Question 5.31: (a) Complete the proof of Theorem 2, Sect. 1.6.2. That is, s...

(a) Complete the proof of Theorem 2, Sect. 1.6.2. That is, show that any divergenceless vector field F can be written as the curl of a vector potential A. What you have to do is find A_{x}, A_{y}, \text { and } A_{z} \text { such that (i) } \partial A_{z} / \partial y-\partial A_{y} / \partial z=F_{x} ; \text { (ii) } \partial A_{x} / \partial z-\partial A_{z} / \partial x=F_{y} ; \text { and (iii) } \partial A_{y} / \partial x-\partial A_{x} / \partial y=F_{z} . Here’s one way to do it: Pick A_{x}=0 \text {, and solve (ii) and (iii) for } A_{y} \text { and } A_{z} . Note that the “constants of integration” are themselves functions of y and z—they’re constant only with respect to x. Now plug these expressions into (i), and use the fact that · F = 0 to obtain 

A_{y}=\int_{0}^{x} F_{z}\left(x^{\prime}, y, z\right) d x^{\prime} ; \quad A_{z}=\int_{0}^{y} F_{x}\left(0, y^{\prime}, z\right) d y^{\prime}-\int_{0}^{x} F_{y}\left(x^{\prime}, y, z\right) d x^{\prime} .

(b) By direct differentiation, check that the A you obtained in part (a) satisfies × A = F. Is A divergenceless? [This was a very asymmetrical construction, and it would be surprising if it were—although we know that there exists a vector whose curl is F and whose divergence is zero.]

(c) As an example,  F =y \hat{ x }+z \hat{ y }+x \hat{ z } Calculate A, and confirm that × A = F. (For further discussion, see Prob. 5.53.)

The Blue Check Mark means that this solution has been answered and checked by an expert. This guarantees that the final answer is accurate.
Learn more on how we answer questions.
\text { (a) }\left\{\begin{array}{c}-\frac{\partial W_{z}}{\partial x}=F_{y} \Rightarrow W_{z}(x, y, z)=-\int_{0}^{x} F_{y}\left(x^{\prime}, y, z\right) d x^{\prime}+C_{1}(y, z) \\\frac{\partial W_{y}}{\partial x}=F_{z} \Rightarrow W_{y}(x, y, z)=+\int_{0}^{x}F_{z}\left(x^{\prime}, y, z\right) d x^{\prime}+C_{2}(y, z)\end{array}\right\}

 

These satisfy (ii) and (iii), for any C_{1} \text { and } C_{2}; it remains to choose these functions so as to satisfy (i):

-\int_{0}^{x} \frac{\partial F_{y}\left(x^{\prime}, y, z\right)}{\partial y} d x^{\prime}+\frac{\partial C_{1}}{\partial y}-\int_{0}^{x} \frac{\partial F_{z}\left(x^{\prime}, y, z\right)}{\partial z} d x^{\prime}-\frac{\partial C_{2}}{\partial z}=F_{x}(x, y, z) . \text { But } \frac{\partial F_{x}}{\partial x}+\frac{\partial F_{y}}{\partial y}+\frac{\partial F_{z}}{\partial z}=0 , so

 

\int_{0}^{x} \frac{\partial F_{x}\left(x^{\prime}, y, z\right)}{\partial x^{\prime}} d x^{\prime}+\frac{\partial C_{1}}{\partial y}-\frac{\partial C_{2}}{\partial z}=F_{x}(x, y, z) . \quad \text { Now } \int_{0}^{x} \frac{\partial F_{x}\left(x^{\prime}, y, z\right)}{\partial x^{\prime}} d x^{\prime}=F_{x}(x, y, z)-F_{x}(0, y, z) , so

 

\frac{\partial C_{1}}{\partial y}-\frac{\partial C_{2}}{\partial z}=F_{x}(0, y, z) \text {. We may as well pick } C_{2}=0, C_{1}(y, z)=\int_{0}^{y} F_{x}\left(0, y^{\prime}, z\right) d y^{\prime} \text {, and we're done, with }

 

W_{x}=0 ; \quad W_{y}=\int_{0}^{x} F_{z}\left(x^{\prime}, y, z\right) d x^{\prime} ; \quad W_{z}=\int_{0}^{y} F_{x}\left(0, y^{\prime}, z\right) d y^{\prime}-\int_{0}^{x} F_{y}\left(x^{\prime}, y, z\right) d x^{\prime} .

 

\text { (b) } \nabla \times W =\left(\frac{\partial W_{z}}{\partial y}-\frac{\partial W_{y}}{\partial z}\right) \hat{ x }+\left(\frac{\partial W_{x}}{\partial z}-\frac{\partial W_{z}}{\partial x}\right) \hat{ y }+\left(\frac{\partial W_{y}}{\partial x}-\frac{\partial W_{x}}{\partial y}\right) \hat{ z }

 

=\left[F_{x}(0, y, z)-\int_{0}^{x} \frac{\partial F_{y}\left(x^{\prime}, y, z\right)}{\partial y} d x^{\prime}-\int_{0}^{x} \frac{\partial F_{z}\left(x^{\prime}, y, z\right)}{\partial z} d x^{\prime}\right] \hat{ x }+\left[0+F_{y}(x, y, z)\right] \hat{ y }+\left[F_{z}(x, y, z)-0\right] \hat{ z } .

 

\text { But } \nabla \cdot F =0, \text { so the } \hat{ x } \text { term is }\left[F_{x}(0, y, z)+\int_{0}^{x} \frac{\partial F_{x}\left(x^{\prime}, y, z\right)}{\partial x^{\prime}} d x^{\prime}\right]=F_{x}(0, y, z)+F_{x}(x, y, z)-F_{x}(0, y, z) ,

 

\text { so } \nabla \times W = F .

 

\nabla \cdot W =\frac{\partial W_{x}}{\partial x}+\frac{\partial W_{y}}{\partial y}+\frac{\partial W_{z}}{\partial z}=0+\int_{0}^{x} \frac{\partial F_{z}\left(x^{\prime}, y, z\right)}{\partial y} d x^{\prime}+\int_{0}^{y} \frac{\partial F_{x}\left(0, y^{\prime}, z\right)}{\partial z} d y^{\prime}-\int_{0}^{x} \frac{\partial F_{y}\left(x^{\prime}, y, z\right)}{\partial z} d x^{\prime} \neq 0 ,

in general.

\text { (c) } W_{y}=\int_{0}^{x} x^{\prime} d x^{\prime}=\frac{x^{2}}{2} ; W_{z}=\int_{0}^{y} y^{\prime} d y^{\prime}-\int_{0}^{x} z d x^{\prime}=\frac{y^{2}}{2}-z x .

 

W =\frac{x^{2}}{2} \hat{ y }+\left(\frac{y^{2}}{2}-z x\right) \hat{ z } . \nabla \times W =\left|\begin{array}{ccc}\hat{ x } & \hat{ y } & \hat{ z } \\\partial / \partial x & \partial / \partial y & \partial / \partial z \\0 & x^{2} / 2 & \left(y^{2} / 2-z x\right)\end{array}\right|=y \hat{ x }+z \hat{ y }+x \hat{ z }= F .

Related Answered Questions