Dimensional Analysis and Units in Engineering Problems

A wrong unit can cost far more than a lost exam mark: in 1999 NASA lost the Mars Climate Orbiter because one team's software produced thrust data in pound-force seconds while another expected newton-seconds. Dimensional analysis is the engineer's first line of defence against such errors, and it is also a powerful tool for discovering how physical quantities relate before doing any detailed theory. This article covers both uses with fully worked problems.

Dimensions Versus Units

A dimension is a type of physical quantity; a unit is a particular way of measuring it. Length is a dimension; the metre, the foot and the kilometre are units of length.

In mechanics, everything can be built from three fundamental dimensions: mass \( [M] \), length \( [L] \) and time \( [T] \). Heat transfer adds temperature \( [\Theta] \), and electrical problems add current \( [I] \).

The Principle of Dimensional Homogeneity

Every term in a physically valid equation must have the same dimensions. You can add metres to metres but never metres to seconds. This gives a quick test of any formula you derive or half-remember.

Checking the Darcy-Weisbach equation

The head loss in a pipe is

$$ h_f = f\,\frac{L}{D}\,\frac{v^2}{2g} $$

The friction factor \( f \) is dimensionless and \( L/D \) is a length ratio, also dimensionless. The remaining term gives

$$ \frac{[v^2]}{[g]} = \frac{L^2T^{-2}}{LT^{-2}} = L $$

So \( h_f \) is a length, as a "head" in metres should be. The equation is homogeneous.

Remember that homogeneity is necessary but not sufficient: an equation can be dimensionally correct and still have a wrong numerical constant.

Unit Conversion Done Safely

The safest method is to multiply by fractions equal to one, so that unwanted units cancel visibly. Convert everything to base SI units before substituting into formulas.

Solved Problem 1: Sizing a Pump Motor

A pump lifts 50 L/s of water through a total head of 20 m. Find the hydraulic power in kW and in horsepower.

Step 1: Convert flow rate to SI.

$$ Q = 50\ \frac{\text{L}}{\text{s}} \times \frac{1\ \text{m}^3}{1000\ \text{L}} = 0.050\ \text{m}^3/\text{s} $$

Step 2: Apply the hydraulic power formula.

$$ P = \rho g Q H = (1000\ \text{kg/m}^3)(9.81\ \text{m/s}^2)(0.050\ \text{m}^3/\text{s})(20\ \text{m}) $$

Step 3: Track units.

$$ \text{kg·m}^{-3} \cdot \text{m·s}^{-2} \cdot \text{m}^3\text{s}^{-1} \cdot \text{m} = \text{kg·m}^2\text{s}^{-3} = \text{W} $$

$$ P = 9810\ \text{W} = 9.81\ \text{kW} $$

Step 4: Convert to horsepower (1 hp = 745.7 W):

$$ P = \frac{9810\ \text{W}}{745.7\ \text{W/hp}} = 13.2\ \text{hp} $$

Sanity check. The units collapsed to watts, confirming the formula. With a typical pump efficiency of about 70%, the motor would need roughly \( 9.81/0.7 = 14 \) kW, so a standard 15 kW motor would be chosen, which is a common size for small municipal or tube-well pumps.

Deriving Relationships: The Rayleigh Method

If you know which variables matter, dimensions alone can often reveal the form of the relationship.

Solved Problem 2: Period of a Pendulum

Suppose the period \( t \) of a simple pendulum depends on its length \( \ell \), the bob mass \( m \) and gravity \( g \). Find the form of the relationship.

Step 1: Assume a power law.

$$ t = C\,\ell^a m^b g^c $$

where \( C \) is a dimensionless constant.

Step 2: Write dimensions of both sides.

$$ T^1 = L^a\,M^b\,\left(LT^{-2}\right)^c = L^{a + c}M^bT^{-2c} $$

Step 3: Match exponents.

• Mass: \( b = 0 \)
• Time: \( -2c = 1 \), so \( c = -\tfrac{1}{2} \)
• Length: \( a + c = 0 \), so \( a = \tfrac{1}{2} \)

Step 4: Result.

$$ t = C\sqrt{\frac{\ell}{g}} $$

Sanity check. Full theory gives \( C = 2\pi \) for small swings, so dimensional analysis found everything except the constant. It also predicted, correctly, that mass does not matter. For a 1 m pendulum, \( t = 2\pi\sqrt{1/9.81} = 2.01 \) s, which is the familiar "seconds pendulum" of old clocks (one second per swing, two per full cycle).

The Buckingham Pi Theorem

When more variables are involved, the Buckingham Pi theorem states that a problem with \( n \) variables and \( k \) fundamental dimensions can be described by \( n - k \) dimensionless groups.

For example, the drag force \( F \) on a sphere depends on fluid density \( \rho \), velocity \( v \), diameter \( D \) and viscosity \( \mu \). That is 5 variables in 3 dimensions, so 2 groups:

$$ \frac{F}{\rho v^2 D^2} = \phi\left(\frac{\rho v D}{\mu}\right) $$

The second group is the famous Reynolds number. Instead of testing every combination of five variables, engineers need only one curve relating two groups.

Solved Problem 3: Ship Model Testing

A ship model is built at a scale of 1:25. The full-size ship will sail at 10 m/s. At what speed should the model be towed so that wave patterns are similar? (Wave effects are governed by the Froude number \( Fr = v/\sqrt{gL} \).)

Step 1: Equate Froude numbers.

$$ \frac{v_m}{\sqrt{gL_m}} = \frac{v_p}{\sqrt{gL_p}} \Rightarrow v_m = v_p\sqrt{\frac{L_m}{L_p}} $$

Step 2: Substitute.

$$ v_m = (10\ \text{m/s})\sqrt{\frac{1}{25}} = \frac{10}{5} = 2.0\ \text{m/s} $$

Sanity check. A smaller model must move more slowly to make waves of the same relative shape, which agrees with everyday observation of small boats. In practice the Reynolds numbers cannot also be matched, so naval architects correct the viscous part of resistance separately.

Did You Know?

• Edgar Buckingham published his Pi theorem in 1914, though similar ideas had been used by Lord Rayleigh and others before him.
• William Froude pioneered towing-tank tests of ship models in England in the 1870s.
• Angles are dimensionless (a radian is arc length divided by radius), which is why \( \sin \) and \( e^x \) need dimensionless arguments.

Key Takeaways

• Every term in a valid equation must have the same dimensions; use this to check formulas.
• Convert to base SI units before substituting, and track units through each step.
• The Rayleigh method finds power-law relationships up to a constant.
• Buckingham Pi reduces \( n \) variables to \( n - k \) dimensionless groups, the basis of model testing.

If unit conversions, dimensional analysis or model-scaling questions are giving you trouble, Amathify's subject experts and PhD scholars can provide step-by-step solutions via WhatsApp or the Ask a Question form.