Module 1 · JEE Physics

Units, Measurement & Motion in a Straight Line

Updated 17 Nov 2026 · 7,022 words · 6 sections
Updated 21 Nov 2026
SI units, dimensional analysis, errors, and position, velocity and acceleration graphs.
Units · Dimensions · Errors · Kinematics
🎯 Take Mock Test

Learning Objectives

  • Understand the fundamental concepts of Units, Measurement & Motion in a Straight Line
  • Apply key formulas and techniques to solve problems
  • Practice with exam-level questions to build speed and accuracy

1. The International System of Units and Fundamental Quantities

The International System of Units, universally abbreviated as SI from the French Système International d'Unités, provides a coherent framework that links every measurable quantity to a small set of base units. The decision to adopt seven base quantities — length, mass, time, electric current, thermodynamic temperature, amount of substance, and luminous intensity — was driven by the need for a system in which derived units emerge algebraically without additional numerical factors. This coherence means that the unit of force, the newton, is exactly one kilogram metre per second squared, and the unit of energy, the joule, is exactly one newton metre, eliminating the conversion constants that plagued earlier systems such as the CGS or the British engineering system.

1.1 Base Units and Their Definitions

The metre is defined by fixing the numerical value of the speed of light in vacuum to be exactly 299 792 458 metres per second, which ties the unit of length to a fundamental constant of nature rather than a physical artefact. The kilogram, uniquely among the base units, was redefined in 2019 by fixing the Planck constant h to 6.626 070 15 × 10⁻³⁴ joule seconds, thereby linking mass to quantum mechanical action. The second is defined by the hyperfine transition frequency of the caesium‑133 atom, set at 9 192 631 770 cycles per second. The ampere follows from fixing the elementary charge e to 1.602 176 634 × 10⁻¹⁹ coulomb, the kelvin from fixing the Boltzmann constant k to 1.380 649 × 10⁻²³ joule per kelvin, the mole from fixing the Avogadro constant Nₐ to 6.022 140 76 × 10²³ per mole, and the candela from fixing the luminous efficacy of monochromatic radiation of frequency 540 × 10¹² hertz to 683 lumen per watt. Each definition is deliberately chosen so that the unit can be realised in any suitably equipped laboratory without reference to a unique prototype.

1.2 Derived Units and Coherent Relationships

Derived units are formed by multiplying and dividing base units according to the algebraic relations that define the corresponding physical quantities. Because the SI is coherent, the numerical factor in every such relation is exactly one. For example, pressure is force per unit area; since force is kilogram metre per second squared and area is metre squared, the pascal becomes kilogram per metre per second squared, with no extra multiplier. This property extends to electrical quantities: the volt is joule per coulomb, the ohm is volt per ampere, and the farad is coulomb per volt. The coherence simplifies dimensional checking, because any equation that is dimensionally consistent in SI will also be numerically consistent provided the same units are used throughout.

1.3 Supplementary Units and the Role of the Radian and Steradian

The radian and the steradian were formerly classified as supplementary units but are now treated as dimensionless derived units. The radian is defined as the ratio of the length of an arc to its radius, yielding a pure number; similarly, the steradian is the ratio of the area of a spherical surface patch to the square of the sphere's radius. Their dimensionless nature means they can be inserted into or omitted from dimensional equations without affecting homogeneity, yet they carry essential geometric information that distinguishes angular measure from ordinary ratios.

2. Dimensional Analysis and Its Applications

Dimensional analysis rests on the principle that any physically meaningful equation must have the same dimensions on both sides, a requirement known as the principle of homogeneity of dimensions. This principle follows from the fact that physical laws cannot depend on the arbitrary choice of units; if an equation were to equate a length to a time, a change of units would alter the numerical equality, contradicting the universality of the law. Consequently, dimensional analysis serves both as a sanity check for derived formulas and as a tool for discovering plausible functional forms when the underlying physics is only partially known.

2.1 Principle of Homogeneity of Dimensions

Every term in an equation must possess identical dimensional formulae. For instance, in the equation for the period T of a simple pendulum, T = 2π √(L/g), the dimensions of the right‑hand side are √(length / (length · time⁻²)) = √(time²) = time, matching the dimension of T. The constant 2π is dimensionless, as are all pure numbers and trigonometric functions of dimensionless arguments. This rule immediately flags equations such as v = u + at² as impossible, because the term at² carries dimensions of length · time⁻¹, not velocity.

2.2 Checking Equations and Deriving Relations

When a student proposes a formula for the viscous drag force F on a sphere moving through a fluid, dimensional reasoning suggests that F must depend on the fluid's viscosity η (dimensions M L⁻¹ T⁻¹), the sphere's radius r (L), and its velocity v (L T⁻¹). The only combination yielding the dimensions of force (M L T⁻²) is η r v, leading to the well‑known Stokes law F = 6π η r v, where the dimensionless factor 6π cannot be fixed by dimensions alone. Similarly, the Rayleigh‑Plesset equation for bubble dynamics can be examined term by term to ensure each additive component shares the dimension of pressure.

2.3 Limitations of Dimensional Analysis

Dimensional analysis cannot determine dimensionless constants, nor can it distinguish between multiple dimensionless groups that may appear in a complete description. For example, the drag coefficient Cᴅ in the equation F = ½ ρ v² A Cᴅ is dimensionless; its value depends on the Reynolds number Re = ρ v L/η, another dimensionless group. Without experimental data or a deeper theoretical model, dimensional analysis alone cannot predict how Cᴅ varies with Re. Moreover, it cannot handle equations involving transcendental functions of dimensional arguments, because such functions are only defined for dimensionless inputs.

Example: Dimensional Consistency of the Kinematic Equation
A student writes the displacement equation for uniformly accelerated motion as s = ut + ½ a t³. Use dimensional analysis to decide whether this equation can be correct.
The left‑hand side s has dimension of length L. The term ut has dimensions (L T⁻¹)·T = L, which is consistent. The term ½ a t³ has dimensions (L T⁻²)·T³ = L T, which is not a length. Because the two terms on the right have different dimensions, they cannot be added, and the equation violates the principle of homogeneity. The correct term is ½ a t², giving dimensions (L T⁻²)·T² = L. Therefore the proposed equation is dimensionally incorrect.
Exam‑Day Insight
When a multiple‑choice question asks you to identify the dimensionally correct formula among several options, first write down the dimensions of each variable using the base symbols M, L, T, I, Θ, N, J. Then test each term; the correct option will be the only one where every additive term shares the same dimensional formula. This technique eliminates algebraic errors that might otherwise go unnoticed.

3. Errors in Measurement and Their Propagation

Every experimental result carries an uncertainty that reflects the limitations of the measuring instrument, the observer, and the environment. Understanding the nature of these uncertainties and how they combine when quantities are manipulated mathematically is essential for reporting results with honest confidence intervals. The classification of errors into systematic and random categories, the quantification of absolute, relative, and percentage errors, and the rules for propagating uncertainties through addition, multiplication, and functional relationships form the core of this topic.

3.1 Types of Errors: Systematic and Random

Systematic errors arise from a consistent bias in the measurement process, such as a zero error in a vernier caliper, an incorrectly calibrated thermometer, or a persistent parallax when reading a scale. Because they shift all readings in the same direction, systematic errors do not diminish with repeated measurements; they must be identified and corrected, often by calibrating against a known standard. Random errors, by contrast, stem from unpredictable fluctuations — thermal noise in an electronic circuit, slight variations in human reaction time, or minute vibrations of the apparatus. These errors follow a statistical distribution, typically Gaussian, and their effect can be reduced by averaging a large number of independent observations, since the standard error of the mean decreases as 1/√n.

3.2 Absolute, Relative and Percentage Errors

The absolute error Δx of a measured quantity x is the magnitude of the difference between the measured value and the true (or accepted) value. The relative error is the ratio Δx / x, a dimensionless quantity that expresses the error as a fraction of the measured magnitude. Multiplying the relative error by 100 yields the percentage error, which is often used in reporting because it is intuitive. For example, a length measured as 12.0 cm with an absolute error of 0.1 cm has a relative error of 0.0083 and a percentage error of 0.83 %. When the true value is unknown, the best estimate of the absolute error is taken from the instrument's least count or from the standard deviation of repeated readings.

3.3 Propagation of Errors in Arithmetic Operations

When two measured quantities A and B are combined, the uncertainty in the result depends on the operation. For addition or subtraction, the absolute errors add in quadrature: Δ(A ± B) = √(ΔA² + ΔB²). For multiplication or division, the relative errors add in quadrature: Δ(A·B) / (A·B) = √[(ΔA/A)² + (ΔB/B)²] and similarly for A/B. If a quantity is raised to a power n, the relative error is multiplied by |n|: Δ(Aⁿ) / Aⁿ = |n|·ΔA/A. These rules assume the errors are independent and random; correlated errors require a more general covariance treatment.

3.4 Significant Figures and Rounding Rules

Significant figures provide a quick, albeit approximate, way to convey the precision of a measurement without explicitly stating an error bar. All non‑zero digits are significant; zeros between non‑zero digits are significant; leading zeros are not; trailing zeros in a number with a decimal point are significant. When performing calculations, the result should be rounded to the same number of significant figures as the least precise factor in multiplication or division, or to the same decimal place as the least precise term in addition or subtraction. Rounding follows the usual rule: if the digit to be dropped is 5 or greater, the preceding digit is increased by one; otherwise it is left unchanged. Over‑rounding intermediate steps can accumulate error, so it is advisable to keep extra guard digits until the final result.

Comparison of Systematic and Random Errors
AspectSystematic ErrorRandom Error
OriginInstrument bias, calibration fault, fixed procedural flawUnpredictable fluctuations, noise, human variability
Effect on Repeated MeasurementsAll readings shifted by same amountReadings scatter around a mean
Reduction MethodCalibration, correction factor, improved techniqueIncrease number of observations, statistical averaging
Statistical TreatmentNot reduced by √n averagingStandard error decreases as 1/√n

4. Kinematics of Motion in a Straight Line

Kinematics describes the geometry of motion without reference to the forces that cause it. In one dimension, the motion of a particle is completely specified by its position coordinate x as a function of time t. From this fundamental function we derive displacement, velocity, and acceleration, each providing a different perspective on how the particle moves. Graphical representations — position‑time, velocity‑time, and acceleration‑time curves — offer intuitive insight and are indispensable for solving problems where the motion is piecewise defined.

4.1 Position, Displacement and Distance

The position vector in one dimension reduces to a scalar coordinate x measured from an chosen origin. Displacement Δx = x₂ − x₁ is a vector quantity (in one dimension it carries a sign) that points from the initial to the final position. Distance travelled, on the other hand, is the total length of the path covered and is always non‑negative. For a particle that moves forward 5 m, then backward 3 m, the displacement is +2 m while the distance travelled is 8 m. This distinction becomes crucial when interpreting the area under a velocity‑time graph: the signed area gives displacement, whereas the total area (ignoring sign) gives distance.

4.2 Velocity: Average and Instantaneous

Average velocity over a time interval Δt is defined as the displacement divided by the interval, v̄ = Δx/Δt. It is a vector whose direction is that of the displacement. Instantaneous velocity v(t) is the limit of the average velocity as Δt → 0, mathematically the derivative dx/dt. Graphically, the instantaneous velocity at any instant equals the slope of the tangent to the position‑time curve at that point. If the position‑time graph is a straight line, the slope is constant and the velocity is uniform; curvature indicates changing velocity. The sign of the velocity tells whether the particle moves in the positive or negative x direction.

4.3 Acceleration: Uniform and Non‑Uniform

Acceleration is the rate of change of velocity. Average acceleration ā = Δv/Δt, while instantaneous acceleration a(t) = dv/dt = d²x/dt². Uniform acceleration means a is constant; the velocity‑time graph is then a straight line with slope a, and the position‑time graph is a parabola. Non‑uniform acceleration produces a curved velocity‑time graph; the instantaneous acceleration at any moment is the slope of the tangent to that curve. A common example is a particle subject to a drag force proportional to velocity, where a = −k v, leading to an exponential decay of velocity.

4.4 Graphical Representation of Motion

The three fundamental graphs are interrelated through differentiation and integration. The slope of the x‑t graph gives the v‑t graph; the slope of the v‑t graph gives the a‑t graph. Conversely, the area under the a‑t graph between two times gives the change in velocity, and the area under the v‑t graph gives the displacement. For a particle thrown vertically upward with initial velocity u under gravity g, the v‑t graph is a straight line descending from u to −u, the a‑t graph is a horizontal line at −g, and the x‑t graph is a symmetric parabola. These graphical tools allow one to solve problems such as finding the maximum height (the vertex of the parabola) or the total time of flight (the intercepts of the v‑t line with the time axis) without writing a single algebraic equation.

Comparison of Average and Instantaneous Quantities
QuantityAverage DefinitionInstantaneous Definition
VelocityΔx / Δt over a finite intervaldx/dt, slope of x‑t tangent
AccelerationΔv / Δt over a finite intervaldv/dt = d²x/dt², slope of v‑t tangent
Graphical MeaningSecant line slopeTangent line slope
Physical SignificanceOverall trend over intervalState at a precise instant

5. Equations of Motion for Uniform Acceleration

When acceleration is constant, the kinematic variables are linked by a set of algebraic equations that can be derived directly from the velocity‑time graph. These equations are the workhorse of one‑dimensional dynamics and appear in countless problems involving free fall, projectile motion along a line, and vehicle braking. Mastery of their derivation, domain of validity, and algebraic manipulation is essential for rapid problem solving.

5.1 Derivation from Velocity‑Time Graph

For uniform acceleration a, the velocity‑time graph is a straight line v = u + a t, where u is the initial velocity at t = 0. The displacement s is the area under this line from 0 to t, which is the area of a rectangle (u t) plus a triangle (½ a t²), giving s = u t + ½ a t². Eliminating t between v = u + a t and s = u t + ½ a t² yields v² = u² + 2 a s. A third form, s = (u + v) t / 2, follows from the fact that the area of a trapezoid equals the average of the parallel sides times the height. These three equations — v = u + a t, s = u t + ½ a t², v² = u² + 2 a s — are completely equivalent and can be used interchangeably depending on which variables are known.

5.2 Application to Free Fall and Projectile Motion (One‑Dimensional)

In free fall near Earth's surface, the acceleration is a = −g, where g ≈ 9.8 m s⁻² directed downward. Taking upward as positive, a projectile launched vertically with initial speed u reaches maximum height when v = 0, giving 0 = u − g tₘₐₓ, so tₘₐₓ = u/g. The maximum height is sₘₐₓ = u²/(2g). The total time of flight until the projectile returns to the launch level is 2u/g. If the launch point is at height h above the ground, the equation s = u t − ½ g t² with s = −h yields a quadratic for the impact time. These results assume no air resistance; including a drag force proportional to velocity makes the acceleration non‑uniform and the simple equations no longer apply.

5.3 Relative Motion in One Dimension

When two particles move along the same line, their relative velocity is the difference of their velocities: v_rel = v₁ − v₂. If both have constant accelerations a₁ and a₂, the relative acceleration is a_rel = a₁ − a₂. The relative motion can therefore be analysed by treating one particle as stationary and assigning the other the relative kinematic quantities. For example, a car A moving at 20 m s⁻¹ with acceleration 2 m s⁻² overtakes car B moving at 15 m s⁻¹ with acceleration 1 m s⁻². The initial relative velocity is 5 m s⁻¹ and the relative acceleration is 1 m s⁻². The time needed for the separation to increase by 100 m is found from 100 = 5 t + ½·1·t², giving t = 10 s. This technique simplifies many pursuit and meeting problems.

6. Practical Problem Solving and Common Pitfalls

Translating a word problem into a mathematical model requires a systematic approach: identify the knowns and unknowns, choose a coordinate system, list the applicable equations, and check dimensional consistency at each step. Many errors arise from sign conventions, confusion between distance and displacement, or misuse of the equations of motion outside their domain of constant acceleration. Developing a habit of verifying the final answer against physical intuition — such as ensuring a time is positive and a speed does not exceed the speed of light in classical contexts — catches a large fraction of mistakes.

6.1 Strategy for Solving Numerical Problems

Begin by drawing a clear diagram showing the origin, positive direction, and all relevant positions, velocities, and accelerations. Write down every given quantity with its symbol and unit. Next, decide which of the three constant‑acceleration equations contains the knowns and the desired unknown; if more than one unknown appears, combine two equations to eliminate the extra variable. Solve algebraically before substituting numbers, which reduces rounding errors and makes dimensional checking easier. Finally, insert the numerical values, compute the result, and express it with the appropriate number of significant figures.

6.2 Dimensional Consistency Check

After obtaining an algebraic expression, replace each symbol by its dimension (M, L, T, etc.) and verify that both sides match. For instance, if the derived formula for the range of a projectile launched from height h is R = u √(2h/g), the dimensions are (L T⁻¹)·√(L / (L T⁻²)) = (L T⁻¹)·√(T²) = L, which is correct for a range. A mismatch signals an algebraic slip, such as a missing square root or an extra factor of time. This quick test is invaluable during examinations when time is limited.

6.3 Error Estimation in Final Results

When the input data carry uncertainties, the final answer should be accompanied by an error estimate. Using the propagation rules from Section 3, compute the absolute or relative error of the result. For a derived quantity like g = 2h/t² obtained from a free‑fall experiment, the relative error in g is Δg/g = √[(Δh/h)² + (2 Δt/t)²]. If h = 1.00 m ± 0.01 m and t = 0.45 s ± 0.01 s, then Δg/g ≈ √[(0.01)² + (2·0.01/0.45)²] ≈ 0.045, so g = 9.88 m s⁻² ± 0.44 m s⁻². Reporting the result with its uncertainty demonstrates a complete understanding of the measurement process.

Mastery of units, dimensions, error analysis, and one‑dimensional kinematics provides the foundation upon which all further physics is built, enabling precise description and prediction of the physical world.

Calculus Based Kinematics for Non‑Uniform Acceleration

When acceleration varies with time, the constant‑acceleration equations no longer apply. The fundamental definitions remain: velocity is the time derivative of position, v(t) = dx/dt, and acceleration is the time derivative of velocity, a(t) = dv/dt. Integrating the acceleration function yields velocity, and a second integration yields position, each step introducing an integration constant fixed by initial conditions. For a given a(t) = a₀ + kt, the velocity becomes v(t) = v₀ + a₀t + ½kt² and the position x(t) = x₀ + v₀t + ½a₀t² + ⅙kt³. This polynomial structure appears frequently in problems with linearly changing forces, such as a rocket losing mass at a constant rate while thrust remains constant.

Derivation of Velocity and Position from Acceleration Function

Start with a(t) expressed analytically. Perform the indefinite integral ∫a(t) dt = v(t) + C₁. Apply the initial velocity v(0) = v₀ to determine C₁. Next integrate v(t) to obtain x(t) = ∫v(t) dt + C₂, using x(0) = x₀ for C₂. If the acceleration is given piecewise, integrate each segment separately and enforce continuity of velocity and position at the segment boundaries. This method automatically handles sign changes in acceleration, which correspond to turning points in the velocity‑time graph.

Example: Motion with Linear Drag

A particle of mass m falls under gravity with a drag force proportional to velocity, F₍drag₎ = −bv. The equation of motion is m dv/dt = mg − bv. Rearranging gives dv/dt + (b/m)v = g. This first‑order linear differential equation has the integrating factor exp((b/m)t). Solving yields v(t) = (mg/b)(1 − e−bt/m) for a particle released from rest. The terminal velocity is vₜ = mg/b. Integrating once more gives the position y(t) = vₜ[t + (m/b)(e−bt/m − 1)]. These expressions reduce to the constant‑acceleration formulas when b → 0, providing a useful consistency check.

Dimensional Analysis in Scaling Laws and Similarity

Beyond checking homogeneity, dimensional analysis can predict the functional form of physical relationships when the governing equations are unknown. The Buckingham π theorem states that a physical problem involving n variables and k fundamental dimensions can be reduced to n − k independent dimensionless groups (π terms). Any complete equation can be written as a relationship among these π groups. This principle underlies model testing in fluid dynamics, where a small‑scale model and a full‑scale prototype share the same dimensionless numbers (Reynolds, Froude, Mach) to ensure dynamic similarity.

Buckingham π Theorem

Identify all relevant variables (e.g., drag force F, density ρ, viscosity μ, velocity v, characteristic length L). Write each variable in terms of M, L, T. Choose k repeating variables that together contain all fundamental dimensions (commonly ρ, v, L). Form π groups by combining the remaining variables with powers of the repeating variables to make each group dimensionless. For drag, π₁ = F/(ρv²L²) and π₂ = μ/(ρvL) = 1/Re. The functional relationship π₁ = f(π₂) is the dimensionless drag coefficient as a function of Reynolds number.

Application to Fluid Flow and Pendulum Period

For a simple pendulum, the period T may depend on length L, mass m, gravitational acceleration g, and amplitude θ₀. Dimensions: [T] = T, [L] = L, [m] = M, [g] = L T⁻², [θ₀] = dimensionless. With k = 3 (M, L, T) and n = 5, we obtain two π groups: π₁ = T√(g/L) and π₂ = θ₀. Hence T = √(L/g) f(θ₀). The exact solution gives f(θ₀) = 2π[1 + ¼ sin²(θ₀/2) + …], confirming the dimensional prediction and showing how the amplitude correction enters only through a dimensionless function.

Worked Example: Determining the Period of a Simple Pendulum Using Dimensional Analysis
Using only the dimensions of length L, mass M, time T, and gravitational acceleration g, find the functional dependence of the period T on L and g.
Assume T = k L^a g^b where k is dimensionless. Dimensions: [T] = L^a (L T⁻²)^b = La+b T-2b. Equating exponents: for T: 1 = -2b → b = -1/2. For L: 0 = a + b → a = 1/2. Hence T = k √(L/g). The constant k = 2π is obtained from the exact solution of the differential equation.
Exam‑Day Strategy
When a question asks for a relationship between physical quantities, first list the dimensions of each quantity. Write a general product of powers, equate exponents, and solve the resulting linear system. This yields the correct powers even if the exact constant is unknown, earning partial credit.

Error Propagation in Graphical Determination of Acceleration

Graphical methods are common in laboratory examinations. The acceleration is often obtained from the slope of a velocity‑time graph or the curvature of a position‑time graph. Each data point carries an uncertainty, and the slope uncertainty must be propagated correctly. For a straight‑line fit v = u + at, the standard error of the slope a is σa = σy / √(∑(tᵢ − t̄)²), where σy is the standard deviation of the residuals. If the time measurements also have uncertainty, a total‑least‑squares (orthogonal distance regression) approach is required, which accounts for errors in both axes.

Uncertainty in Slope of Velocity‑Time Graph

Given N measurements (tᵢ, vᵢ) with identical vertical uncertainty Δv, the best‑fit slope a and its standard error are obtained from the usual least‑squares formulas. The relative error in acceleration is Δa/a = σa/|a|. When the data points are not equally spaced in time, the denominator ∑(tᵢ − t̄)² becomes smaller, increasing the slope uncertainty. This highlights the importance of choosing measurement times that maximise the spread of the independent variable.

Weighted Least‑Squares Fit for Non‑Uniform Data

If individual velocity measurements have different uncertainties Δvᵢ, a weighted fit minimises χ² = ∑[(vᵢ − u − atᵢ)² / Δvᵢ²]. The resulting slope variance is σa² = 1 / ∑[wᵢ(tᵢ − t̄w)²], where wᵢ = 1/Δvᵢ² and t̄w = ∑wᵢtᵢ / ∑wᵢ. This procedure gives more influence to precise points and yields a realistic error estimate that can be compared with the theoretical prediction.

Redefinition of SI Base Units and Its Impact on Measurements

In 2019 the General Conference on Weights and Measures adopted definitions of the kilogram, ampere, kelvin, and mole based on fixed values of fundamental constants: the Planck constant h, the elementary charge e, the Boltzmann constant kB, and the Avogadro constant NA. The metre and second were already defined via the speed of light c and the caesium hyperfine transition frequency ΔνCs. This shift makes the SI system entirely constant‑based, eliminating artefacts such as the International Prototype Kilogram.

2019 Redefinition of Kilogram, Ampere, Kelvin, Mole

The kilogram is now defined by fixing h = 6.62607015 × 10⁻³⁴ J·s exactly. The ampere follows from fixing e = 1.602176634 × 10⁻¹⁹ C. The kelvin uses kB = 1.380649 × 10⁻²³ J·K⁻¹, and the mole uses NA = 6.02214076 × 10²³ mol⁻¹. These exact values have zero uncertainty by definition; the uncertainty now resides in the realization experiments (e.g., Kibble balance for mass, Josephson and quantum Hall effects for voltage and resistance).

Practical Consequences for Laboratory Work

For most undergraduate experiments the numerical values of the constants remain unchanged, so existing data and formulas are unaffected. However, the uncertainty budget of a measurement traceable to the SI must now include the realization uncertainty of the relevant constant. For example, a mass measured with a Kibble balance inherits the relative uncertainty of the Planck constant realization, currently a few parts in 10⁸. In contrast, a mass calibrated against a stainless‑steel kilogram prototype carries the additional uncertainty of the prototype’s stability. Understanding this hierarchy is essential when a problem asks for the combined standard uncertainty of a derived quantity such as the gravitational constant G measured via a torsion balance.

Relative Motion with Time‑Dependent Accelerations

When both particles experience accelerations that vary with time, the relative acceleration is simply the difference of the instantaneous accelerations: arel(t) = a₁(t) − a₂(t). The relative velocity and separation are obtained by integrating arel(t) with appropriate initial conditions. This approach reduces a two‑body pursuit problem to a single‑body problem in a non‑inertial frame attached to one particle, provided the frame’s acceleration is accounted for.

Relative Acceleration as Difference of Instantaneous Accelerations

Let particle 1 have position x₁(t) and particle 2 have position x₂(t). The separation s(t) = x₁(t) − x₂(t). Differentiating twice gives d²s/dt² = a₁(t) − a₂(t). If a₁(t) = αt and a₂(t) = β (constant), then arel(t) = αt − β. Integrating: vrel(t) = vrel(0) + ½αt² − βt, and s(t) = s(0) + vrel(0)t + ⅙αt³ − ½βt². The meeting condition s(t) = 0 yields a cubic equation whose real positive root gives the encounter time.

Solving Pursuit Problems with Variable Acceleration

Consider a police car starting from rest with acceleration ap(t) = a₀(1 − e−t/τ) chasing a speeder moving at constant velocity vs. The relative acceleration is arel(t) = a₀(1 − e−t/τ). Integrating twice with zero initial relative velocity and initial separation s₀ gives s(t) = s₀ − vst + a₀[t²/2 − τt + τ²(1 − e−t/τ)]. The capture time satisfies s(t) = 0. This transcendental equation is solved numerically, but the dimensional form shows that the only dimensionless groups are a₀τ/vs and s₀/(vsτ). Recognising these groups allows a quick estimate of whether capture occurs before the police car reaches its terminal acceleration.

  • Assuming constant‑acceleration formulas when the problem explicitly states a time‑varying force or drag.
  • Neglecting the integration constants when deriving velocity or position from a given acceleration function.
  • Using the simple slope‑error formula for a velocity‑time graph when the time measurements have non‑negligible uncertainties.
  • Treating the 2019 SI redefinition as a change in numerical values of constants rather than a change in the uncertainty hierarchy.
  • Forgetting that the relative acceleration is the difference of instantaneous accelerations, not the difference of average accelerations over an interval.
  • Applying dimensional analysis to a problem that already contains a dimensionless constant (e.g., 2π) and expecting the method to produce that constant.
  • Misidentifying the repeating variables in the Buckingham π theorem, leading to an incorrect set of dimensionless groups.

Non‑Inertial Reference Frames and Pseudo Forces in One Dimension

When the observer’s frame accelerates with acceleration aframe relative to an inertial frame, Newton’s second law acquires an additional term. For a particle of mass m the equation of motion in the non‑inertial frame becomes m arel = Freal − m aframe, where arel is the acceleration measured in the accelerating frame and Freal is the sum of all physical forces (gravity, tension, friction, etc.). The term −m aframe is the pseudo force; it acts on every mass proportionally to its inertia and is directed opposite to the frame’s acceleration. In one dimension the sign convention is crucial: if the frame accelerates in the positive x direction, the pseudo force points in the negative x direction. A classic illustration is a block on a frictionless horizontal surface inside a truck that accelerates forward with aframe = 2 m s⁻². In the truck’s frame the block experiences a pseudo force −m·2 m s⁻² and therefore accelerates backward relative to the truck with arel = −2 m s⁻², even though no real horizontal force acts on it. When the frame’s acceleration is time‑dependent, aframe(t), the pseudo force becomes a known function of time and the relative motion is obtained by integrating arel(t) = Freal(t)/m − aframe(t). This technique reduces many pursuit or elevator problems to a single‑body problem in a non‑inertial frame, provided the initial conditions are transformed correctly. A common mistake is to treat the pseudo force as a reaction force that obeys Newton’s third law; it does not have an action‑reaction pair because it originates from the frame’s acceleration, not from an interaction with another body.

Variable Mass Systems: Rocket Equation and Thrust

Systems that gain or lose mass while moving require a modified momentum balance. Consider a rocket of instantaneous mass m(t) ejecting exhaust at a constant relative speed u (measured in the rocket frame) with a mass flow rate ṁ = −dm/dt > 0. In an inertial frame the momentum of the rocket plus expelled mass is conserved if external forces are neglected. Differentiating the total momentum gives m dv/dt = u ṁ + Fext. The term u ṁ is the thrust; it is positive when mass is ejected backward (ṁ > 0, u > 0). Integrating with Fext = 0 yields the Tsiolkovsky rocket equation: Δv = u ln(m0/mf), where m0 and mf are the initial and final masses. If gravity acts, Fext = −mg, the velocity after burnout becomes v = u ln(m0/mf) − g tburn for constant g. When the exhaust velocity varies, the integral generalises to Δv = ∫ u(t) d(ln m). A frequent pitfall is to write F = dp/dt = m dv/dt + v dm/dt and then set F = 0 for a free rocket; this incorrectly treats the momentum of the ejected mass as part of the rocket’s momentum. The correct approach isolates the rocket’s momentum and accounts for the momentum carried away by the exhaust. Dimensional analysis confirms the thrust term: [u ṁ] = (m s⁻¹)(kg s⁻¹) = kg m s⁻² = N. In problems where mass is accreted (e.g., a raindrop falling through a cloud), the sign of ṁ reverses and the thrust term opposes the motion, leading to a terminal velocity that depends on the accretion rate.

Worked Example: Rocket with Constant Exhaust Velocity
A rocket of initial mass 1200 kg (including 800 kg of fuel) ejects gas at a constant speed of 2500 m s⁻¹ relative to the rocket. The fuel burns at a constant rate of 40 kg s⁻¹. Neglect gravity and air resistance. Find the velocity of the rocket when the fuel is exhausted.
The mass flow rate is ṁ = 40 kg s⁻¹, so burn time tb = 800 kg / 40 kg s⁻¹ = 20 s. The final mass mf = 1200 kg − 800 kg = 400 kg. Using the Tsiolkovsky equation Δv = u ln(m0/mf) with u = 2500 m s⁻¹, m0 = 1200 kg, mf = 400 kg: Δv = 2500 × ln(1200/400) = 2500 × ln(3) ≈ 2500 × 1.0986 = 2746.5 m s⁻¹. The rocket’s velocity at burnout is therefore approximately 2.75 km s⁻¹.
Exam‑Day Strategy
When a variable‑mass problem appears, first identify whether mass is ejected or accreted and the sign of the mass flow rate. Write the thrust term as u |ṁ| with the correct direction, then apply the rocket equation or integrate the differential form. Check dimensions: thrust must have units of newton. If gravity or drag is present, include them as external forces in the momentum balance before integrating.

Numerical Integration of Equations of Motion: Euler and Runge‑Kutta Methods

Analytical solutions exist only for a limited class of acceleration functions. For arbitrary a(t, x, v) one must resort to numerical integration. The simplest scheme is the explicit Euler method: given xn, vn at time tn, compute an = a(tn, xn, vn), then vn+1 = vn + an Δt and xn+1 = xn + vn Δt. This first‑order method accumulates a global error proportional to Δt. The velocity‑Verlet (or symplectic Euler) variant updates position with the new velocity, improving energy conservation for conservative forces. A far more accurate workhorse is the classical fourth‑order Runge‑Kutta (RK4) method. For the system dx/dt = v, dv/dt = a(t, x, v), RK4 evaluates four slopes per step: k1v = a(tn, xn, vn), k1x = vn; k2v = a(tn+Δt/2, xn+k1xΔt/2, vn+k1vΔt/2), k2x = vn+k1vΔt/2; similarly k3 and k4. The updates are vn+1 = vn + (Δt/6)(k1v+2k2v+2k3v+k4v) and xn+1 = xn + (Δt/6)(k1x+2k2x+2k3x+k4x). RK4’s global error scales as Δt⁴, allowing much larger steps for a given accuracy. Adaptive step‑size control (e.g., embedded Runge‑Kutta pairs) estimates the local truncation error by comparing a fourth‑ and fifth‑order estimate and adjusts Δt to keep the error within a tolerance. In examinations, you may be asked to perform one or two steps of Euler or RK4 by hand; the key is to organise the intermediate values in a table and keep track of units. A common error is to update position using the old velocity after the velocity has already been updated; the correct order depends on the specific scheme (Euler uses old velocity, velocity‑Verlet uses new velocity).

Correlated Uncertainties and Covariance in Error Propagation

The standard propagation formulas σf² = Σ (∂f/∂xi)² σi² assume the input quantities xi are statistically independent. When measurements share a common systematic effect (e.g., the same voltmeter used for two voltage readings), their errors are correlated and the covariance term must be included: σf² = Σ (∂f/∂xi)² σi² + 2 Σi<j</sub> (∂f/∂xi)(∂f/∂xj) cov(xi, xj). The covariance cov(xi, xj) = ρij σi σj, where ρij is the correlation coefficient (−1 ≤ ρ ≤ 1). If two length measurements are made with the same ruler whose calibration error dominates, ρ ≈ 1 and the covariance term adds constructively, increasing the combined uncertainty. Conversely, if one measurement tends to be high when the other is low (ρ < 0), the covariance reduces the total uncertainty. In a least‑squares fit of a straight line y = mx + c to data with correlated y errors, the covariance matrix of the parameters (m, c) is obtained from (Xᵀ W X)⁻¹, where W is the weight matrix (inverse of the data covariance matrix). Ignoring off‑diagonal elements of W underestimates the parameter uncertainties when positive correlations exist. A practical rule: always examine the experimental setup for shared instruments, common environmental fluctuations, or common calibration standards; if present, estimate ρ from repeated calibrations or manufacturer specifications and include the covariance term. In graphical analysis, the slope and intercept of a fitted line are often strongly anti‑correlated (ρ ≈ −0.9); reporting only the individual standard errors without the covariance gives a misleading picture of the joint confidence region.

Relativistic Corrections to Classical Kinematics at High Speeds

Classical kinematics assumes velocities add linearly and kinetic energy is ½mv². When speeds approach a significant fraction of the speed of light c = 2.99792458 × 10⁸ m s⁻¹, these formulas break down. The Lorentz factor γ = 1/√(1 − v²/c²) modifies momentum to p = γmv and total energy to E = γmc². The kinetic energy becomes K = (γ − 1)mc², which reduces to ½mv² for v ≪ c via a Taylor expansion. Velocity addition in one dimension follows u' = (u − v)/(1 − uv/c²), where u is the particle’s velocity in the lab frame and v is the relative velocity of the moving frame. For a rocket ejecting mass at relativistic exhaust speed ue, the relativistic rocket equation is Δv = c tanh[(ue/c) ln(m0/mf)]. In particle accelerators, the time dilation factor γ extends the lifetime of fast muons, allowing them to reach the Earth’s surface despite their short proper lifetime. When solving problems that mix classical and relativistic regimes, identify the dimensionless parameter β = v/c. If β < 0.01, classical formulas are accurate to better than 0.01 %; if β > 0.1, relativistic corrections exceed 0.5 % and must be included. A frequent trap is to apply the classical work‑energy theorem W = ΔK with K = ½mv² at high speeds; the correct work‑energy relation uses the relativistic kinetic energy. Dimensional analysis still holds: γ is dimensionless, mc² has dimensions of energy, and the argument of the hyperbolic tangent is dimensionless.

  • Assuming pseudo forces obey Newton’s third law and come in action‑reaction pairs.
  • Using the classical rocket equation when the exhaust velocity is comparable to the speed of light.
  • Updating position with the new velocity in the explicit Euler scheme instead of the old velocity.
  • Neglecting covariance terms when two measured quantities share a common systematic error source.
  • Applying the classical kinetic energy formula ½mv² for speeds above 0.1 c without relativistic correction.
  • Treating the correlation coefficient as zero when the same instrument calibrates multiple readings.
  • Forgetting that the relativistic velocity addition formula reduces to simple subtraction only when both speeds are much smaller than c.

Practice Questions

← Modern Physics Motion in a Plane & Projectile Motion →