Problems
  1. Prove in the sense of distributions
    limε0επ(x2+ε2)=δ0.\lim_{\varepsilon \to 0} \frac{\varepsilon}{\pi (x^2 + \varepsilon^2)} = \delta_0.

  2. For any ε>0\varepsilon > 0, 1x+iε\frac{1}{x+i\varepsilon} is a locally integrable function on R\mathbb{R}. As a distribution on R\mathbb{R}, prove
    limε0+1x+iε=Dvp1xiπδ0.\lim_{\varepsilon \to 0^+} \frac{1}{x+i\varepsilon} \overset{\mathcal{D}'}{=} \text{vp} \frac{1}{x} - i\pi \delta_0.
    The limit distribution can also be written as 1x+i0\frac{1}{x+i0}.

  3. As a distribution on R\mathbb{R}, un(x)=x1sin(nx)u_n(x) = x^{-1} \sin(nx), compute the limit in the distribution space D(R)\mathcal{D}'(\mathbb{R}) of limnun\lim_{n \to \infty} u_n.

  4. We denote the surface area of the unit sphere Sn1|\mathbf S^{n-1}| in Rn\mathbb{R}^n by Sn1\mathbf S^{n-1}. Let
    E=x2n(2n)Sn1.E = \frac{|x|^{2-n}}{(2-n)|\mathbf S^{n-1}|}.
    Prove in the sense of distributions that
    ΔE=Dδ0.\Delta E \overset{\mathcal{D}'}{=} \delta_0.

  5. Assume ΩR3\Omega \subset \mathbb{R}^3 is an open set and uD(Ω)u \in \mathcal{D}'(\Omega), satisfying
    Δu=f,\Delta u = f,
    where fC(Ω)f \in C^\infty(\Omega) is a smooth function. Hence, uC(Ω)u \in C^\infty(\Omega).

  6. Suppose fE(R3)f \in \mathcal{E}'(\mathbb{R}^3) (a distribution with compact support). Prove the existence of a constant C1>0C_1 > 0 (which may depend on the distribution ff), such that, as x|x| \to \infty,
    Ef(x)Cx.|E * f(x)| \leq \frac{C}{|x|}.

  7. Assume fE(R3)f \in \mathcal{E}'(\mathbb{R}^3). Prove that u=Efu = E * f is a solution to the potential equation
    Δu=f.\Delta u = f.
    Furthermore, a constant C2>0C_2 > 0 exists such that, as x|x| \to \infty,
    u(x)f,14πxC2x2.\left| u(x) - \frac{\langle f, 1 \rangle}{4\pi |x|} \right| \leq \frac{C_2}{|x|^2}.
    Depending on the physical context, f,1\langle f, 1 \rangle represents the total mass or total charge of uu.

  8. Suppose uC(R3)u \in C^\infty(\mathbb{R}^3) is a real-valued function, and define the sign function sign(u)\operatorname{sign}(u) as
    sign(u)(x)={1,u(x)>0;0,u(x)=0;1,u(x)<0.\operatorname{sign}(u)(x) = \begin{cases} 1, & u(x) > 0; \\ 0, & u(x) = 0; \\ -1, & u(x) < 0. \end{cases}
    Prove, in the sense of distributions that
    ΔuDΔusign(u),\Delta |u| \overset{\mathcal{D}'}{\geq} \Delta u \cdot \operatorname{sign}(u),
    i.e., for any non-negative real-valued function φD(R3)\varphi \in \mathcal{D}(\mathbb{R}^3),
    Δu,φΔusign(u),φ,\langle \Delta |u|, \varphi \rangle \geq \langle \Delta u \cdot \operatorname{sign}(u), \varphi \rangle,
    Hint: First approximate uu in the sense of distributions with u2+ε2\sqrt{u^2 + \varepsilon^2}.