MathLabs
TheoremProved

Weak maximum principle for the heat equation

Statement

Let uu be continuous on [0,L]×[0,T][0,L]\times[0,T] and satisfy ut=αuxxu_t = \alpha u_{xx} on the open rectangle 0<x<L, 0<t≤T0<x<L,\ 0<t\le T. Then the maximum of uu over the whole closed rectangle is already attained on the "parabolic boundary" — the initial edge t=0t=0 or one of the two sides x=0x=0, x=Lx=L — never at an interior point with t>0t>0 unless uu is constant there.

Why is it true?

Heat has nowhere to come from except the initial data and the boundary temperatures: if a hidden interior point were the single hottest spot at some later time, heat would have to be spontaneously created there, which the diffusion process forbids — diffusion only ever moves heat from hot to cold, it never manufactures a new peak in empty space.

Proof sketch

Fix ε>0\varepsilon>0 and define v(x,t)=u(x,t)−εtv(x,t) = u(x,t) - \varepsilon t, so vt−αvxx=ut−ε−αuxx=−ε<0v_t - \alpha v_{xx} = u_t - \varepsilon - \alpha u_{xx} = -\varepsilon < 0 everywhere on the open rectangle. Suppose, for contradiction, that vv attains its maximum over the closed rectangle at an interior point (x0,t0)(x_0,t_0) with 0<x0<L0<x_0<L and 0<t0≤T0<t_0\le T — i.e. off the parabolic boundary.

Because x0x_0 is an interior spatial maximum, the second-derivative test gives vxx(x0,t0)≤0v_{xx}(x_0,t_0)\le 0. Because t0t_0 is where the maximum over t∈[0,T]t\in[0,T] is reached (either an interior critical time, giving vt(x0,t0)=0v_t(x_0,t_0)=0, or the endpoint t0=Tt_0=T approached from the left, giving vt(x0,t0)≥0v_t(x_0,t_0)\ge 0), in either case vt(x0,t0)≥0v_t(x_0,t_0)\ge 0.

Combining the two inequalities, vt(x0,t0)−αvxx(x0,t0)≥0−α⋅0=0v_t(x_0,t_0) - \alpha v_{xx}(x_0,t_0) \ge 0 - \alpha\cdot 0 = 0, which contradicts vt−αvxx=−ε<0v_t-\alpha v_{xx}=-\varepsilon<0 at every point. Hence no such interior maximum exists, so vv's maximum lies on the parabolic boundary Γ\Gamma (the initial edge together with the two lateral edges), giving u(x,t)≤εt+max⁡Γuu(x,t) \le \varepsilon t + \max_{\Gamma} u everywhere. Letting ε→0+\varepsilon \to 0^+ proves the maximum principle for uu itself.

Topics that use this theorem

Step-by-step proofs

No step-by-step proof yet for this theorem.

References

  1. Lawrence C. Evans (2010). Partial Differential Equations
  2. James Ward Brown, Ruel V. Churchill (2011). Fourier Series and Boundary Value Problems