betaFeedback

WorksMathematics

Second variation of the fractional torsion energy and linear stability of the ball

Abstract

We compute the second shape derivative of the fractional torsion. It is an explicit quadratic form on the boundary, which involves a Dirichlet-to-Neumann operator for half-harmonic functions blowing up at the boundary. We deduce an explicit formula for volume preserving variations on the ball and show that as in the local case, the shape Hessian is negative and coercive in H1/2H^{1/2} modulo translations.

Introduction

The aim of this paper is to show, in arguably the simplest meaningful situation, how second order shape derivatives can be computed for shape optimization problems involving operators of fractional order. To this end we compute the second shape derivative of the fractional torsion energy for smooth domains and the exponent s=1/2s = 1/2. While second order shape derivatives are well understood in the local setting, see e.g. [6, 20], and first order derivatives for fractional problems have received considerable attention, see e.g. [7, 9, 10, 23], second order derivatives in the fractional setting have been less studied. We expect the approach developed here to be useful for related problems, such as fractional capacities and first Dirichlet eigenvalues, and for other exponents s∈(0,1)s \in(0, 1).

Our initial motivation was quantitative stability of the ball. For problems in which the ball is a volume-constrained optimizer (a minimizer or a maximizer, depending on the problem), computing the second variation and establishing its coercivity are essential steps in carrying out the program of [6], namely a proof of sharp quantitative stability through the study of nearly spherical sets and a selection principle. We were indeed on this path when we found, in [15], a much shorter and less technical proof of quantitative stability, adapting the strategy of [19]. We nevertheless decided to write down a rigorous derivation of the second variation, since it may be useful for other purposes. With this in mind we also prove its coercivity at the ball under volume-preserving variations.

Following the full strategy of [6] would require further results. First, one would need weaker regularity assumptions on the domains, or a regularity bootstrap for the nearly spherical sets arising in the contradiction argument; see [4, 22]. Second, one would need continuity estimates for the second variation in the relevant Sobolev norm, here H1/2H^{1/2}. This might require the framework of [16, 17, 18], instead of the regularity results of [3] used here. Finally, the selection principle would require an adaptation of the regularity theory for fractional Bernoulli free boundary problems developed in [8].

Let Ω⊂Rn\Omega\subset\mathbb{R}^n, n≥2n \ge2, be a bounded open set with C∞C^\infty boundary. The fractional torsion function uΩu_\Omega is the solution of

(−Δ)1/2uΩ=1 in Ω,uΩ=0 in Rn∖Ω,(1)(-\Delta)^{1/2}u_\Omega=1 \text{ in }\Omega,\qquad u_\Omega=0 \text{ in }\mathbb{R}^n\setminus\Omega, \tag*{(1)}

where (−Δ)1/2(-\Delta)^{1/2} is the fractional Laplacian with Fourier symbol ∣ξ∣|\xi|,

(−Δ)1/2v(x)=cnP.V.⁡∫Rnv(x)−v(y)∣x−y∣n+1 dy,cn=Γ((n+1)/2)π(n+1)/2,(-\Delta)^{1/2}v(x)=c_n\operatorname{P.V.}\int_{\mathbb{R}^n}\frac{v(x)-v(y)}{|x-y|^{n+1}}\,\mathrm{d}y,\qquad c_n=\frac{\Gamma((n+1)/2)}{\pi^{(n+1)/2}},

and the fractional torsional rigidity, or torsion energy, of Ω\Omega is

T(Ω)=∫ΩuΩ.T(\Omega)=\int_\Omega u_\Omega.

The fractional Saint-Venant inequality states that balls are the (unique) maximizers of the fractional torsion under volume constraint, see e.g. [5, 15]. It is then a natural question to compute the corresponding first and second order shape derivatives. Let us first recall what is known for the first order derivatives from the works [7, 9, 10, 11, 23]. For this we need to introduce some notation.

Problem (1) is understood in the energy space

V={v∈H1/2(Rn):v=0 a.e. in Rn∖Ω},V=\{v\in H^{1/2}(\mathbb{R}^n):v=0\ \text{a.e. in }\mathbb{R}^n\setminus\Omega\},
a(v1,v2)=cn2∬R2n(v1(y)−v1(z))(v2(y)−v2(z))∣y−z∣n+1 dy dz,(2)a(v_1,v_2)=\frac{c_n}{2}\iint_{\mathbb{R}^{2n}}\frac{(v_1(y)-v_1(z))(v_2(y)-v_2(z))}{|y-z|^{n+1}}\,\mathrm{d}y\,\mathrm{d}z, \tag*{(2)}

so that a(v1,v2)=∫Rnv1(−Δ)1/2v2a(v_1,v_2)=\int_{\mathbb{R}^n}v_1(-\Delta)^{1/2}v_2 for smooth functions. Then uΩu_\Omega is the unique element of VV with a(uΩ,v)=∫Ωva(u_\Omega,v)=\int_\Omega v for every v∈Vv\in V, and

T(Ω)=max⁡v∈V{2∫Ωv−a(v,v)},(3)T(\Omega)=\max_{v\in V}\left\{2\int_\Omega v-a(v,v)\right\}, \tag*{(3)}

the maximum being attained at v=uΩv=u_\Omega.

We write ν\nu for the outer unit normal to ∂Ω\partial\Omega, H=div⁡∂ΩνH=\operatorname{div}_{\partial\Omega}\nu for the mean curvature and dd for the distance to ∂Ω\partial\Omega. We fix r0>0r_0>0 such that (x,r)↦x−rν(x)(x,r)\mapsto x-r\nu(x) is a diffeomorphism from ∂Ω×(−2r0,2r0)\partial\Omega\times(-2r_0,2r_0) onto a neighbourhood of ∂Ω\partial\Omega, and a function δ∈C∞(Ω‾)\delta\in C^\infty(\overline{\Omega}), positive in Ω\Omega, with δ=d\delta=d on {y∈Ω:d(y)<r0}\{y\in\Omega:d(y)<r_0\}. Thus d(x−rν(x))=δ(x−rν(x))=rd(x-r\nu(x))=\delta(x-r\nu(x))=r for x∈∂Ωx\in\partial\Omega and 0<r<r00<r<r_0.

The torsion function behaves like d\sqrt{d} at the boundary [24]: by Lemma 2.1 below, uΩ=δ Uu_\Omega=\sqrt{\delta}\,U with U∈C∞(Ω‾)U\in C^\infty(\overline{\Omega}). The function UU depends on the choice of δ\delta away from ∂Ω\partial\Omega, but its behaviour near ∂Ω\partial\Omega does not, since there U(x−rν(x))=r−1/2uΩ(x−rν(x))U(x-r\nu(x))=r^{-1/2}u_\Omega(x-r\nu(x)). The torsion function enters our formulas only through two boundary quantities: the weighted trace ψ\psi of uΩu_\Omega and its first normal coefficient κ\kappa, defined for x∈∂Ωx\in\partial\Omega by

ψ(x)=lim⁡r↓0uΩ(x−rν(x))r=U(x),κ(x)=∂rU(x−rν(x))∣r=0,\psi(x)=\lim_{r\downarrow0}\frac{u_\Omega(x-r\nu(x))}{\sqrt{r}}=U(x),\qquad\kappa(x)=\left.\partial_rU(x-r\nu(x))\right|_{r=0},

so that uniformly in x∈∂Ωx\in\partial\Omega

uΩ(x−rν(x))=ψ(x)r+κ(x)r3/2+O(r5/2).(4)u_\Omega(x-r\nu(x))=\psi(x)\sqrt{r}+\kappa(x)r^{3/2}+O(r^{5/2}). \tag*{(4)}

In terms of these coefficients, the first variation of TT is as follows. Let Φt\Phi_t be any smooth family of diffeomorphisms of Rn\mathbb{R}^n with Φ0=Id⁡\Phi_0=\operatorname{Id}, let X=∂tΦt∣t=0X=\left.\partial_t\Phi_t\right|_{t=0} and φ=X⋅ν\varphi=X\cdot\nu on ∂Ω\partial\Omega. Let Ωt=Φt(Ω)\Omega_t=\Phi_t(\Omega). Then

ddtT(Ωt)∣t=0=c∗∫∂Ωψ2φ dHn−1,c∗:=Γ(3/2)2=π4.(5)\left.\frac{d}{dt}T(\Omega_t)\right|_{t=0}=c_*\int_{\partial\Omega}\psi^2\varphi\,d\mathcal{H}^{n-1},\qquad c_*:=\Gamma(3/2)^2=\frac{\pi}{4}. \tag*{(5)}

As in [10] where the fractional Dirichlet eigenvalue is considered, we recover the first variation formula (5) through a short application of the generalized fractional Pohozaev identity of [10]; see Remark 3.6.

We now turn to the second order derivative. The difficulty comes from understanding the PDE which is solved by the Eulerian derivative q=∂tuΩt∣t=0q=\left.\partial_tu_{\Omega_t}\right|_{t=0}. Indeed, by differentiating the PDE solved by uΩtu_{\Omega_t} we can guess that it is 1/21/2-harmonic in Ω\Omega and equal to zero outside Ω\Omega. What makes it both non-trivial and non-standard is that it blows up like the inverse square root of the distance to the boundary. In particular, it lies outside the energy space in which (1) is posed. These are ”large solutions” in the language of [1, 26]. One of the main points of the analysis is thus to understand a certain Dirichlet-to-Neumann-type operator DΩ\mathcal{D}_{\Omega}, defined in (7) below, which acts on the boundary data of these large 1/21/2-harmonic functions. Here we will only need an elementary construction of these functions and of DΩ\mathcal{D}_{\Omega} (Proposition 2.4) which we now describe. For every f∈C∞(∂Ω)f\in C^\infty(\partial\Omega) we prove that there is a unique function QfQ_f, vanishing outside Ω\Omega and with δQf∈C∞(Ω‾)\sqrt{\delta}Q_f\in C^\infty(\overline{\Omega}), such that

(−Δ)1/2Qf=0 in Ω,(δQf)∣∂Ω=f.(6)(-\Delta)^{1/2}Q_f=0\ \text{in }\Omega,\qquad\left.(\sqrt{\delta}Q_f)\right|_{\partial\Omega}=f. \tag*{(6)}

Here the equation holds in the sense of distributions: QfQ_f is integrable, but it blows up like d−1/2d^{-1/2} at ∂Ω\partial\Omega and is not in L2L^2 unless f=0f=0. We set, for x∈∂Ωx\in\partial\Omega,

DΩf(x)=∂r(rQf(x−rν(x)))∣r=0,(7)\mathcal{D}_{\Omega}f(x)=\left.\partial_r\left(\sqrt{r}Q_f(x-r\nu(x))\right)\right|_{r=0}, \tag*{(7)}

that is, Qf(x−rν(x))=f(x)r−1/2+DΩf(x)r+O(r3/2)Q_f(x-r\nu(x))=f(x)r^{-1/2}+\mathcal{D}_{\Omega}f(x)\sqrt{r}+O(r^{3/2}). The construction of DΩ\mathcal{D}_{\Omega} in Proposition 2.4 only uses the regularity result Lemma 2.1.

We now give the main statement. We will limit ourselves to the computation in the case of normal graphs. This is very common in the study of shape derivatives, see [20]. Fix φ∈C∞(∂Ω)\varphi\in C^\infty(\partial\Omega). For x∈∂Ωx\in\partial\Omega and ∣r∣≤r0/2|r|\le r_0/2 we set

X(x−rν(x))=φ(x)ν(x)(8)X(x-r\nu(x))=\varphi(x)\nu(x) \tag*{(8)}

which we extend to a smooth vector field XX compactly supported in {d<r0}\{d<r_0\}. We then set

References

  1. [1]Abatangelo, N. Large s-harmonic functions and boundary blow-up solutions for the fractional Laplacian. Discrete Contin. Dyn. Syst. 35, 12 (2015), 5555–5607.arxiv.org/abs/1310.3193
  2. [2]Abatangelo, N., Jarohs, S., and Saldaña, A. Integral representation of solutions to higher-order fractional Dirichlet problems on balls. Commun. Contemp. Math. 20, 8 (2018), 1850002, 36 pp.DOI
  3. [3]Abatangelo, N., and Ros-Oton, X. Obstacle problem for integro-differential operators: higher regularity of free boundaries. Adv. Math. 360 (2020), 106931, 61 pp.arxiv.org/abs/1909.10339
  4. [4]Barrios, B., Ros-Oton, X., and Weidner, M. Higher regularity in nonlocal free boundary problems. Preprint, arXiv:2507.20581, 2025.arxiv.org/abs/2507.20581
  5. [5]Brasco, L., Cinti, E., and Vita, S. A quantitative stability estimate for the fractional Faber–Krahn inequality. J. Funct. Anal. 279, 3 (2020), 108560.DOI
  6. [6]Brasco, L., De Philippis, G., and Velichkov, B. Faber–Krahn inequalities in sharp quantitative form. Duke Math. J. 164, 9 (2015), 1777–1831.DOI
  7. [7]Dalibard, A.-L., and Gérard-Varet, D. On shape optimization problems involving the fractional Laplacian. ESAIM Control Optim. Calc. Var. 19, 4 (2013), 976–1013.arxiv.org/abs/1202.4920
  8. [8]De Silva, D., Savin, O., and Sire, Y. A one-phase problem for the fractional Laplacian: regularity of flat free boundaries. Bull. Inst. Math. Acad. Sin. (N.S.) 9 (2014), 111–145.arxiv.org/abs/1401.6443
  9. [9]Djitte, S. M., Fall, M. M., and Weth, T. A fractional Hadamard formula and applications. Calc. Var. Partial Differential Equations 60, 6 (2021), Paper No. 231, 31 pp.arxiv.org/abs/2002.07719
  10. [10]Djitte, S. M., Fall, M. M., and Weth, T. A generalized fractional Pohozaev identity and applications. Adv. Calc. Var. 17, 1 (2024), 237–253.arxiv.org/abs/2112.10653
  11. [11]Djitte, S. M., and Sueur, F. Pointwise Hadamard variational formula for the fractional Laplacian. Preprint, arXiv:2602.07214, 2026.arxiv.org/abs/2602.07214
  12. [12]Dyda, B. Fractional calculus for power functions and eigenvalues of the fractional Laplacian. Fract. Calc. Appl. Anal. 15, 4 (2012), 536–555.DOI
  13. [13]Fiscella, A., Servadei, R., and Valdinoci, E. Density properties for fractional Sobolev spaces. Ann. Acad. Sci. Fenn. Math. 40, 1 (2015), 235–253.DOI
  14. [14]Getoor, R. K. First passage times for symmetric stable processes in space. Trans. Amer. Math. Soc. 101 (1961), 75–90.DOI
  15. [15]Goldman, M., Novaga, M., and Ruffini, B. The fractional Faber–Krahn inequality in sharp quantitative form. Preprint.arxiv.org/abs/2610.40232
  16. [16]Grubb, G. Functional Calculus of Pseudodifferential Boundary Problems, 2 ed., vol. 65 of Progress in Mathematics. Birkhäuser, Boston, 1996.DOI
  17. [17]Grubb, G. Local and nonlocal boundary conditions for μ-transmission and fractional elliptic pseudodifferential operators. Anal. PDE 7 (2014), 1649–1682.DOI
  18. [18]Grubb, G. Fractional Laplacians on domains, a development of Hörmander’s theory of μ-transmission pseudodifferential operators. Adv. Math. 268 (2015), 478–528.DOI
  19. [19]Guerra, A., Machado, J. M., and Ramos, J. P. G. On the effective sharp stability of the Faber–Krahn inequality. Preprint, arXiv:2609.07772, 2026.arxiv.org/abs/2609.07772
  20. [20]Henrot, A., and Pierre, M. Shape variation and optimization. A geometrical analysis, vol. 28 of EMS Tracts in Mathematics. European Mathematical Society, Zürich, 2018.
  21. [21]Iannizzotto, A., Mosconi, S., and Squassina, M. Global Hölder regularity for the fractional p-Laplacian. Rev. Mat. Iberoam. 32, 4 (2016), 1353–1392.DOI
  22. [22]Lyu, R. C∞ regularity for the free boundary of one-phase fractional Laplacian problem. Preprint, arXiv:2509.00609, 2025.arxiv.org/abs/2509.00609
  23. [23]Muratov, C. B., Novaga, M., and Ruffini, B. Conducting flat drops in a confining potential. Archive for Rational Mechanics and Analysis 243, 3 (2022), 1773–1810.DOI
  24. [24]Ros-Oton, X., and Serra, J. The Dirichlet problem for the fractional Laplacian: regularity up to the boundary. J. Math. Pures Appl. (9) 101, 3 (2014), 275–302.
  25. [25]Ros-Oton, X., and Serra, J. The Pohozaev identity for the fractional Laplacian. Arch. Ration. Mech. Anal. 213, 4 (2014), 587–628.
  26. [26]Ros-Oton, X., and Weidner, M. Regularity for nonlocal equations with local Neumann boundary conditions. Analysis & PDE 19 (2026), 353–411.

Paper details

vlnc:W_agcpseaa
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 2

A view counts once per browser per day, a download each time the PDF is requested.

  • Mathematics

a large language model (LLM)

View statement
Looking for related papers…
Contents