Taylor expansion in proof of weak maximum principle

110 Views Asked by At

Picture below is part of proof of weak maximum principle. On the red line ,I don't know how to use the Taylor expansion to get $-u''(x_0) \le 0$. I think the Taylor expansion of $u(x)$ at $x_0$ is $$ u(x)=u(x_0)+u'(x_0)(x-x_0)+\frac{u''(x_0)}{2}(x-x_0)^2+o((x-x_0)^2) $$ Let $x\in \partial \Omega$ and $u(x)=m$, then $$ m=u(x_0)+\frac{u''(x_0)}{2}(x-x_0)^2+o((x-x_0)^2) $$ So, $$ 0> \frac{u''(x_0)}{2}(x-x_0)^2+o((x-x_0)^2) $$ then $$ u''(x_0) > 2\frac{o((x-x_0)^2)}{(x-x_0)^2} $$ But the $x$ has been choice , how to show $-u''(x_0) \le 0$ ?

Picture below is from the 7th page of this .

enter image description here