12 Applications of Multivariable Calculus

A connected guide to modeling geometry, motion, mass, fields, work, flux, and optimization with the main tools of multivariable calculus.

A modeling framework for applications

Multivariable calculus extends one-variable ideas to quantities that depend on several inputs. The central modeling choices are:

  • A scalar field assigns one number to each point, such as temperature, density, or elevation.

  • A vector field assigns a vector to each point, such as velocity, force, or an electric field.

  • A parametrized curve describes motion or a wire.

  • A surface or solid describes a geometric region over which quantities can be accumulated.

The main operations match the geometry of the application: derivatives describe local change, multiple integrals accumulate over regions, line integrals accumulate along curves, and flux integrals measure passage through surfaces.

A useful workflow is to identify the physical quantity first, choose the mathematical object that represents it, select coordinates suited to the geometry, determine bounds carefully, and then check units and interpretation.

Takeaway: The same calculus framework can describe geometry, motion, mass, force, fluid flow, and optimization by matching each situation with the right field, curve, surface, or integral.

Surfaces, tangent planes, and area

A surface given explicitly by z=f(x,y)z=f(x,y) has tangent plane at (x0,y0,z0)\left(x_0,y_0,z_0\right), where z0=f(x0,y0)z_0=f(x_0,y_0), given by

z−z0=fx(x0,y0)(x−x0)+fy(x0,y0)(y−y0).z-z_0=f_x(x_0,y_0)(x-x_0)+f_y(x_0,y_0)(y-y_0).

The of ff is

∇f=⟨fx,fy⟩.\nabla f=\langle f_x,f_y\rangle.

It is perpendicular to the level curve f(x,y)=cf(x,y)=c, so it gives a normal direction to the curve. An implicit surface F(x,y,z)=0F(x,y,z)=0 has normal vector

∇F=⟨Fx,Fy,Fz⟩.\nabla F=\langle F_x,F_y,F_z\rangle.

A normal line through a point PP can therefore be written as

r(t)=P+t∇F(P).\mathbf r(t)=P+t\nabla F(P).

For a surface above a region RR, the surface area is

A=∬R1+fx2+fy2 dA.A=\iint_R \sqrt{1+f_x^2+f_y^2}\,dA.

The square-root factor corrects for the enlargement of a tilted surface patch relative to its projection onto the xyxy-plane.

Takeaway: Partial derivatives determine tangent behavior, while gradients provide normal directions and steepest increase.

Motion along curves

The position of a particle moving through space is represented by a vector-valued function

r(t)=⟨x(t),y(t),z(t)⟩.\mathbf r(t)=\langle x(t),y(t),z(t)\rangle.

Differentiation gives velocity and acceleration:

v(t)=r′(t),a(t)=r′′(t).\mathbf v(t)=\mathbf r'(t), \qquad \mathbf a(t)=\mathbf r''(t).

Speed is the magnitude of velocity, ∥v(t)∥\|\mathbf v(t)\|, and distance traveled from t=at=a to t=bt=b is

L=∫ab∥v(t)∥ dt.L=\int_a^b \|\mathbf v(t)\|\,dt.

For example, if

r(t)=⟨3t,4t,12t⟩,\mathbf r(t)=\langle 3t,4t,12t\rangle,

then v(t)=⟨3,4,12⟩\mathbf v(t)=\langle 3,4,12\rangle, the speed is 1313, and the distance traveled over the interval from aa to bb is 13(b−a)13(b-a) units.

The unit tangent vector is

T(t)=v(t)∥v(t)∥.\mathbf T(t)=\frac{\mathbf v(t)}{\|\mathbf v(t)\|}.

It gives the instantaneous direction of motion. Curvature measures how rapidly this direction changes with respect to distance traveled, while the principal normal direction points toward the local bending of the path.

Takeaway: Position, velocity, acceleration, speed, distance, and path direction are successive geometric interpretations of a vector-valued function and its derivatives.

Multiple integrals and coordinates

A double integral accumulates a quantity over a planar region RR:

∬Rf(x,y) dA.\iint_R f(x,y)\,dA.

A triple integral accumulates throughout a solid EE:

∭Ef(x,y,z) dV.\iiint_E f(x,y,z)\,dV.

If a solid lies between z=g(x,y)z=g(x,y) and z=f(x,y)z=f(x,y), its volume is

V=∬R[f(x,y)−g(x,y)] dA.V=\iint_R\bigl[f(x,y)-g(x,y)\bigr]\,dA.

For a general three-dimensional region, volume can be written as

V=∭E1 dV.V=\iiint_E 1\,dV.

Coordinate selection should follow the geometry:

  • Cartesian coordinates use dA=dx dydA=dx\,dy and dV=dx dy dzdV=dx\,dy\,dz.

  • Polar coordinates use x=rcos⁡θx=r\cos\theta, y=rsin⁡θy=r\sin\theta, and dA=r dr dθdA=r\,dr\,d\theta.

  • Cylindrical coordinates use x=rcos⁡θx=r\cos\theta, y=rsin⁡θy=r\sin\theta, and dV=r dz dr dθdV=r\,dz\,dr\,d\theta.

  • Spherical coordinates use x=ρsin⁡ϕcos⁡θx=\rho\sin\phi\cos\theta, y=ρsin⁡ϕsin⁡θy=\rho\sin\phi\sin\theta, z=ρcos⁡ϕz=\rho\cos\phi, and dV=ρ2sin⁡ϕ dρ dϕ dθdV=\rho^2\sin\phi\,d\rho\,d\phi\,d\theta.

A corrects for the stretching caused by a coordinate change. In particular, the factor rr appears in polar and cylindrical area or volume elements, and the factor ρ2sin⁡ϕ\rho^2\sin\phi appears in spherical volume.

Takeaway: Set up the region before integrating, and use coordinates that simplify its boundaries and symmetry.

Mass, balance, and rotation

For a lamina occupying a region RR with surface density ρ(x,y)\rho(x,y), the mass is

m=∬Rρ(x,y) dA.m=\iint_R \rho(x,y)\,dA.

Its moments about the coordinate axes are

Mx=∬Ryρ(x,y) dA,My=∬Rxρ(x,y) dA.M_x=\iint_R y\rho(x,y)\,dA, \qquad M_y=\iint_R x\rho(x,y)\,dA.

The is then

xˉ=Mym,yˉ=Mxm.\bar{x}=\frac{M_y}{m}, \qquad \bar{y}=\frac{M_x}{m}.

For constant density, the is the centroid, which depends only on the shape of the region. For a solid with volume density ρ(x,y,z)\rho(x,y,z),

m=∭Eρ(x,y,z) dV.m=\iiint_E \rho(x,y,z)\,dV.

The corresponding coordinates are

xˉ=Myzm,yˉ=Mxzm,zˉ=Mxym.\bar{x}=\frac{M_{yz}}{m}, \qquad \bar{y}=\frac{M_{xz}}{m}, \qquad \bar{z}=\frac{M_{xy}}{m}.

Moment of inertia measures resistance to rotation. For a lamina,

Ix=∬Ry2ρ dA,Iy=∬Rx2ρ dA,I_x=\iint_R y^2\rho\,dA, \qquad I_y=\iint_R x^2\rho\,dA,

and the polar moment about the origin is

I0=∬R(x2+y2)ρ dA=Ix+Iy.I_0=\iint_R (x^2+y^2)\rho\,dA=I_x+I_y.

For a solid rotating about the zz-axis,

Iz=∭E(x2+y2)ρ dV.I_z=\iiint_E (x^2+y^2)\rho\,dV.

The squared distance from the axis explains why mass farther away contributes disproportionately to rotational resistance.

Takeaway: Density supplies the weighting, moments locate the balance point, and squared distance determines rotational resistance.

Fields, sources, and rotation

A Vector field assigns a vector to every point. In three dimensions it can be written as

F(x,y,z)=⟨P(x,y,z),Q(x,y,z),R(x,y,z)⟩.\mathbf F(x,y,z)=\langle P(x,y,z),Q(x,y,z),R(x,y,z)\rangle.

A velocity field describes both the direction and speed of a fluid at each location. A curve whose tangent follows the field is a flow line or streamline.

Two local measurements are especially important. Divergence is

∇⋅F=Px+Qy+Rz.\nabla\cdot\mathbf F=P_x+Q_y+R_z.

Positive divergence indicates local expansion or a source, while negative divergence indicates local compression or a sink. Curl is

∇×F=⟨Ry−Qz,  Pz−Rx,  Qx−Py⟩.\nabla\times\mathbf F =\left\langle R_y-Q_z,\;P_z-R_x,\;Q_x-P_y\right\rangle.

It measures the local tendency of the field to rotate.

A field is conservative when there is a potential function ff such that

F=∇f.\mathbf F=\nabla f.

For a conservative field, work depends only on the initial and final points, not on the route between them.

Takeaway: Divergence describes local spreading or gathering, curl describes local rotation, and a potential function explains path-independent work.

Accumulation along curves and work

A accumulates a quantity along a curve. For a scalar function ff and a parametrized curve r(t)\mathbf r(t),

∫Cf ds=∫abf(r(t))∥r′(t)∥ dt.\int_C f\,ds =\int_a^b f(\mathbf r(t))\|\mathbf r'(t)\|\,dt.

If λ\lambda is the linear density of a wire, its mass is

m=∫Cλ ds.m=\int_C \lambda\,ds.

For a vector field, the is

∫CF⋅dr=∫abF(r(t))⋅r′(t) dt.\int_C \mathbf F\cdot d\mathbf r =\int_a^b \mathbf F(\mathbf r(t))\cdot\mathbf r'(t)\,dt.

When F\mathbf F is a force field, this is the work done along the path. Equivalently,

∫CF⋅dr=∫CP dx+Q dy+R dz.\int_C \mathbf F\cdot d\mathbf r =\int_C P\,dx+Q\,dy+R\,dz.

For a constant force F=⟨2,3⟩\mathbf F=\langle 2,3\rangle moving an object from A=(1,1)A=(1,1) to B=(4,5)B=(4,5),

W=F⋅(B−A)=⟨2,3⟩⋅⟨3,4⟩=18.W=\mathbf F\cdot(B-A) =\langle 2,3\rangle\cdot\langle 3,4\rangle =18.

If F=∇f\mathbf F=\nabla f, the fundamental theorem for line integrals gives

∫CF⋅dr=f(B)−f(A).\int_C \mathbf F\cdot d\mathbf r=f(B)-f(A).

Thus, conservative fields allow work to be computed from endpoint values alone.

Takeaway: Scalar line integrals measure accumulation along a curve, while vector line integrals measure directional effects such as work.

Flux and integral theorems

Flux measures how much of a vector field crosses an oriented curve or surface. For an oriented surface SS with unit normal n\mathbf n,

∬SF⋅n dS\iint_S \mathbf F\cdot\mathbf n\,dS

is the flux through the surface. For a fluid velocity field, it represents net flow crossing the surface.

The relates the flux across a closed surface to the divergence inside the enclosed solid:

∬∂EF⋅n dS=∭E∇⋅F dV.\iint_{\partial E}\mathbf F\cdot\mathbf n\,dS =\iiint_E \nabla\cdot\mathbf F\,dV.

Green's theorem converts a positively oriented planar circulation integral into a double integral:

∮CP dx+Q dy=∬R(Qx−Py) dA.\oint_C P\,dx+Q\,dy =\iint_R (Q_x-P_y)\,dA.

Stokes' theorem extends this idea to a surface in space:

∮∂SF⋅dr=∬S(∇×F)⋅n dS.\oint_{\partial S}\mathbf F\cdot d\mathbf r =\iint_S (\nabla\times\mathbf F)\cdot\mathbf n\,dS.

These theorems connect local differential behavior with global circulation or flux. They are often used to replace a difficult boundary integral by an easier regional or surface integral, provided the orientation and boundary conditions are handled correctly.

Takeaway: Green's theorem, Stokes' theorem, and the translate between boundary behavior and what happens throughout the enclosed region.

Extrema and constrained optimization

For an unconstrained function f(x,y)f(x,y), an interior critical point occurs where

∇f=⟨fx,fy⟩=0,\nabla f=\langle f_x,f_y\rangle=\mathbf 0,

or where a partial derivative does not exist. The second-derivative test uses

D=fxxfyy−(fxy)2.D=f_{xx}f_{yy}-(f_{xy})^2.

At a critical point:

  • If D>0D>0 and fxx>0f_{xx}>0, the point is a local minimum.

  • If D>0D>0 and fxx<0f_{xx}<0, the point is a local maximum.

  • If D<0D<0, the point is a saddle point.

  • If D=0D=0, the test is inconclusive.

To find an absolute maximum or minimum on a closed and bounded region, evaluate the function at every interior critical point and on every boundary segment, including endpoints where appropriate.

is the standard method for a constraint

g(x,y,z)=c.g(x,y,z)=c.

Solve

∇f=λ∇g,g=c.\nabla f=\lambda\nabla g, \qquad g=c.

The equation says that the objective and constraint have parallel normal vectors at a constrained extremum. The solutions are candidates, so their objective values must still be compared.

Takeaway: Critical-point tests classify local behavior, boundary checks are required for absolute extrema, and incorporate equality constraints.