14. Directional Derivatives and Gradients

c.2. Applications of Directional Derivatives

The temperature in a room is given by \(T=(275+\sin(\pi x)\cos(\pi y)+\sin(\pi z))^\circ K\). Find the rate of change of \(T\) at \((2,-1,2)\) in the direction toward \((3,-3,4)\). Assume \(x\), \(y\) and \(z\) are in meters and \(T\) is in \(^\circ\text{K}\).

We need to find the rate of change of the temperature in the direction of the vector from \(P=(2,-1,2)\) to \(Q=(3,-3,4)\). This is a directional derivative. The vector from \(P\) to \(Q\) is: \[ \overrightarrow{PQ}=\langle 1,-2,2\rangle \] Its length is \(\left|\overrightarrow{PQ}\right|=\sqrt{1+4+4}=3\) and the unit vector is: \[ \widehat{PQ} =\left\langle \dfrac{1}{3},\dfrac{-2}{3},\dfrac{2}{3}\right\rangle \] The gradient of the temperature is: \[\begin{aligned} \vec\nabla T &=\left\langle \dfrac{\partial T}{\partial x}, \dfrac{\partial T}{\partial y}, \dfrac{\partial T}{\partial z}\right\rangle \\ &=\langle \pi\cos(\pi x)\cos(\pi y), -\pi\sin(\pi x)\sin(\pi y), \pi\cos(\pi z)\rangle \end{aligned}\] At \(P=(2,-1,2)\) this is \[\begin{aligned} \left.\vec\nabla T\right|_P &=\langle \pi\cos(2\pi)\cos(-\pi), -\pi\sin(2\pi)\sin(-\pi), \pi\cos(2\pi)\rangle \\ &=\langle-\pi,0,\pi\rangle \end{aligned}\] So the directional derivative is: \[\begin{aligned} \nabla_{\widehat{PQ}}T &=\widehat{PQ}\cdot\left.\vec\nabla T\right|_P =\left\langle \dfrac{1}{3},\dfrac{-2}{3},\dfrac{2}{3}\right\rangle \cdot\langle-\pi,0,\pi\rangle \\ &=-\,\dfrac{\pi}{3}+0+\dfrac{2\pi}{3}=\dfrac{\pi}{3} \end{aligned}\] The derivative \(\dfrac{\partial T}{\partial x}\) has units \(\dfrac{^\circ\text{K}}{\text{m}}\). So those are the units of the gradient. The displacement vector, \(\overrightarrow{PQ}\), and its length \(|\overrightarrow{PQ}|\), have units of \(\text{m}\). So the direction vector, \(\widehat{PQ}\), is dimensionless. So the directional derivative, \(\widehat{PQ}\cdot\left.\vec\nabla T\right|_P\), has units \(\dfrac{^\circ\text{K}}{\text{m}}\). So the answer is \[ \nabla_{\widehat{PQ}}T=\dfrac{\pi}{3}\,\dfrac{^\circ\text{K}}{\text{m}} \]

Find the rate of change of the temperature \(T=xy^2+yz^2\) at the point \(P=(3,1,2)\) if we move toward the point \(Q=(1,3,3)\) with speed \(6\,\dfrac{\text{m}}{\text{sec}}\). Assume \(x\), \(y\) and \(z\) are in meters and \(T\) is in \(^\circ\text{K}\).

Get the velocity by combining the speed with the direction.

\(\nabla_{\vec v}T=44\,\dfrac{^\circ\text{K}}{\text{sec}}\)

The vector from \(P\) to \(Q\) is \(\overrightarrow{PQ}=Q-P=\langle -2,2,1\rangle\). Its length is \(|\overrightarrow{PQ}|=\sqrt{4+4+1}=3\). So its direction is: \[ \hat v=\widehat{PQ} =\left\langle \dfrac{-2}{3},\dfrac{2}{3},\dfrac{1}{3}\right\rangle \] The speed is \(|\vec v|=6\). So the velocity is \[ \vec v=|\vec v|\hat v =\langle -4,4,2\rangle \] The gradient of \(T\) is \(\vec\nabla T=\langle y^2,2xy+z^2,2yz\rangle\). At the point \(P=(3,1,2)\), this is: \[ \left.\vec\nabla T\right|_P =\langle 1,10,4\rangle \] So the derivative of \(T\) with respect to \(\vec v\) is: \[\begin{aligned} \nabla_{\vec v}T &=\vec v\cdot\left.\vec\nabla T\right|_P =\langle -4,4,2\rangle\cdot\langle 1,10,4\rangle \\ &-4+40+8=44\,\dfrac{^\circ\text{K}}{\text{sec}} \end{aligned}\] Alternatively, we can find the directional derivative of \(T\) along \(\hat v\) which is: \[\begin{aligned} \nabla_{\hat v}T &=\hat v\cdot\left.\vec\nabla T\right|_P =\left\langle \dfrac{-2}{3}, \dfrac{2}{3},\dfrac{1}{3}\right\rangle\cdot\langle 1,10,4\rangle \\ &=\dfrac{-2}{3}+\dfrac{20}{3}+\dfrac{4}{3} =\dfrac{22}{3}\,\dfrac{^\circ\text{K}}{\text{m}} \end{aligned}\] and multiply by the speed to get: \[ \nabla_{\vec v}T =|\vec v|\nabla_{\hat v}T =6\,\dfrac{\text{m}}{\text{sec}} \cdot\dfrac{22}{3}\,\dfrac{^\circ\text{K}}{\text{m}} =44\,\dfrac{^\circ\text{K}}{\text{sec}} \]

It is interesting to note the difference in units between the answers to the example and the exercise. The directional derivative in the example gives rate of change with respect to distance while the derivative along a vector in the exercise gives the rate of change with respect to time.

© MYMathApps

Supported in part by NSF Grant #1123255