Dimensional Analysis and Vector Calculus
CSIR NET Physical Sciences: dimensions and the Buckingham pi theorem, vector algebra, index notation, grad, div, curl, Gauss and Stokes theorems, practice.
- 20 min
On this page · 12 sections
Prerequisites
- Partial derivatives, multiple integrals and the Jacobian, as in Metric Spaces and Several Variables.
- Determinants (the scalar triple product is one), from Rank, Determinants and Linear Equations.
- The Physical Sciences overview for how the paper is marked.
The Hook: How Big Is Quantum Gravity?
Dimensions and Dimensional Homogeneity
Every physical quantity has dimensions built from the SI base quantities: mass \( M \), length \( L \), time \( T \), current \( A \), temperature \( K \) (plus amount of substance and luminous intensity, rarely needed here). Two rules do almost all the work:
- Homogeneity. Every term in a correct equation has the same dimensions. You can only add, subtract or equate like quantities.
- Transcendental arguments are dimensionless. The argument of \( \sin \), \( \exp \), \( \ln \) or any power series must be a pure number. So in \( e^{-E/k_BT} \) the ratio \( E/k_BT \) is dimensionless, and in \( \sin(kx - \omega t) \) both \( kx \) and \( \omega t \) are.
Dimensions worth knowing cold
| Quantity | Dimensions | Quick derivation |
|---|---|---|
| Force | \( MLT^{-2} \) | \( ma \) |
| Energy, work, torque | \( ML^2T^{-2} \) | \( Fd \) |
| Power | \( ML^2T^{-3} \) | energy per time |
| Pressure, stress, energy density | \( ML^{-1}T^{-2} \) | force per area |
| Planck constant \( h \), \( \hbar \), angular momentum | \( ML^2T^{-1} \) | \( E/\nu \) |
| Gravitational constant \( G \) | \( M^{-1}L^3T^{-2} \) | \( Fr^2/m^2 \) |
| Boltzmann constant \( k_B \) | \( ML^2T^{-2}K^{-1} \) | energy per kelvin |
| Surface tension | \( MT^{-2} \) | force per length |
| Dynamic viscosity \( \eta \) | \( ML^{-1}T^{-1} \) | stress per velocity gradient |
| Charge | \( AT \) | current times time |
| Potential difference | \( ML^2T^{-3}A^{-1} \) | power per current |
| Resistance | \( ML^2T^{-3}A^{-2} \) | \( V/I \) |
| Magnetic field \( B \) | \( MT^{-2}A^{-1} \) | \( F = qvB \) |
| Permittivity \( \varepsilon_0 \) | \( M^{-1}L^{-3}T^4A^2 \) | \( q^2/(4\pi F r^2) \) |
| Permeability \( \mu_0 \) | \( MLT^{-2}A^{-2} \) | \( 1/(\varepsilon_0c^2) \) |
Some combinations are dimensionless and worth recognising on sight: the fine-structure constant \( \alpha = e^2/(4\pi\varepsilon_0\hbar c) \approx 1/137 \), the Reynolds number \( \rho vL/\eta \), and the ratio \( E/k_BT \).
Rayleigh’s method
Assume the quantity is a product of powers of the variables it depends on, and match exponents.
Simple pendulum. Suppose the period depends on mass \( m \), length \( l \) and \( g \): \( T_p = C\,m^a l^b g^c \). Then \( T = M^aL^{b}(LT^{-2})^c \), so \( a = 0 \), \( b + c = 0 \), \( -2c = 1 \). Hence \( c = -\tfrac{1}{2} \), \( b = \tfrac{1}{2} \) and
$$T_p = C\sqrt{\frac{l}{g}}.$$The method cannot give \( C \) (it is \( 2\pi \) for small swings), and it cannot tell you how \( T_p \) depends on the amplitude, because the amplitude is already dimensionless.
The Buckingham pi theorem
If a physical relation involves \( n \) variables built from \( k \) independent base dimensions, it can be rewritten as a relation among \( n - k \) independent dimensionless groups \( \Pi_1, \ldots, \Pi_{n-k} \):
$$f(q_1, \ldots, q_n) = 0 \quad\Longleftrightarrow\quad F(\Pi_1, \ldots, \Pi_{n-k}) = 0.$$For the pendulum with amplitude: variables \( T_p, m, l, g, \theta_0 \) (\( n = 5 \)), dimensions \( M, L, T \) (\( k = 3 \)), so two groups: \( \Pi_1 = T_p\sqrt{g/l} \) and \( \Pi_2 = \theta_0 \). The law must be \( T_p = \sqrt{l/g}\,\phi(\theta_0) \), and mass has dropped out because nothing else carries \( M \).
What dimensional analysis cannot do
- It gives no dimensionless constants (\( 2\pi \), \( 6\pi \) in Stokes’s law, \( \tfrac{1}{2} \)).
- It cannot separate two variables with the same dimensions (two lengths, say a radius and a wavelength): any function of their ratio is allowed.
- It cannot decide between \( \sin \) and \( \cos \), or between \( e^{x} \) and \( 1 + x \).
- It needs you to list the right variables. Leave one out and the answer is wrong; add an irrelevant one and you get an undetermined function.
Vector Algebra
Products
For \( \mathbf A, \mathbf B, \mathbf C \) in three dimensions:
- Dot product: \( \mathbf A\cdot\mathbf B = AB\cos\theta = A_iB_i \) (repeated indices summed). Zero means perpendicular.
- Cross product: \( \lvert\mathbf A\times\mathbf B\rvert = AB\sin\theta \), the area of the parallelogram; direction by the right-hand rule. Zero means parallel.
- Scalar triple product: \( \mathbf A\cdot(\mathbf B\times\mathbf C) \) is the determinant of the three component rows, the signed volume of the parallelepiped. It is unchanged by cyclic permutation and changes sign under a swap. Zero means the three vectors are coplanar (linearly dependent).
- Vector triple product (“BAC minus CAB”):
- Lagrange identity: \( (\mathbf A\times\mathbf B)\cdot(\mathbf C\times\mathbf D) = (\mathbf A\cdot\mathbf C)(\mathbf B\cdot\mathbf D) - (\mathbf A\cdot\mathbf D)(\mathbf B\cdot\mathbf C) \). With \( \mathbf C = \mathbf A \), \( \mathbf D = \mathbf B \) it gives \( \lvert\mathbf A\times\mathbf B\rvert^2 = A^2B^2 - (\mathbf A\cdot\mathbf B)^2 \).
Index notation
Two symbols turn vector identities into algebra:
- Kronecker delta \( \delta_{ij} \) (1 if \( i = j \), else 0), with \( \delta_{ii} = 3 \).
- Levi-Civita symbol \( \varepsilon_{ijk} \): \( +1 \) for even permutations of \( (1,2,3) \), \( -1 \) for odd ones, 0 if any index repeats. Then \( (\mathbf A\times\mathbf B)_i = \varepsilon_{ijk}A_jB_k \).
The one contraction identity to memorise:
$$\boxed{\varepsilon_{ijk}\varepsilon_{imn} = \delta_{jm}\delta_{kn} - \delta_{jn}\delta_{km}}$$Contracting further: \( \varepsilon_{ijk}\varepsilon_{ijn} = 2\delta_{kn} \) and \( \varepsilon_{ijk}\varepsilon_{ijk} = 6 \). The BAC-CAB rule is the first identity in one line: \( [\mathbf A\times(\mathbf B\times\mathbf C)]_i = \varepsilon_{ijk}A_j\varepsilon_{kmn}B_mC_n = (\delta_{im}\delta_{jn} - \delta_{in}\delta_{jm})A_jB_mC_n = B_i(A_jC_j) - C_i(A_jB_j) \).
Vector Calculus
Gradient, divergence and curl
With \( \nabla = \hat{\mathbf x}\,\partial_x + \hat{\mathbf y}\,\partial_y + \hat{\mathbf z}\,\partial_z \):
| Operation | Acts on | Gives | Meaning |
|---|---|---|---|
| \( \nabla f \) | scalar | vector | direction and rate of steepest increase; normal to level surfaces |
| \( \nabla\cdot\mathbf F \) | vector | scalar | net outflow per unit volume (source density) |
| \( \nabla\times\mathbf F \) | vector | vector | circulation per unit area (local rotation) |
| \( \nabla^2 f = \nabla\cdot\nabla f \) | scalar | scalar | how \( f \) differs from its local average |
The directional derivative of \( f \) along a unit vector \( \hat{\mathbf n} \) is \( \hat{\mathbf n}\cdot\nabla f \); its maximum over directions is \( \lvert\nabla f\rvert \).
Identities
Two second-derivative identities hold for every smooth field:
$$\boxed{\nabla\times(\nabla f) = \mathbf 0, \qquad \nabla\cdot(\nabla\times\mathbf A) = 0}$$and one rewrites the curl of a curl:
$$\nabla\times(\nabla\times\mathbf A) = \nabla(\nabla\cdot\mathbf A) - \nabla^2\mathbf A.$$Product rules (prove any of them with \( \varepsilon_{ijk} \)):
- \( \nabla(fg) = f\nabla g + g\nabla f \)
- \( \nabla\cdot(f\mathbf A) = f\,\nabla\cdot\mathbf A + \mathbf A\cdot\nabla f \)
- \( \nabla\times(f\mathbf A) = f\,\nabla\times\mathbf A + \nabla f\times\mathbf A \)
- \( \nabla\cdot(\mathbf A\times\mathbf B) = \mathbf B\cdot(\nabla\times\mathbf A) - \mathbf A\cdot(\nabla\times\mathbf B) \)
The position vector: results to know
With \( \mathbf r = (x, y, z) \) and \( r = \lvert\mathbf r\rvert \):
| Expression | Value |
|---|---|
| \( \nabla\cdot\mathbf r \) | 3 |
| \( \nabla\times\mathbf r \) | \( \mathbf 0 \) |
| \( \nabla r^n \) | \( nr^{n-2}\mathbf r \) |
| \( \nabla\cdot(r^n\mathbf r) \) | \( (n+3)r^n \) (for \( r \neq 0 \)) |
| \( \nabla\cdot\hat{\mathbf r} \) | \( 2/r \) |
| \( \nabla(1/r) \) | \( -\hat{\mathbf r}/r^2 \) |
| \( (\mathbf A\cdot\nabla)\mathbf r \) | \( \mathbf A \) |
| \( \nabla\cdot(\hat{\mathbf r}/r^2) \) | \( 4\pi\delta^3(\mathbf r) \) |
| \( \nabla^2(1/r) \) | \( -4\pi\delta^3(\mathbf r) \) |
The last two are the most tested. For \( r \neq 0 \), \( \nabla\cdot(\hat{\mathbf r}/r^2) = 0 \) by the \( (n+3)r^n \) rule with \( n = -3 \). Yet the flux of \( \hat{\mathbf r}/r^2 \) through any sphere centred on the origin is \( 4\pi \). The resolution is that all the divergence sits at the origin, as a three-dimensional delta function. This is Coulomb’s law turned into Gauss’s law.
Integral theorems
| Theorem | Statement | Use |
|---|---|---|
| Gradient theorem | \( \int_a^b\nabla f\cdot d\mathbf l = f(b) - f(a) \) | line integrals of conservative fields |
| Divergence (Gauss) | \( \int_V(\nabla\cdot\mathbf F)\,dV = \oint_S\mathbf F\cdot d\mathbf a \) | flux through a closed surface |
| Curl (Stokes) | \( \int_S(\nabla\times\mathbf F)\cdot d\mathbf a = \oint_C\mathbf F\cdot d\mathbf l \) | circulation round a closed loop |
| Green (plane) | \( \oint_C(P\,dx + Q\,dy) = \iint_R\left(\partial_xQ - \partial_yP\right)dx\,dy \) | 2D circulation; area \( = \tfrac{1}{2}\oint(x\,dy - y\,dx) \) |
Orientation matters: in Stokes’s theorem the loop is traversed counter-clockwise when seen from the side the normal points to. The surface in Stokes’s theorem is any surface bounded by the loop, so pick the flattest one.
Conservative fields
On a simply connected region, the following are equivalent:
- \( \nabla\times\mathbf F = \mathbf 0 \) everywhere in the region.
- \( \mathbf F = \nabla\phi \) for some scalar potential \( \phi \).
- \( \int\mathbf F\cdot d\mathbf l \) depends only on the endpoints.
- \( \oint\mathbf F\cdot d\mathbf l = 0 \) round every closed loop.
Similarly \( \nabla\cdot\mathbf B = 0 \) everywhere means \( \mathbf B = \nabla\times\mathbf A \) for some vector potential. The classic counter-example to “curl-free implies conservative” is
$$\mathbf F = \frac{-y\,\hat{\mathbf x} + x\,\hat{\mathbf y}}{x^2 + y^2},$$whose curl vanishes for \( (x, y) \neq (0, 0) \) but whose circulation round any loop enclosing the \( z \)-axis is \( 2\pi \). The plane minus the axis is not simply connected. This is the field of a line current, and the \( 2\pi \) is Ampère’s law.
Curvilinear coordinates
For orthogonal coordinates \( (u_1, u_2, u_3) \) with scale factors \( h_i \) (so that \( dl^2 = \sum h_i^2du_i^2 \)):
| System | Coordinates | Scale factors | Volume element |
|---|---|---|---|
| Cartesian | \( x, y, z \) | \( 1, 1, 1 \) | \( dx\,dy\,dz \) |
| Cylindrical | \( s, \phi, z \) | \( 1, s, 1 \) | \( s\,ds\,d\phi\,dz \) |
| Spherical | \( r, \theta, \phi \) | \( 1, r, r\sin\theta \) | \( r^2\sin\theta\,dr\,d\theta\,d\phi \) |
In spherical coordinates:
$$\nabla\cdot\mathbf A = \frac{1}{r^2}\frac{\partial(r^2A_r)}{\partial r} + \frac{1}{r\sin\theta}\frac{\partial(\sin\theta\,A_\theta)}{\partial\theta} + \frac{1}{r\sin\theta}\frac{\partial A_\phi}{\partial\phi}$$$$\nabla^2f = \frac{1}{r^2}\frac{\partial}{\partial r}\left(r^2\frac{\partial f}{\partial r}\right) + \frac{1}{r^2\sin\theta}\frac{\partial}{\partial\theta}\left(\sin\theta\frac{\partial f}{\partial\theta}\right) + \frac{1}{r^2\sin^2\theta}\frac{\partial^2f}{\partial\phi^2}$$In cylindrical coordinates: \( \nabla\cdot\mathbf A = \frac{1}{s}\partial_s(sA_s) + \frac{1}{s}\partial_\phi A_\phi + \partial_zA_z \) and \( \nabla^2f = \frac{1}{s}\partial_s(s\,\partial_sf) + \frac{1}{s^2}\partial_\phi^2f + \partial_z^2f \).
A quick check of the radial part: for \( f = 1/r \), \( r^2f' = -1 \), a constant, so \( \nabla^2(1/r) = 0 \) away from the origin, consistent with the delta-function result above. The angular part of \( \nabla^2 \) is where Legendre polynomials and spherical harmonics come from (see Linear ODEs and Special Functions).
Memory Tricks
Don’t Confuse
| This | vs | That |
|---|---|---|
| \( \nabla\cdot(\hat{\mathbf r}/r^2) = 0 \) for \( r \neq 0 \) | Flux of \( \hat{\mathbf r}/r^2 \) through a sphere about the origin \( = 4\pi \) | |
| \( \nabla\times\mathbf F = 0 \) on a simply connected region: conservative | \( \nabla\times\mathbf F = 0 \) on a region with a hole: maybe not | |
| \( \nabla\cdot\mathbf r = 3 \) (in 3D) | \( \nabla\cdot\mathbf r = 2 \) (in a 2D problem) | |
| Scalar triple product: volume, a scalar | Vector triple product: a vector in the plane of \( \mathbf B \) and \( \mathbf C \) | |
| \( h \) and angular momentum: \( ML^2T^{-1} \) | Energy and torque: \( ML^2T^{-2} \) | |
| Dimensionally correct | Correct (constants and functions of dimensionless groups are invisible) |
Common Mistakes
Practice
All questions below are original practice questions written for this page. They are not from any CSIR NET paper.
Level 1: Foundation
- What are the dimensions of \( \hbar/(mc) \), where \( m \) is a mass?
- Evaluate \( \nabla\cdot(x\,\hat{\mathbf x} + y\,\hat{\mathbf y} + z\,\hat{\mathbf z}) \).
- A relation involves 6 variables built from \( M \), \( L \) and \( T \). How many independent dimensionless groups does the Buckingham theorem predict?
Answers
- Length. \( [\hbar] = ML^2T^{-1} \), \( [mc] = MLT^{-1} \); the ratio is \( L \). (It is the reduced Compton wavelength.)
- 3. Each of \( \partial_xx \), \( \partial_yy \), \( \partial_zz \) is 1.
- 3. \( n - k = 6 - 3 \).
Level 2: Part B Style
Practice Q1. Which combination of \( G \), \( \hbar \) and \( c \) has the dimensions of length?
(a) \( \sqrt{\hbar c/G} \) (b) \( \sqrt{\hbar G/c^3} \) (c) \( \sqrt{\hbar G/c^5} \) (d) \( \hbar G/c^2 \)
Answer: (b)
Write \( \ell = \hbar^aG^bc^d \): \( M^{a-b}L^{2a+3b+d}T^{-a-2b-d} \). Mass: \( a = b \). Time: \( -3a - d = 0 \Rightarrow d = -3a \). Length: \( 5a - 3a = 1 \Rightarrow a = \tfrac{1}{2} \). So \( \ell = \sqrt{\hbar G/c^3} \approx 1.6\times10^{-35} \) m. Option (a) is a mass and (c) a time; (d) has dimensions \( L^3T^{-1} \).
Practice Q2. For \( \mathbf F = y\,\hat{\mathbf x} - x\,\hat{\mathbf y} \), the line integral \( \oint\mathbf F\cdot d\mathbf l \) round the unit circle in the \( xy \)-plane, traversed counter-clockwise, is:
(a) \( 2\pi \) (b) \( -2\pi \) (c) \( \pi \) (d) 0
Answer: (b) −2π
\( (\nabla\times\mathbf F)_z = \partial_xF_y - \partial_yF_x = -1 - 1 = -2 \). By Stokes’s theorem with the normal along \( +\hat{\mathbf z} \) (counter-clockwise loop), the circulation is \( -2\times\pi\times1^2 = -2\pi \). Directly: on \( x = \cos t \), \( y = \sin t \), \( \mathbf F\cdot d\mathbf l = (-\sin^2t - \cos^2t)\,dt = -dt \). Option (a) has the orientation reversed.
Practice Q3. The vector field \( \mathbf F = r^n\mathbf r \) has zero divergence everywhere except the origin when \( n \) equals:
(a) \( -1 \) (b) \( -2 \) (c) \( -3 \) (d) 0
Answer: (c) −3
\( \nabla\cdot(r^n\mathbf r) = r^n\nabla\cdot\mathbf r + \mathbf r\cdot\nabla r^n = 3r^n + \mathbf r\cdot nr^{n-2}\mathbf r = (n + 3)r^n \). It vanishes for \( n = -3 \), that is \( \mathbf F = \hat{\mathbf r}/r^2 \), the Coulomb field.
Practice Q4. The value of \( \varepsilon_{ijk}\varepsilon_{ijk} \) (summed over all indices from 1 to 3) is:
(a) 0 (b) 3 (c) 6 (d) 27
Answer: (c) 6
Only the 6 permutations of \( (1,2,3) \) give non-zero entries, each squared to 1. Equivalently, \( \varepsilon_{ijk}\varepsilon_{ijn} = 2\delta_{kn} \), and \( 2\delta_{kk} = 6 \).
Level 3: Part C Style
Practice Q5. Let \( \mathbf F = x^2\,\hat{\mathbf x} + y^2\,\hat{\mathbf y} + z^2\,\hat{\mathbf z} \) and let \( S \) be the surface of the unit cube \( 0 \le x, y, z \le 1 \), outward normal. Consider:
- \( \oint_S\mathbf F\cdot d\mathbf a = 3 \).
- The faces \( x = 0 \), \( y = 0 \), \( z = 0 \) contribute nothing to the flux.
- \( \nabla\times\mathbf F = \mathbf 0 \), so \( \mathbf F \) is the gradient of \( (x^3 + y^3 + z^3)/3 \).
- The flux through the face \( x = 1 \) alone is \( \tfrac{1}{3} \).
Which are correct?
(a) 1 and 2 only (b) 1, 2 and 3 only (c) 2, 3 and 4 only (d) all four
Answer: (b) 1, 2 and 3 only
Statement 1: \( \nabla\cdot\mathbf F = 2x + 2y + 2z \). Over the cube, \( \int_0^1\!\int_0^1\!\int_0^1 2x\,dx\,dy\,dz = 1 \), and the same for \( y \) and \( z \), so the flux is 3. True.
Statement 2: on \( x = 0 \) the normal component \( F_x = x^2 = 0 \), and similarly for the other two. True.
Statement 3: each component depends only on its own coordinate, so every cross-derivative is zero and the curl vanishes; \( \nabla\big[(x^3 + y^3 + z^3)/3\big] = \mathbf F \). True.
Statement 4: on \( x = 1 \), \( F_x = 1 \) over unit area, so that face carries flux 1 (and the three far faces give 3 in all, consistent with statement 1). False.
Practice Q6. The frequency \( \omega \) of small shape oscillations of a liquid drop is assumed to depend only on the surface tension \( \sigma \), the density \( \rho \) and the radius \( R \). If the radius is doubled with \( \sigma \) and \( \rho \) unchanged, \( \omega \) becomes approximately what fraction of its original value?
(a) 0.25 (b) 0.35 (c) 0.50 (d) 0.71
Answer: (b) 0.35
\( \omega = C\sigma^a\rho^bR^c \): \( T^{-1} = (MT^{-2})^a(ML^{-3})^bL^c \). Mass: \( a + b = 0 \). Time: \( -2a = -1 \Rightarrow a = \tfrac{1}{2} \), so \( b = -\tfrac{1}{2} \). Length: \( -3b + c = 0 \Rightarrow c = -\tfrac{3}{2} \).
\( \omega \propto \sqrt{\sigma/(\rho R^3)} \), so doubling \( R \) multiplies \( \omega \) by \( 2^{-3/2} \approx 0.354 \). Option (d) is \( 2^{-1/2} \), from treating \( \omega \propto R^{-1/2} \); (a) and (c) come from guessing integer powers. Here \( n = 4 \) variables and \( k = 3 \) dimensions give exactly one group, \( \omega^2\rho R^3/\sigma \), which must be a constant.
Practice Q7. Consider \( \mathbf F = yz\,\hat{\mathbf x} + xz\,\hat{\mathbf y} + xy\,\hat{\mathbf z} \). Which of the following are correct?
- \( \nabla\times\mathbf F = \mathbf 0 \).
- \( \nabla\cdot\mathbf F = 0 \).
- The line integral of \( \mathbf F \) from \( (0,0,0) \) to \( (1,2,3) \) is 6, along any path.
- The potential \( \phi \) with \( \mathbf F = \nabla\phi \) satisfies Laplace’s equation.
(a) 1 and 3 only (b) 1, 2 and 3 only (c) 1, 3 and 4 only (d) all four
Answer: (d) all four
Statement 1: \( (\nabla\times\mathbf F)_x = \partial_y(xy) - \partial_z(xz) = x - x = 0 \), and cyclically. True.
Statement 2: \( \partial_x(yz) + \partial_y(xz) + \partial_z(xy) = 0 \). True.
Statement 3: \( \phi = xyz \) (check: \( \nabla(xyz) = (yz, xz, xy) \)). The field is curl-free on all of space (simply connected), so the integral is \( \phi(1,2,3) - \phi(0,0,0) = 6 \) on every path. True.
Statement 4: \( \nabla^2\phi = \nabla\cdot\mathbf F = 0 \). True. A field that is both irrotational and solenoidal always has a harmonic potential.
Quick Revision Table
| Item | Result |
|---|---|
| Homogeneity | every term has the same dimensions; arguments of exp, sin, log dimensionless |
| Buckingham | \( n \) variables, \( k \) dimensions: \( n - k \) groups |
| Planck length | \( \sqrt{\hbar G/c^3} \approx 1.6\times10^{-35} \) m |
| BAC-CAB | \( \mathbf A\times(\mathbf B\times\mathbf C) = \mathbf B(\mathbf A\cdot\mathbf C) - \mathbf C(\mathbf A\cdot\mathbf B) \) |
| Epsilon-delta | \( \varepsilon_{ijk}\varepsilon_{imn} = \delta_{jm}\delta_{kn} - \delta_{jn}\delta_{km} \) |
| Zero identities | \( \nabla\times\nabla f = 0 \), \( \nabla\cdot(\nabla\times\mathbf A) = 0 \) |
| Curl curl | \( \nabla(\nabla\cdot\mathbf A) - \nabla^2\mathbf A \) |
| \( \nabla\cdot(r^n\mathbf r) \) | \( (n+3)r^n \) |
| Point source | \( \nabla\cdot(\hat{\mathbf r}/r^2) = 4\pi\delta^3(\mathbf r) \) |
| Gauss / Stokes | volume of divergence = flux; surface of curl = circulation |
| Spherical scale factors | \( 1, r, r\sin\theta \) |
Key Takeaways
What’s Next
- Linear Algebra, Matrices and Eigenvalues: the next topic in this chapter.
- Mathematical Methods of Physics (Core): chapter overview and syllabus map.
- Quantum Mechanics (Core), where \( \nabla^2 \) in spherical coordinates separates the hydrogen atom.
- The same unit-checking habit that saves marks here is drilled for Part A in Numerical Ability.