02.12.2012 Views

Quasilinear parabolic problems with nonlinear boundary conditions

Quasilinear parabolic problems with nonlinear boundary conditions

Quasilinear parabolic problems with nonlinear boundary conditions

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Combining (5.38) <strong>with</strong> (5.39), setting<br />

and using (5.37) leads to<br />

q0 := 1<br />

2A(1 ∗ gw) − 1 2<br />

2A(b − 3a) ∗ γp1 + γ∂yw1<br />

q0 = p0 + 1 2<br />

2A(b − 3a) ∗ p0 + γ∇x · vp<br />

= Aa ∗ p0+ 1<br />

2<br />

= 1<br />

2<br />

= 1<br />

2<br />

Further,<br />

A(b + 4<br />

3<br />

A(b − 2<br />

3<br />

a) ∗ p0−∆xF −2<br />

1<br />

4 A(b + 3a) ∗ p0 − ∆xF −2<br />

1<br />

(5.40)<br />

−2<br />

−2 1<br />

a) ∗ p0+∆xF 1 p0−∆xF 1 A(b + 3a) ∗ (F −G)(F +G)−1p0 �<br />

1<br />

−Aa ∗ (F +G) + A(b +<br />

� A(b − 2<br />

3<br />

3 a) ∗ (F −G)� (F +G) −1 p0<br />

a) ∗ F − A(b + 4<br />

3 a) ∗ G� (F + G) −1 p0.<br />

2Ka ∗ q0 = � I + 2DnF −2<br />

1 [K(b − 2<br />

3 a) ∗ F − G](F + G)−1� p0<br />

= � (A + 2Dn + 2DnK(b − 2<br />

3 a)∗)F + AG� (F + G) −1 F −2<br />

1 p0<br />

= � A + [2Dn + 2DnK(b − 2<br />

3a)∗]F (F + G)−1� F −2<br />

1 p0<br />

= � A + [2Dn(K(b + 4<br />

3a)∗) + 2DnK(b − 2<br />

3a)∗]F (F + G)−1� F −2<br />

1 p0<br />

= � A + DnK(4b + 4<br />

3a) ∗ F (F + G)−1� F −2<br />

1 p0 = Lp0, (5.41)<br />

where the operator L is defined by<br />

L = � A + DnK(4b + 4<br />

3 a) ∗ F (F + G)−1� F −2<br />

1 . (5.42)<br />

We claim now that q0 ∈ 0Y . Indeed, in virtue of (5.18),(5.19), there exist ψ1, ψ2 ∈ Y<br />

such that<br />

and<br />

q0 = 1<br />

2 (ψ1 + Ab ∗ ψ2) − 1 2<br />

2A(b − 3a) ∗ γp1 + γ∂yw1<br />

= 1<br />

2 (ψ1 + A(b − 2<br />

3a) ∗ ψ2) + 1<br />

3ψ2 − 1 2<br />

2A(b − 3a) ∗ γp1 + γ∂yw1<br />

= 1 2<br />

2A(b − 3a) ∗ (ψ2 − γp1) + ( 1<br />

2ψ1 + 1<br />

3ψ2 + γ∂yw1),<br />

ψ1|t=0 = 2<br />

3 γ∇x · v0 − 4<br />

3 γ∂yw0, ψ2|t=0 = −γ∇x · v0 − γ∂yw0 (5.43)<br />

in case p > 1 + 2/δa. But from (5.43) and the definition of p1 and w1, we deduce that<br />

ψ2 − γp1 , 1<br />

2 ψ1 + 1<br />

3 ψ2 + γ∂yw1 ∈ 0Y.<br />

Hence the claim follows, because A(b∗) ∈ B(0Y ).<br />

From q0 ∈ 0Y and K(a∗) ∈ B(0Y ) we conclude further that K(a ∗ q0) ∈ 0Y . That is,<br />

to solve (5.41) for p0, we have to show that L has a bounded inverse on 0Y . To achieve<br />

this, we shall use, aside from extension and restriction, the joint (causal) H∞ (Σ π<br />

2 +η ×Ση)<br />

- calculus (0 < η < π/2) of the pair (∂t, Dn) in Lp(R+ × Rn ), cf. Example 2.4.1.<br />

For this purpose we look at the symbol l(z, ξ) of L (in Lp(R+ × Rn )). Taking the<br />

Laplace-transform in t and the Fourier-transform in x we obtain for l(z, ξ) :<br />

⎛<br />

⎜<br />

l(z, ξ) = ⎝<br />

1<br />

+<br />

â(z)|ξ| 2<br />

� 1<br />

â(z)|ξ| 2 + 1<br />

4ˆb(z)+ 4<br />

3 â(z)<br />

ˆ 4<br />

b(z)+ 3 â(z)<br />

�<br />

1<br />

â(z)|ξ| 2 + 1 + �<br />

1<br />

( ˆb(z)+ 4 + 1<br />

â(z))|ξ|2 3<br />

88<br />

⎞<br />

�<br />

�−1 ⎟ 1<br />

⎠ + 2 . (5.44)<br />

â(z)|ξ| 2

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!