= Solution
Write $q(x)=1/[x(1+(\ln x)^2)]$. The <vector field> is undefined at $x=0$ because $q$ diverges there, but this component is tangential to that face. The face fluxes have finite <improper integral> limits. In particular, the substitution $s=\ln x$ gives
$$
J=\int_0^1q(x)\,dx=\int_{-\infty}^0\frac{ds}{1+s^2}=\frac\pi2.
$$
Using the outward <unit normal> on each face, the <flux> contributions are
$$
\boxed{\begin{array}{c|c|c}
\text{face}&\mathbf F\cdot\hat{\mathbf n}&\text{outward flux}\\\hline
x=0&-1\text{ (normal trace)}&-1\\
x=1&e^{-1}&e^{-1}\\
y=0&-q(x)&-\pi/2\\
y=1&q(x)&\pi/2\\
z=0&0&0\\
z=1&1&1
\end{array}}
$$
Hence
$$
\boxed{\text{total outward flux}=e^{-1}}.
$$
For a precise justification at the singular face, first use the truncated cube $\varepsilon\leq x\leq1$. Its opposite $y$-face integrals cancel, the $x$-faces contribute $e^{-1}-e^{-\varepsilon}$, and the upper $z$-face contributes $1-\varepsilon$. The limit as $\varepsilon\downarrow0$ is $e^{-1}$. On each truncated cube the <divergence theorem> gives the same answer, since $\nabla\cdot\mathbf F=1-e^{-x}$. Thus the result is a <flux of a vector field with an integrable tangential singularity>, rather than an application of the classical smooth-field theorem directly at $x=0$.
Back to article page
