Blocking of 2D bistable reaction-diffusion fronts by obstacles Preprint: APS/123-QED
J.-G. Caputo, G. Cruz-Pacheco, J. Gatlik, Benoit Sarels · 2026
Read Blocking of 2D bistable reaction-diffusion fronts by obstacles Preprint: APS/123-QED by J.-G. Caputo, G. Cruz-Pacheco, J. Gatlik, Benoit Sarels (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-2604-15246.txt.
Introduction
Caputo and others model how obstacles block 2D bistable reaction-diffusion fronts. Using a conservation law, they show the total of the reaction term acts as the front driving force. Paired with the exact 1D traveling wave, this yields a reduced model that predicts blocking points. Clear rules cover a waveguide joined to a cone-shaped region for widths below 4, and the model matches test runs and reaches checkerboard blocks.
Catalog record
| CREATOR | J.-G. Caputo, G. Cruz-Pacheco, J. Gatlik, Benoit Sarels |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 24678 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.15246
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Blocking of 2D bistable reaction-diffusion fronts by obstacles Preprint: APS/123-QED J.-G. Caputo, G. Cruz-Pacheco, J. Gatlik, Benoit Sarels Abstract We investigate numerically the blocking of two-dimensional bistable reaction–diffusion fronts by geometric obstacles. Our goal is to derive quantitative criteria for front propagation in the presence of spatial heterogeneities. Using a conservation-law approach, we show that the integral of the reaction term acts as an effective driving force for the front. Combining this insight with the exact one-dimensional traveling wave solution, we construct a reduced analytical model that predicts blocking thresholds. In particular, we obtain explicit conditions for front propagation in a waveguide connected to a conical region of angle θ \theta , valid for widths w ≲ 4 w\lesssim 4 . The model captures the influence of both geometry and nonlinearity, and shows good agreement with numerical simulations. Finally, we extend the analysis to more complex geometries, including checkerboard-like obstacles, and derive simple heuristic rules governing front propagation. I Introduction Reaction diffusion equations are ubiquitous in chemistry and biology, see the reviews by Scott Scott , Murray Murray , and the more recent review by Volpert Volpert . Examples are fungus propagation in crops of wheat or barley, the propagation of an electrical impulse in a nerve and the propagation of an epidemic in a geographic network. Many of these equations can only be solved numerically. In 1D however, two important models have exact solutions: the Zeldovich and the Fisher equations with nonlinearities that are cubic and quadratic respectively yielding bistable respectively monostable stationary states. These exact solutions provide valuable insight into the dynamics and serve as effective approximations of the system in the presence of perturbations. A fundamental question concerns the interaction of such fronts with spatial heterogeneities. Using a perturbation method together with extensive numerical calculations the present authors analyzed cs11 the pinning of a kink by abrupt large amplitude spatially localized defects for the cubic bistable model. In ccs21 we analyzed stopping of a bistable kink by a no reaction zone. On the other hand, a monostable front will always cross a non reaction region ccs21 . Another important defect is an obstacle corresponding to an inaccessible region in the domain. For example, a bistable reaction-diffusion front propagating in an inhomogeneous waveguide can be blocked by a sudden enlargement, see Chapuisat and Grenier cg05 and Berestycki et al bbc16 . In physiology, the nerve impulse in a neuron can be stopped by a sudden enlargement of the axon, see the nice discussion in bbc16 and also thesebouhours . See also the preliminary analysis in c18 . The studies bbc16 and thesebouhours prove passage or not passage depending on geometric conditions. However they do not give precise values for the widths of the waveguide leading to blocking for a given nonlinearity. Therefore, the objective of the present study is to derive quantitative thresholds for front propagation in heterogeneous geometries. We address the following questions: (i) What are the critical widths leading to blocking? (ii) How do more complex obstacles affect propagation? (iii) Can front blocking be predicted by a reduced analytical model ? Using an approximate analysis based on the 1D front solution, we obtained an analytical model describing the blocking of the front in the junction between two waveguides of different widths (w_{1},w_{2}) or a waveguide connected to a cone of angle \theta . This model involves the integral of the reaction term over the front. Using it, we can derive explicit values of the widths (w_{1},w_{2}) or (w_{1},\theta) causing blocking; this value also depends on the parameter a of the bistable nonlinearity. The article is organized as follows: section II presents the model and the numerical method, section III introduces an analysis where we approximate the solution to obtain reduced models. These models are compared to numerical results in section IV. There we also describe two parallel waveguides connected to a cavity, as well as a checkerboard geometry. Conclusions are drawn in section V. II The model We consider the bistable reaction-diffusion model u_{t}-\nabla(b\nabla u)+u(1-u)(u-a)=0, (1) in a 2D rectangular domain \Omega=[0,L]\times[0,L] with homogeneous Neumann boundary conditions. The nonlinearity R(u)=u(1-u)(u-a), is the standard bistable nonlinearity. The term b(x,y) is spatially dependent. For b=1 in 1D cs11 we have the exact front solution u={1\over 1+\exp[\sqrt{1\over 2}(x-ct)]}. (2) The speed and width of the front are given by c=\sqrt{1\over 2}(1-2a),~~~w=\sqrt{2}. (3) Note that c=0 for a=0.5 . In this study, we will consider specifically an inhomogeneous b . More precisely the spatial dependance of b(x,y) allows to represent different types of obstacles: b=1 outside the obstacle and b<<1 inside the obstacle. We will examine two different numbers associated to the solution u : the integral of u and of the reaction term R(u) <u>(t)\equiv{\int_{\Omega}dxdyu\over\int_{\Omega}dxdy}. (4) \mathbf{R}\equiv\int_{\Omega}R(u)dxdy. (5) II.1 Radial solutions In 2D, an important class of solutions are radial fronts. In this section, we show how they are related to the 1D exact solution ( 2 ). Consider the static problem ( 1 ), \Delta u+R(u)=0, (6) on a radial 2D domain of size L . Two numerical schemes were tested, a Newton-Raphson algorithm and the relaxation scheme presented in bl1980 : -\Delta u^{k+1}+Ku^{k+1}=Ku^{k}+R(u^{k}), (7) where K>0 is a constant. We found that imposing u_{r}=0 at r=0 always leads to a zero solution. This is consistent with the fact that a radial front will always propagate so that there are no static radial fronts. To obtain a nonzero static solution, we need to fix u=u_{0} at r=0 . Results are shown in Fig. 1 for L=2 (left panel) and L=10 (right panel). We chose n=400 grid points and used the finite difference scheme of Strickwerda sn86 with K=2 . The scheme converges rapidly, reaching a Cauchy residual of 10^{-9} in about 100 iterations. For this type of nonlinearity, there is no minimal radius to obtain a solution bl1980 , so there is always a solution for any L>0 . Clearly for L=2 , the domain is too narrow to ”fit” the solution, we need to increase it so that the solution and its tangent are close to zero for r=L . In the right panel of Fig. 1 , we included the kink profile centered on x_{0}=0 and with a width w_{0}=0.5 . Note the good agreement with the static solution. The typical scale of variation of the 2D solution is about 5, we then expect 2D effects to appear for inhomogeneities larger than that value. This will be confirmed in the following sections. Figure 1: Solution of the static problem for L=2 (left) and L=10 (right). The parameter a=0.3 . III Analysis In a recent article Berestycki, Bouhours and Chapuisat bbc16 proved that a bistable front will always propagate in a waveguide with decreasing cross section. By contrast, if the waveguide opens up abruptly, they proved that there can be blocking. They consider ”blocked” solutions and show that these can exist in some configurations. In this article, we use a direct argument based on conservation laws derived from the partial differential equation to distinguish between blocking and propagation of the front. III.1 Waveguide with sharp transition To simplify the discussion, we start with a junction between a left waveguide of cross-section w_{1} and a right waveguide of