Physics: Dimensional Analysis and Unit Consistency

Abstract

Dimensional analysis is a foundational tool in physics that constrains the form of physical relationships by requiring consistency across units and dimensions. This article examines the principles underlying dimensional reasoning, demonstrates how dimensional constraints guide problem-solving, and illustrates the method through applications in mechanics. The approach is particularly valuable when deriving relationships between physical quantities without solving differential equations explicitly.

Background

Physical quantities are characterized not only by their numerical values but by their dimensions—the fundamental categories (length, mass, time, temperature, etc.) that define what they measure. Every equation in physics must be dimensionally consistent: both sides of an equality must have the same dimensions ^{[rolling-without-slipping]}.

This requirement is not merely a bookkeeping convenience. It reflects a deep principle: the laws of physics are independent of the choice of units. Whether we measure length in meters or feet, the underlying relationship between quantities remains unchanged. Dimensional analysis exploits this invariance to constrain the possible forms of physical laws.

The fundamental dimensions in classical mechanics are:

• Length: $[L]$
• Mass: $[M]$
• Time: $[T]$

All other mechanical quantities can be expressed as products of powers of these three. For example, velocity has dimensions $[L T^{-1}]$, acceleration has $[L T^{-2}]$, and force has $[M L T^{-2}]$.

Key Results

Dimensional Consistency as a Constraint

If a physical relationship involves quantities with dimensions $Q_1, Q_2, \ldots, Q_n$, then any valid equation relating them must satisfy:

$[Q_1]^{a_1} [Q_2]^{a_2} \cdots [Q_n]^{a_n} = [1]$

where the exponents $a_i$ are chosen so that the product is dimensionless. This constraint alone can eliminate many proposed relationships and guide the form of correct ones.

The Buckingham Pi Theorem

When a physical phenomenon depends on $n$ quantities with $k$ independent fundamental dimensions, the relationship can be expressed in terms of $n - k$ dimensionless groups (called $\pi$ groups). Each $\pi$ group is a dimensionless combination of the original quantities. This reduces the complexity of the problem: instead of analyzing $n$ variables, one analyzes $n - k$ dimensionless ratios.

For instance, if a quantity $Q$ depends on three variables $a$, $b$, and $c$ with two independent dimensions, then $Q$ can be expressed as a function of a single dimensionless combination of $a$, $b$, and $c$.

Application to Kinematic Relationships

Consider a rolling object. The motion of a rigid body combines translation of the center of mass with rotation about that center ^{[center-of-mass-motion]}. For rolling without slipping, the linear velocity $v$ of the center of mass and the angular velocity $\omega$ are related by ^{[rolling-without-slipping]}:

$v = r \omega$

We can verify this is dimensionally consistent:

• $[v] = [L T^{-1}]$
• $[r \omega] = [L] \cdot [T^{-1}] = [L T^{-1}]$ ✓

The form of this relationship—that $v$ is proportional to $\omega$ with the radius as the proportionality constant—can be anticipated through dimensional reasoning. If $v$ depends on $\omega$ and $r$, and we have two quantities with two independent dimensions ($[L]$ and $[T]$), then by the Buckingham Pi theorem, there is one dimensionless group. The only way to construct a dimensionless combination from $v$, $\omega$, and $r$ is $v / (r \omega)$, which must be a dimensionless constant. Thus $v = C \cdot r \omega$ for some dimensionless constant $C$. Physical reasoning (or detailed kinematics) determines that $C = 1$.

Worked Examples

Example 1: Period of a Pendulum

A simple pendulum's period $T$ depends on its length $\ell$, the mass $m$ of the bob, and gravitational acceleration $g$. What is the form of $T(\ell, m, g)$?

Dimensional analysis:

• $[T] = [T]$ (time)
• $[\ell] = [L]$
• $[m] = [M]$
• $[g] = [L T^{-2}]$

We have 4 quantities and 3 independent dimensions, so there is $4 - 3 = 1$ dimensionless group. Assume:

$T = k \cdot \ell^a m^b g^c$

where $k$ is dimensionless. Then:

$[T] = [L]^a [M]^b [L T^{-2}]^c = [L]^{a+c} [M]^b [T]^{-2c}$

Matching exponents:

• Length: $a + c = 0 \Rightarrow a = -c$
• Mass: $b = 0$
• Time: $-2c = 1 \Rightarrow c = -1/2$

Thus $a = 1/2$, and:

$T = k \sqrt{\frac{\ell}{g}}$

Dimensional analysis cannot determine $k$ (which equals $2\pi$ from detailed analysis), but it constrains the functional form powerfully: the period is independent of mass and proportional to the square root of the length-to-gravity ratio.

Example 2: Drag Force on a Sphere

A sphere moving through a fluid experiences drag. The drag force $F_d$ depends on the sphere's radius $r$, its velocity $v$, and the fluid's viscosity $\eta$ (with dimensions $[M L^{-1} T^{-1}]$). What form does $F_d$ take?

Dimensional analysis:

• $[F_d] = [M L T^{-2}]$
• $[r] = [L]$
• $[v] = [L T^{-1}]$
• $[\eta] = [M L^{-1} T^{-1}]$

We have 4 quantities and 3 dimensions, so 1 dimensionless group. Assume $F_d = k \cdot r^a v^b \eta^c$:

$[M L T^{-2}] = [L]^a [L T^{-1}]^b [M L^{-1} T^{-1}]^c = [M]^c [L]^{a+b-c} [T]^{-b-c}$

Matching exponents:

• Mass: $c = 1$
• Length: $a + b - c = 1 \Rightarrow a + b = 2$
• Time: $-b - c = -2 \Rightarrow b = 1$

Thus $a = 1$, and:

$F_d = k \cdot r v \eta$

This is Stokes's law for low-Reynolds-number flow. Again, dimensional analysis determines the exponents but not the numerical coefficient $k$ (which is $6\pi$ for a sphere).

References

^{[rolling-without-slipping] [center-of-mass-motion]}

AI Disclosure

This article was drafted with the assistance of an AI language model. The structure, synthesis, and presentation were guided by AI, though all factual claims are grounded in the cited class notes and standard physics pedagogy. The worked examples and dimensional reasoning are original applications of the principles documented in the source notes.