17. Divergence, Curl and Potentials

f. Scalar Potentials

Given a vector field \(\vec F\), a scalar potential for \(\vec F\) is any scalar field \(f\) whose gradient is \(\vec F\): \[ \vec\nabla f=\vec F \]

1. Finding Scalar Potentials

If you can find a scalar potential by inspection, that's great. If not then you will need to do some work.

Find a scalar potential for \(\vec F=\left\langle yz,xz,xy\right\rangle\).

We need to find a function \(f\) satisfying \(\vec\nabla f=\vec F\). In more detail, \(f\) must satisfy \[ \partial_x f=yz \qquad \partial_y f=xz \qquad \partial_z f=xy \] , we see that a function that works is \(f=xyz\).

"By inspection" is a nebulous term. It actually means we took
  the \(x\)-antiderivative of \(\partial_x f=yz\),
  the \(y\)-antiderivative of \(\partial_y f=xz\) and
  the \(z\)-antiderivative of \(\partial_z f=xy\)
and got \(f=xyz\) in all three cases.

Find a scalar potential for \(\vec F=\left\langle 2x,2y,2z\right\rangle\).

\(f=x^2+y^2+z^2\)

We need to find a function \(f\) satisfying \(\vec\nabla f=\vec F\), or \[ \partial_x f=2x \qquad \partial_y f=2y \qquad \partial_z f=2z \] By , we see that a function that works is \(f=x^2+y^2+z^2\).

Actually, we take
  the \(x\)-antiderivative of \(\partial_x f=2x\) to get \(f(x,y,z)=x^2+g(y,z)\) where \(g\) is an arbitrary function which is independent of \(x\),
  the \(y\)-antiderivative of \(\partial_y f=2y\) to get \(f(x,y,z)=y^2+h(x,z)\) where \(h\) is an arbitrary function which is independent of \(y\), and
  the \(z\)-antiderivative of \(\partial_z f=2z\) to get \(f(x,y,z)=z^2+k(x,y)\) where \(k\) is an arbitrary function which is independent of \(z\).
Then \(f=x^2+y^2+z^2\) works in all three cases.

Caution: The scalar potential is NOT always just the sum of the \(x\)-antiderivative of \(F_1\), the \(y\)-antiderivative of \(F_2\) and the \(z\)-antiderivative of \(F_3\). A counter example happens in the next example and exercise which also show the process of finding all scalar potentials when inspection does not work.

Find all scalar potentials for: \[ \vec F =\left\langle \sin(y)-\sin(x)+\cos(z),x\cos(y)+e^z+e^y, y\,e^z-x\sin(z)+1\right\rangle \]

We need to find a function \(f\) satisfying \(\vec\nabla f=\vec F\) or: \[\begin{aligned} \partial_x f&=\sin(y)-\sin(x)+\cos(z)\quad&&\text{(1)} \\ \partial_y f&=x\cos(y)+e^z+e^y\quad&&\text{(2)} \\ \partial_z f&=y\,e^z-x\sin(z)+1\quad&&\text{(3)} \end{aligned}\] Since the scalar potential is not obvious by inspection, we solve these successively and build up a solution.

  1. We first take the \(x\)-antiderivative of (1) to get: \[ f=x\sin(y)+\cos(x)+x\cos(z)+g(y,z) \] where \(g\) is an arbitrary function which is independent of \(x\).
    (Note: When we take an antiderivative, we normally add a constant \(C\). However, since this is an \(x\)-antiderivative, the "constant", \(g\), only needs to be constant in \(x\), i.e. a function of \(y\) and \(z\).)
  2. Now this function \(f\) must also satisfy (2) and (3). We first take its \(y\)-derivative: \[ \partial_y f=x\cos(y)+\partial_y g \] Comparing this to (2) we find: \[ \partial_y g=e^z+e^y \] We take the \(y\)-antiderivative of this to get: \[ g=y\,e^z+e^y+h(z) \] where \(h\) is an arbitrary function of \(z\).
    (Note: Even though we are taking a \(y\)-antiderivative, we don't allow \(h\) to be a function of \(x\) as well as \(z\), because we already know \(g\) must be independent of \(x\).)
    Then we substitute this \(g\) into \(f\) to get: \[ f=x\sin(y)+\cos(x)+x\cos(z)+y\,e^z+e^y+h(z) \]
  3. Finally, this function \(f\) must also satisfy (3). We take its \(z\)-derivative: \[ \partial_z f=-x\sin(z)+y\,e^z+\dfrac{dh}{dz} \] Comparing this to (3) we find: \[ \dfrac{dh}{dz}=1 \] So: \[ h=z+C \] where \(C\) is really constant. Substituting this back into \(f\), we find the most general scalar potential is: \[ f=x\sin(y)+\cos(x)+x\cos(z)+y\,e^z+e^y+z+C \]

Find all scalar potentials for: \[ \vec F=\left\langle 2xy^3+z^3,3x^2y^2-3y^2z,3xz^2+4z^3-y^3\right\rangle \]

\(f=x^2y^3+xz^3-y^3z+z^4+C\)

We need to find a function \(f\) satisfying \(\vec\nabla f=\vec F\), or: \[\begin{aligned} \partial_x f&=2xy^3+z^3\quad&&\text{(1)} \\ \partial_y f&=3x^2y^2-3y^2z\quad&&\text{(2)} \\ \partial_z f&=3xz^2+4z^3-y^3\quad&&\text{(3)} \end{aligned}\] We solve these successively.

  1. We first take the \(x\)-antiderivative of (1) to get: \[ f=x^2y^3+xz^3+g(y,z) \]
  2. Next, we take the \(y\)-derivative of this: \[ \partial_y f=3x^2y^2+\partial_y g \] and compare it to (2) to get: \[ \partial_y g=-3y^2z \] We take the \(y\)-antiderivative of this to get: \[ g=-y^3z+h(z) \] and substitute this \(g\) into \(f\) to get: \[ f=x^2y^3+xz^3-y^3z+h(z) \]
  3. Finally, we take the \(z\)-derivative: \[ \partial_z f=3xz^2-y^3+\dfrac{dh}{dz} \] and compare it to (3) to get: \[ \dfrac{dh}{dz}=4z^3 \] So: \[ h=z^4+C \] Substituting this back into \(f\), we find the most general scalar potential is: \[ f=x^2y^3+xz^3-y^3z+z^4+C \]

We check by computing the gradient of the scalar potential: \[\begin{aligned} \text{grad}f &=\vec\nabla(x^2y^3+xz^3-y^3z+z^4) \\ &=\left\langle 2xy^3+z^3,3x^2y^2-3y^2z,3xz^2-y^3+4z^3\right\rangle \end{aligned}\] which is the vector field we started with.

Non-Uniqueness of the Scalar Potential

The previous example and exercise demonstrate the following:

If \(f\) and \(g\) are both scalar potentials for a vector field \(\vec F\), then \(f=g+C\) for some constant \(C\).

Since \(\vec F=\vec\nabla f=\vec\nabla g\), we have \(\vec\nabla(f-g)=0\). So the \(x\), \(y\) and \(z\) partial derivatives of \(f-g\) are all \(0\). This says \(f-g\) must be a constant \(C\).

© MYMathApps

Supported in part by NSF Grant #1123255