26.10.2014 Aufrufe

Weak Convergence Methods for Nonlinear Partial Differential ...

Weak Convergence Methods for Nonlinear Partial Differential ...

Weak Convergence Methods for Nonlinear Partial Differential ...

MEHR ANZEIGEN
WENIGER ANZEIGEN

Erfolgreiche ePaper selbst erstellen

Machen Sie aus Ihren PDF Publikationen ein blätterbares Flipbook mit unserer einzigartigen Google optimierten e-Paper Software.

Figure 3.4: A “rarefaction wave”.<br />

then<br />

∂ t Φ(u) = DΦ(u) ∂ t u<br />

= −DΦ(u) DF(u) ∂ x u (u is a solution)<br />

= −DΨ(u) ∂ x u<br />

= − div Ψ(u).<br />

That is, Φ(u) satisfies a scalar conservation law with flux Ψ(u):<br />

∂ t Φ(u) + div Ψ(u) = 0.<br />

Now <strong>for</strong> a non-smooth u, we cannot expect this to be true in general.<br />

The guiding idea <strong>for</strong> a selection criterion is that in<strong>for</strong>mation (about the solution)<br />

is transported ( ) along characteristics of the equation, i.e., along the characteristic<br />

curves , where<br />

x(t)<br />

t<br />

x = x(t) = F ′ (y(t)) = F ′ (y(0)).<br />

Such in<strong>for</strong>mation is lost in shocks and characteristic curves whose origin lies in a<br />

shock will in general not carry physically relevant bits of in<strong>for</strong>mation.<br />

As a paradigm we consider a shock in a single equation (i.e., m = 1) along<br />

a curve C, parametrized by x = s(t). Suppose <strong>for</strong> simplicity that F is convex.<br />

Then F ′ is increasing and in order that the characteristic curves hit at C we must<br />

have u 1 > u 2 . Suppose Φ, Ψ satisfy Φ ′ F ′ = Ψ ′ .<br />

Claim: We claim that<br />

whenever Φ is convex.<br />

Ψ(u 1 ) − Ψ(u 2 ) ≥ F(u 1) − F(u 2 )<br />

u 1 − u 2<br />

(Φ(u 1 ) − Φ(u 2 ))<br />

60

Hurra! Ihre Datei wurde hochgeladen und ist bereit für die Veröffentlichung.

Erfolgreich gespeichert!

Leider ist etwas schief gelaufen!