singular and degenerate fully nonlinear elliptic equations. In [ 2 ] , the authors established global \displaystyle C^{1,\alpha} regularity for viscosity solutions to the Dirichlet problem associated with ( 1.2 ) on \displaystyle C^{2} domains. For the singular case corresponding to the power-type choice \displaystyle\Phi(x,t)=t^{\gamma} with \displaystyle-1<\gamma<0 , [ 12 ] proved global Lipschitz regularity under homogeneous Neumann boundary conditions on \displaystyle C^{2} domains. For the degenerate case \displaystyle\Phi(x,t)=t^{\gamma} with \displaystyle\gamma>0 , [ 3 ] obtained global \displaystyle C^{1,\alpha} regularity under nonhomogeneous Neumann boundary conditions on \displaystyle C^{2} domains. For oblique boundary conditions, global \displaystyle C^{1,\alpha} regularity was established in [ 6 ] for the same power-type choice \displaystyle\Phi(x,t)=t^{\gamma} with \displaystyle\gamma>0 on \displaystyle C^{1} domains. For further references, including those concerning uniformly elliptic equations, we refer the reader to [ 13 , 4 , 5 , 11 , 10 , 7 ] and the references therein. Our goal in this paper is to establish global \displaystyle C^{1,\alpha} regularity for viscosity solutions to a general class of equations of the form ( 1.1 ), under nonhomogeneous oblique boundary conditions on \displaystyle C^{1} domains. The novelty of our result lies in treating both the degenerate case ( \displaystyle i(\Phi)\geq 0 ) and the singular case ( \displaystyle-1<i(\Phi)<0 ) within a unified framework. In the degenerate regime, our result builds on [ 6 ] and applies to a broader class of nonlinearities beyond the power-type setting. In the singular regime, we improve upon the result of [ 12 ] , where global Lipschitz regularity was established via the Ishii–Lion’s method under homogeneous Neumann boundary conditions. We note that, in this singular regime, even the extension to nonhomogeneous Neumann boundary conditions is not straightforward. We now briefly explain the strategy used in the proof. To establish global \displaystyle C^{1,\alpha} regularity, we show that a solution \displaystyle u can be approximated by a sequence of affine functions with an error of order \displaystyle r^{k(1+\alpha)} at each scale \displaystyle r^{k} (for some fixed \displaystyle r\in(0,1) ) via an iterative argument. More precisely, we obtain a sequence of affine functions \displaystyle\left\{l_{k}\right\} satisfying \displaystyle\displaystyle\left\lVert u-l_{k}\right\rVert_{L^{\infty}\left(\Omega\cap B_{r^{k}}\right)}\leq r^{k(1+\alpha)}\qquad\text{ for all }\;k\geq 0. The case \displaystyle k=1 follows from a compactness argument. By iterating this procedure—namely, by applying the compactness argument repeatedly to the scaled functions \displaystyle v_{k} defined by \displaystyle\displaystyle v_{k}(x):=\frac{(u-l_{k})(r^{k}x)}{r^{k(1+\alpha)}}, we obtain the estimate for any \displaystyle k . Note that \displaystyle v_{k} does not satisfy the original equation ( 1.1 ). Instead, it solves an equation of the form (1.3) \displaystyle\displaystyle\widetilde{\Phi}(|Dv_{k}-q|)F(D^{2}v_{k})=\tilde{f}, where \displaystyle q=-\nabla l_{k}/r^{k\alpha} and \displaystyle\widetilde{\Phi} satisfies (A2): after an appropriate scaling (see Section 2.3 for details). Hence, to implement the compactness argument, we need an equicontinuity estimate for solutions to ( 1.3 ) that remains uniform in \displaystyle q . In [ 9 ] , the authors introduced the sliding cusp method to establish Hölder estimates for uniformly elliptic equations that hold only where the gradient is large. In [ 6 ] , this result was extended to the oblique boundary setting. To address the difficulties arising from oblique boundary conditions, the authors combined the sliding paraboloid and cusp methods and exploited a modified scaling property adapted to the equation. As a consequence, they obtained Hölder estimates for uniformly elliptic equations that hold only where the gradient is far from some point on \displaystyle C^{1} domains. By combining the above result with the structural condition (A2): on \displaystyle\Phi , we obtain the desired estimate in the degenerate case. In the singular case, after a suitable transformation (see Lemma 2.2 for more details), a term of the form \displaystyle|Du|^{\gamma} with \displaystyle\gamma\in[0,1) appears, which creates a major difficulty in the analysis. More specifically, ( 1.3 ) takes the form \displaystyle\displaystyle\Phi\left(x,|Dv-q|\right)F(D^{2}v)=f|Dv-q|^{\gamma}, for an arbitrary \displaystyle q\in\mathbb{R}^{n} . As mentioned before, we need to find a uniform analysis in \displaystyle q . However, as \displaystyle|q| becomes very large, the right-hand side may grow indefinitely, hindering the derivation of uniform estimates independent of \displaystyle q . To circumvent this difficulty, we restrict our attention to the regime where \displaystyle|q|^{\gamma}\left\lVert f\right\rVert_{L^{\infty}} is bounded; in this case, the arguments from [ 8 , 6 ] remain applicable, without resorting to the Ishii-Lion’s method. We now state the main results of this paper. Theorem 1.1 . Let \displaystyle\Omega be a bounded \displaystyle C^{1} domain and \displaystyle 0\in\partial\Omega . Assume (A1): – (A3): (see Section 2 ). In addition, assume that \displaystyle f\in C(\Omega) , \displaystyle g\in C^{\alpha}(\partial\Omega) , and \displaystyle\beta\in C^{\alpha}(\partial\Omega;\mathbb{R}^{n}) for some \displaystyle\alpha\in(0,1) . Let \displaystyle u be a viscosity solution to ( 1.1 ). Then \displaystyle u\in C^{1,\overline{\alpha}}(\overline{\Omega\cap B_{1/2}}) with the estimate \displaystyle\displaystyle\left\lVert u\right\rVert_{C^{1,\overline{\alpha}}\left(\overline{\Omega\cap B_{1/2}}\right)}\leq c\left(\left\lVert u\right\rVert_{L^{\infty}(\Omega\cap B_{1})}+\left\lVert f\right\rVert_{L^{\infty}(\Omega\cap B_{1})}^{1/(1+i(\Phi))}+\left\lVert g\right\rVert_{C^{\alpha}(\partial\Omega\cap B_{1})}\right), where (1.4) \overline{\alpha}\in\begin{cases}(0,\alpha_{0})\cap\left(0,\frac{1}{1+s(\Phi)}\right]\cap(0,\alpha]\quad&\text{if }i(\Phi)\geq 0,\\ (0,\alpha_{0})\cap\left(0,\frac{1}{1+s(\Phi)-i(\Phi)}\right]\cap(0,\alpha]\quad&\text{if }-1<i(\Phi)<0,\end{cases} \displaystyle c=c(n,\lambda,\Lambda,i(\Phi),M,\nu_{0},\delta_{0},[\beta]_{C^{\alpha}(\partial\Omega\cap B_{1})},C^{1}\text{ modulus of }\partial\Omega\cap B_{1},\overline{\alpha}) , and \displaystyle\alpha_{0} will be specified in Remark 1.1 . Remark 1.1 . The constant \displaystyle\alpha_{0}=\alpha_{0}(n,\lambda,\Lambda,\delta_{0})\in(0,1] in the statement of the theorem denotes the optimal exponent of regularity theory for a homogeneous equation with a constant oblique boundary condition. More precisely, any viscosity solution \displaystyle h of \left\{\begin{aligned} F(D^{2}h)&=0&&\text{in }B_{1}^{+},\\ \beta_{0}\cdot Dh&=0&&\text{on }T_{1},\end{aligned}\right. where \displaystyle\beta_{0} is a constant vector satisfying (A3): , belongs to the class \displaystyle C^{1,\alpha_{0}}_{\text{loc}}(B^{+}_{1}) and satisfies the estimate \displaystyle\displaystyle\left\lVert h\right\rVert_{C^{1,\alpha_{0}}(B^{+}_{3/4})}\leq C_{e}\left\lVert h\right\rVert_{L^{\infty}(B^{+}_{1})}, for some constant \displaystyle C_{e}=C_{e}(n,\lambda,\Lambda,\delta_{0})>1 . The authors in [ 10 ] proved that

Read the complete text