调和方程(拉普拉斯方程)基本解和边界元方法的积分计算

前言

调和方程的基本解以及边界元方法中某积分的解析结果。

调和方程

u ( x 1 , . . . , x n ) = f ( r ) u(x_1, ..., x_n) = f(r) u(x1,...,xn)=f(r) (其中 r = x 1 2 + . . . + x n 2 r = \sqrt{x_1^2 + ... + x_n^2} r=x12+...+xn2 )是 n n n 维调和函数 (即满足方程 ∂ 2 u ∂ x 1 2 + . . . , + ∂ 2 u ∂ x n 2 = 0 \frac{\partial^2u}{\partial{x_1^2}} + ..., + \frac{\partial^2u}{\partial{x_n^2}} = 0 x122u+...,+xn22u=0),证明如下等式成立
f ( r ) = c 1 + c 2 r n − 2          ( n ≠ 2 ) f(r) = c_1 + \frac{c_2}{r^{n-2}} \ \ \ \ \ \ \ \ (n \neq 2) f(r)=c1+rn2c2        (n=2)
f ( r ) = c 1 + c 2 l n 1 r        ( n = 2 ) f(r) = c_1 + c_2ln\frac{1}{r} \ \ \ \ \ \ (n = 2) f(r)=c1+c2lnr1      (n=2)
其中 c 1 , c 2 c_1, c_2 c1,c2 为任意常数。
证明:
u = f ( r ) ,     ∂ u ∂ x i = f ′ ( r ) ⋅ ∂ r ∂ x i = f ′ ( r ) ⋅ x i r u = f(r), \ \ \ \frac{\partial{u}}{\partial{x_i}} = f^{'}(r) \cdot \frac{\partial r}{\partial x_i} = f^{'}(r) \cdot \frac{x_i}{r} u=f(r),   xiu=f(r)xir=f(r)rxi

∂ 2 u ∂ x i 2 = f ′ ′ ( r ) ⋅ x i 2 r 2 + f ′ ( r ) ⋅ 1 r − f ′ ( r ) ⋅ x i 2 r 3 \frac{\partial^2{u}}{\partial{x^2_i}} = f^{''}(r) \cdot \frac{x^2_i}{r^2} + f^{'}(r) \cdot \frac{1}{r} - f^{'}(r) \cdot \frac{x^2_i}{r^3} xi22u=f(r)r2xi2+f(r)r1f(r)r3xi2

∑ i = 1 n ∂ 2 u ∂ x i 2 = f ′ ′ ( r ) ⋅ ∑ i = 1 n x i 2 r 2 + f ′ ( r ) ⋅ n r − f ′ ( r ) ⋅ ∑ i = 1 n x i 2 r 3 = f ′ ′ ( r ) + n − 1 r f ′ ( r ) \sum^{n}_{i = 1} \frac{\partial^2u}{\partial{x^2_i}} = f^{''}(r) \cdot \frac{\sum^{n}_{i = 1}{x^2_i}}{r^2} + f^{'}(r) \cdot \frac{n}{r} - f^{'}(r) \cdot \frac{\sum^{n}_{i = 1}{x^2_i}}{r^3} = f^{''}(r) + \frac{n-1}{r} f^{'}(r) i=1nxi22u=f(r)r2i=1nxi2+f(r)rnf(r)r3i=1nxi2=f(r)+rn1f(r)
即方程 Δ u = 0 \Delta u = 0 Δu=0 化为 f ′ ′ ( r ) + n − 1 r f ′ ( r ) = 0 f^{''}(r) + \frac{n-1}{r} f^{'}(r) = 0 f(r)+rn1f(r)=0

f ′ ′ ( r ) f ′ ( r ) = − n − 1 r            ( 1 ) \frac{f^{''}(r)}{f^{'}(r)} = -\frac{n-1}{r} \ \ \ \ \ \ \ \ \ \ (1) f(r)f(r)=rn1          (1)

所以

f ′ ( r ) = A 1 r − ( n − 1 )            ( 2 ) f^{'}(r) = A_1r^{-(n-1)} \ \ \ \ \ \ \ \ \ \ (2) f(r)=A1r(n1)          (2)

n ≠ 2 n \neq 2 n=2,积分得

f ( r ) = A 1 − n + 2 r − n + 2 + c 1 f(r) = \frac{A_1}{-n+2} r^{-n+2} + c_1 f(r)=n+2A1rn+2+c1

n ≠ 2 n \neq 2 n=2, 则 f ( r ) = c 1 + c 2 r n − 2 f(r) = c_1 + \frac{c_2}{r^{n-2}} f(r)=c1+rn2c2

n = 2 n = 2 n=2,则 f ′ ( r ) = A 1 r f^{'}(r) = \frac{A_1}{r} f(r)=rA1 f ( r ) = c 1 + A 1 l n r f(r) = c_1 + A_1lnr f(r)=c1+A1lnr

n = 2 n = 2 n=2,则 f ( r ) = c 1 + c 2 l n 1 r f(r) = c_1 + c_2ln\frac{1}{r} f(r)=c1+c2lnr1

上面的证明中 (1) 到 (2) 的推导如下:
可令 z = f ′ ( r ) z = f^{'}(r) z=f(r), 则等式 (1) 变成 z ′ z = − n − 1 r \frac{z^{'}}{z} = -\frac{n-1}{r} zz=rn1

下面就类似求常微分方程 d y d x = P ( x ) y \frac{dy}{dx} = P(x)y dxdy=P(x)y 的通解。
变量分离法:
d y y = P ( x ) d x \frac{dy}{y} = P(x) dx ydy=P(x)dx
两边积分,即得
l n ∣ y ∣ = ∫ P ( x ) d x + c 1 ln \left | y \right | = \int P(x) dx + c_1 lny=P(x)dx+c1
这里 c 1 c_1 c1 是任意常数,由对数定义,即有
∣ y ∣ = e ∫ P ( x ) d x + c 1 \left | y \right | = e^{\int P(x)dx + c_1} y=eP(x)dx+c1

y = ± e c 1 ⋅ e ∫ P ( x ) d x y = \pm e^{c_1} \cdot e^{\int P(x)dx} y=±ec1eP(x)dx
± e c 1 = c \pm e^{c_1} = c ±ec1=c 得到
y = c e ∫ P ( x ) d x y = ce^{\int P(x)dx} y=ceP(x)dx

边界元方法的基本知识

二维空间的拉普拉斯方程的基本解为

u s ∗ ( r ) = − 1 2 π l n r u^*_s(r) = -\frac{1}{2\pi}lnr us(r)=2π1lnr

q s ∗ ( r ) = ∂ u s ∗ ∂ n ⃗ = − ( r ⃗ , n ⃗ ) 2 π r 2 q^*_s(r) = \frac{\partial u^{*}_s}{\partial {\vec{n}}} = -\frac{(\vec{r}, \vec{n})}{2\pi r^2} qs(r)=n us=2πr2(r ,n )

三维空间的拉普拉斯方程的基本解为

u s ∗ ( r ) = 1 4 π r u^*_s(r) = \frac{1}{4\pi r} us(r)=4πr1

q s ∗ ( r ) = ∂ u s ∗ ∂ n ⃗ = − ( r ⃗ , n ⃗ ) 4 π r 3 q^*_s(r) = \frac{\partial u^{*}_s}{\partial {\vec{n}}} = -\frac{(\vec{r}, \vec{n})}{4\pi r^3} qs(r)=n us=4πr3(r ,n )

积分计算

如下图所示计算积分

I = ∫ L r ⃗ ⋅ n ⃗ r 2 d l I = \int_{L} \frac{\vec{r} \cdot \vec{n}}{r^2} dl I=Lr2r n dl

其中 r ⃗ = ( x , y ) ,       L = A B ‾ \vec{r} = (x, y), \ \ \ \ \ L= \overline{AB} r =(x,y),     L=AB

经计算可得
I = β ,         其 中 β 为 线 段 O A 与 线 段 O B 的 夹 角 的 弧 度 值 I = \beta, \ \ \ \ \ \ \ 其中 \beta 为 线段OA 与线段 OB的夹角的弧度值 I=β,       β线OA线OB

在这里插入图片描述
注:可用如下网址验证 Integral Calculator
在这里插入图片描述

  • 1
    点赞
  • 10
    收藏
    觉得还不错? 一键收藏
  • 0
    评论
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值