Engineering Math Series 20: Classification and Meaning of PDEs

While ODEs dealt with a single state changing over time, partial differential equations handle phenomena where time and space are intertwined. Phenomena such as heat spreading, waves propagating, and electric potential distributing across space are mostly represented as PDEs.

Why Are ODEs Not Enough

Consider the temperature of a rod. It changes not only over time but also varies along the position of the rod. Therefore the unknown function is no longer $y(t)$ but

$u(x,t)$

depending on multiple variables.

This requires partial derivatives.

$\frac{\partial u}{\partial t}, \quad \frac{\partial^2 u}{\partial x^2}$

appear together.

The Three Representative PDEs

The most important PDEs in introductory engineering math are usually three types.

Heat Equation

$u_t = k u_{xx}$

Represents diffusion phenomena where values spread and smooth out over time.

Wave Equation

$u_{tt} = c^2 u_{xx}$

Deals with vibration and propagation.

Laplace Equation

$u_{xx} + u_{yy} = 0$

Deals with steady-state distributions.

Why Classification Matters

PDEs are commonly classified as elliptic, parabolic, or hyperbolic. At the introductory level, remembering them as follows is sufficient.

• Parabolic: spreading and stabilization
• Hyperbolic: propagation and waves
• Elliptic: equilibrium and static distribution

This classification is not mere labeling but tells you about the nature of solutions and the type of boundary conditions needed.

Physical Interpretation

Heat-type problems

Local temperature differences smooth out over time. Sharp changes gradually becoming smooth is the characteristic feature.

Wave-type problems

Information propagates at finite speed. The picture of vibration at one point being transmitted to the side is important.

Elliptic-type problems

These deal not with time changes but with an already balanced state. The electrostatic potential distribution is a representative example.

A Simple Example by Hand

In steady-state 1D heat conduction, the time change vanishes, giving

$u_{xx}=0$

Integrating twice,

$u(x)=Ax+B$

With boundary conditions

$u(0)=0, \quad u(1)=100$

we get

$B=0, \quad A=100$

so

$u(x)=100x$

This example, though very simple, nicely shows that PDEs include equilibrium problems that can give very meaningful simple results.

Engineering Applications

Heat Transfer

The heat equation is fundamental to chip cooling, battery thermal management, and building insulation.

Vibration and Acoustics

String vibration, membrane vibration, and sound wave propagation in pipes are connected to the wave equation.

Electric Fields and Potential

Electrostatics and fluid potential problems are frequently treated from the Laplace equation perspective.

Common Mistakes

Viewing PDEs as merely "equations with one more variable" than ODEs

It is not just that a variable is added; the type of phenomenon itself changes. Spatial structure and boundary conditions are central.

Only memorizing the classification

Parabolic, hyperbolic, and elliptic classifications must be connected to the physical character of the solutions to be meaningful.

Taking boundary conditions lightly

For PDEs, boundary conditions constitute virtually half of the problem definition.

One-Line Summary

Partial differential equations handle phenomena involving both time and space, and their classification tells you about the physical character of the solutions.

Next Post Preview

In the next post, we will compare the three most representative PDEs -- the heat equation, wave equation, and Laplace equation -- more concretely.

References

• Erwin Kreyszig, Advanced Engineering Mathematics, 10th Edition
• Lawrence C. Evans, Partial Differential Equations
• Richard Haberman, Applied Partial Differential Equations