Calculus III: Partial Derivatives

In engineering and physics, most real-world systems depend on more than one variable—from the stress on a bridge beam to the temperature distribution in a heat sink. Partial derivatives provide the mathematical toolkit to analyze how these systems change with respect to each individual variable, holding others constant. Mastering this concept is essential for modeling, optimizing, and solving multidimensional problems you will encounter in fields like thermodynamics, fluid mechanics, and machine learning.

Foundations: Defining and Notating Partial Derivatives

A partial derivative of a function of multiple variables measures its instantaneous rate of change with respect to one variable, while all other independent variables are held fixed. Consider a function $z=f(x,y)$. The partial derivative of $f$ with respect to $x$ is denoted by $\frac{\partial f}{\partial x}$, $f_x$, or ${\partial }_{x}f$. The notation $\frac{\partial }{\partial x}$ is the partial derivative operator, distinguishing it from the ordinary derivative $\frac{d}{dx}$.

To compute $\frac{\partial f}{\partial x}$, you treat $y$ as a constant and differentiate with respect to $x$ using standard single-variable calculus rules. For example, if $f(x,y)=3x^{2}y+\sin (y)$, then $\frac{\partial f}{\partial x}=6xy$. Here, $\sin (y)$ is treated as a constant when differentiating with respect to $x$, so its derivative is zero. Similarly, $\frac{\partial f}{\partial y}=3x^2+\cos (y)$, where $3x^2$ is treated as a constant coefficient with respect to $y$. This process extends to functions with any number of variables; for $f(x_1,x_2,...,x_n)$, $\frac{\partial f}{\partial x_i}$ is found by differentiating with respect to $x_i$ while treating all other $x_j$ (where $j\ne i$) as constants.

Visualizing Change: The Geometric Interpretation

The geometric interpretation of a partial derivative is the slope of a tangent line to a cross-section of the function's surface. Imagine the surface defined by $z=f(x,y)$. If you slice this surface with a vertical plane where $y$ is held constant at $y=c$, you create a curve: $z=f(x,c)$. The partial derivative $\frac{\partial f}{\partial x}$ evaluated at a point $(a,c)$ is precisely the slope of the tangent line to this curve at $x=a$. In essence, it tells you how steeply the surface rises or falls as you move in the $x$-direction, while your $y$-position is locked.

Conversely, $\frac{\partial f}{\partial y}$ gives the slope of the tangent line to the cross-section formed by holding $x$ constant. This interpretation is powerful for engineering design. For instance, if $f(x,y)$ models the elevation of a terrain, $f_x$ indicates the slope in the east-west direction, crucial for assessing landslide risk or planning road gradients. Visualizing these cross-sections helps you intuit how multivariable functions behave locally, forming the basis for more advanced concepts like the gradient and directional derivatives.

Higher-Order Partials and Clairaut's Theorem

Just as with ordinary derivatives, you can take partial derivatives of partial derivatives, known as higher-order partial derivatives. For $f(x,y)$, the second partial derivatives include the direct second derivatives, $\frac{\partial }{\partial x}(\frac{\partial f}{\partial x})=\frac{{\partial }^{2}f}{\partial x^2}=f_{xx}$ and $\frac{{\partial }^{2}f}{\partial y^2}=f_{yy}$, as well as the mixed partial derivatives, $\frac{\partial }{\partial y}(\frac{\partial f}{\partial x})=\frac{{\partial }^{2}f}{\partial y\partial x}=f_{xy}$ and $\frac{\partial }{\partial x}(\frac{\partial f}{\partial y})=\frac{{\partial }^{2}f}{\partial x\partial y}=f_{yx}$.

Clairaut's theorem (also called Schwarz's theorem) states that if the mixed partial derivatives $f_{xy}$ and $f_{yx}$ are continuous on an open region, then they are equal at every point in that region: $f_{xy}=f_{yx}$. This theorem simplifies calculations and is frequently invoked in engineering analysis where continuity is assured. For example, consider $f(x,y)=x^3{y}^2+e^{xy}$. First, compute $f_x=3x^2{y}^2+y{e}^{xy}$ and then $f_{xy}=6x^{2}y+e^{xy}+xy{e}^{xy}$. Alternatively, compute $f_y=2x^{3}y+x{e}^{xy}$ and then $f_{yx}=6x^{2}y+e^{xy}+xy{e}^{xy}$. The results are identical, confirming Clairaut's theorem. This property is foundational for solving partial differential equations that model phenomena like wave propagation or heat diffusion.

Implicit Partial Differentiation

Often in engineering, relationships between variables are defined implicitly by an equation like $F(x,y,z)=0$, rather than an explicit function $z=f(x,y)$. Implicit partial differentiation allows you to find partial derivatives such as $\frac{\partial z}{\partial x}$ without explicitly solving for $z$. You differentiate the entire equation with respect to the variable of interest, applying the chain rule and treating other variables as functions where necessary.

Suppose an equation defines $z$ implicitly as a function of $x$ and $y$: $x^{2}z+y\ln (z)=xyz$. To find $\frac{\partial z}{\partial x}$, differentiate both sides with respect to $x$, treating $y$ as constant and $z$ as $z(x,y)$: $$\frac{\partial }{\partial x}(x^{2}z)+\frac{\partial }{\partial x}(y\ln (z))=\frac{\partial }{\partial x}(xyz).$$ Applying the product and chain rules: $$2xz+x^2\frac{\partial z}{\partial x}+y\cdot \frac{1}{z}\frac{\partial z}{\partial x}=yz+xy\frac{\partial z}{\partial x}.$$ Now, gather terms containing $\frac{\partial z}{\partial x}$: $$x^2\frac{\partial z}{\partial x}+\frac{y}{z}\frac{\partial z}{\partial x}-xy\frac{\partial z}{\partial x}=yz-2xz.$$ Factor out $\frac{\partial z}{\partial x}$ and solve: $$\frac{\partial z}{\partial x}\left(x^2+\frac{y}{z}-xy\right)=yz-2xz,$$ so $$\frac{\partial z}{\partial x}=\frac{yz-2xz}{x^2+\frac{y}{z}-xy}.$$ This technique is indispensable when dealing with complex constraints in thermodynamics or control systems where explicit solutions are impractical.

Engineering Applications: Putting Theory into Practice

Partial derivatives are the workhorses of quantitative engineering analysis. One key application is in optimization problems, such as minimizing the material cost of a tank or maximizing the efficiency of a wing design. For a function $f(x,y)$ representing cost or performance, its critical points are found by solving $f_x=0$ and $f_y=0$ simultaneously. The second partial derivatives then determine whether these points are minima, maxima, or saddle points using the second derivative test.

In heat transfer, the temperature $T(x,y,t)$ in a plate might satisfy the heat equation: $\frac{\partial T}{\partial t}=\alpha \left(\frac{{\partial }^{2}T}{\partial x^2}+\frac{{\partial }^{2}T}{\partial y^2}\right)$, where $\alpha$ is thermal diffusivity. Here, partial derivatives describe how temperature changes over time and space. Similarly, in fluid dynamics, the Navier-Stokes equations involve partial derivatives of velocity and pressure fields. Another common use is in sensitivity analysis, where $\frac{\partial P}{\partial r}$ might indicate how a project's profit $P$ changes with respect to interest rate $r$, aiding in financial risk assessment for engineering projects.

Consider a concrete scenario: designing a rectangular beam with width $w$, height $h$, and length $L$. The bending stress $\sigma$ is given by $\sigma =\frac{6M}{w{h}^2}$, where $M$ is the constant bending moment. To see how stress changes with dimensions, compute $\frac{\partial \sigma }{\partial w}=-\frac{6M}{w^2{h}^2}$ and $\frac{\partial \sigma }{\partial h}=-\frac{12M}{w{h}^3}$. These partials show that stress decreases more rapidly with increases in height than in width, guiding material-efficient design choices.

Common Pitfalls

1. Confusing Partial and Total Derivatives: A common error is using the ordinary derivative $\frac{d}{dx}$ when $\frac{\partial }{\partial x}$ is required. Remember, partial derivatives apply to functions of multiple variables, holding others constant. For instance, if $f(x,y)=x^2+y$, then $\frac{\partial f}{\partial x}=2x$, but $\frac{df}{dx}$ is not defined unless $y$ is a function of $x$. In engineering contexts, always check if variables are independent or related.

1. Misapplying Clairaut's Theorem: Students often assume mixed partials are always equal without verifying continuity. If $f(x,y)$ has discontinuous second derivatives, $f_{xy}$ and $f_{yx}$ may differ. For example, functions defined piecewise can violate the theorem's conditions. Always ensure the function's partials are continuous in the region of interest before equating mixed derivatives.

1. Chain Rule Errors in Implicit Differentiation: When finding $\frac{\partial z}{\partial x}$ from $F(x,y,z)=0$, it's easy to forget that $z$ is a function of $x$ and $y$. Differentiating a term like $yz$ with respect to $x$ requires the product rule: $\frac{\partial }{\partial x}(yz)=y\frac{\partial z}{\partial x}$, since $y$ is constant. Missing this leads to incorrect algebraic solutions.

1. Neglecting Units in Applied Problems: In engineering calculations, partial derivatives often have physical units. For example, $\frac{\partial T}{\partial x}$ has units of temperature per length. Dimensional analysis can help catch mistakes. If your computed partial derivative's units don't match the expected rate of change, revisit your differentiation steps.

Summary

• Partial derivatives, denoted $\frac{\partial f}{\partial x}$, measure the rate of change of a multivariable function with respect to one variable, treating all others as constants. They are fundamental to analyzing systems with multiple independent parameters.
• Geometrically, $\frac{\partial f}{\partial x}$ represents the slope of the tangent line to the cross-section of the surface $z=f(x,y)$ formed by holding $y$ fixed, providing a visual tool for understanding local behavior.
• Higher-order partial derivatives include mixed partials like $f_{xy}$. Under continuity conditions, Clairaut's theorem ensures $f_{xy}=f_{yx}$, simplifying calculations in many engineering contexts.
• Implicit partial differentiation allows you to find derivatives from equations where variables are not explicitly isolated, using the chain rule to handle interdependent variables.
• In engineering, partial derivatives are critical for optimizing designs, modeling physical phenomena (e.g., heat flow, fluid dynamics), and conducting sensitivity analyses to understand how outputs respond to input changes.