UY1: Resistive Forces
A UY1 treatment of resistive forces: linear vs quadratic drag, terminal velocity, time constants, and velocity-time solutions.
Continue where you stopped
The core idea
On this page
Learning objectives
- Formulate motion and force models with explicit coordinates, assumptions, and units.
Resistive forces (drag) matter when an object moves through a fluid (air/water) and the fluid exerts a force opposite the motion. This lesson focuses on the standard 1D vertical-fall models and what they predict for v(t) and terminal speed.
At a glance
- Prerequisites: Newton’s 2nd law, vectors (Mathematics for Undergraduate Physics), First-order differential equations (optional refresher)
- You will learn: how to model drag as f ∝ v or f ∝ v²; how terminal speed emerges; what the time constant τ means; how to sanity-check v(t)
- Key results (downward positive): vₜ = mg/k (linear drag), vₜ = square root of (mg/b) (quadratic drag)
- Common trap: forgetting drag always opposes the velocity (sign errors), or swapping the meaning of C_D (dimensionless) with a lumped constant in bv²
Setup (model + sign convention)
These results are for idealized models. State your model before you calculate:
- 1D particle motion: vertical motion only; treat the object as a point mass.
- Inertial frame + uniform gravity: take g constant (near Earth, small height changes).
- Fluid at rest: drag depends on velocity relative to the fluid (wind/current changes the relative velocity).
- Constant parameters: treat k or b as constants (in reality they can vary with speed/orientation).
- Choose the drag law by regime: linear drag is typical at low Reynolds number (Stokes regime); quadratic drag is typical at higher Reynolds number.
We model the object as a particle falling vertically through a fluid.
Choose the + y axis downward, so:
- velocity v(t) ≥ 0 means “moving downward”,
- weight is + mg,
- drag is negative (upward) when the object moves downward.
Neglect buoyancy unless stated (we revisit it briefly at the end).
Drag always opposes the velocity relative to the fluid. With + y downward:
- if the object moves downward (v > 0), then F_drag < 0;
- if the object moves upward (v < 0), then F_drag > 0.
With signed v (downward positive), it is often cleanest to write drag laws in a way that automatically handles the sign:
- Linear drag: F_drag = -kv
- Quadratic drag: F_drag = -b v|v|
(If you know the motion stays downward, v ≥ 0 so v|v| = v² and the quadratic law becomes F_drag = -bv².)
Optional: when is drag ∝ v vs ∝ v^2? (Reynolds number)
A standard way to decide which drag model is appropriate is the Reynolds number: Re = (ρ v L)/η where ρ is fluid density, η is dynamic viscosity, and L is a characteristic length (e.g. object diameter).
- Creeping / Stokes regime (Re≪ 1): viscous effects dominate, and drag is approximately proportional to v.
- Inertial / high-Re regime (Re≫ 1): pressure/inertial effects dominate, and drag is approximately proportional to v².
In the transitional regime, the drag coefficient (and therefore the effective k or b) can vary with speed and orientation, so the “constant k / constant b” models become approximations.
1) Linear drag: F_drag = -kv
This model is often used for small objects at low speeds (or as a first approximation).
Governing equation
With signed v (downward positive), Newton’s 2nd law gives: mdv/dt = mg-kv.
Optional: where does linear drag come from? (Stokes’ law)
For a small sphere of radius r moving slowly through a viscous fluid, the Stokes drag is F_drag = 6πη r v (opposite the motion), where η is the dynamic viscosity. In that regime you can identify k = 6πη r.
Terminal speed
Terminal speed means the acceleration is zero: dv/dt = 0. mg-kvₜ = 0 ⇒ vₜ = mg/k.
Units check: since kv is a force, [k] = N/(m s⁻¹) = kg s⁻¹.
Solution for v(t) and the time constant
The ODE dv/dt = g-(k/m)v has a standard “relaxation” form. Since k/m has units of s⁻¹, define the time constant τ = m/k ⇒ k/m = 1/τ. Then the ODE becomes dv/dt = g-(1/τ)v.
Now use the terminal-speed relation vₜ = mg/k = gτ to rewrite it as: dv/dt = (vₜ-v)/τ. This makes the physics clear: the acceleration is proportional to the “gap” to terminal speed.
The general solution (with v(0) = v₀) is: v(t) = vₜ + (v₀-vₜ)e^(-t/τ), vₜ = mg/k, τ = m/k.
Optional derivation (integrating factor)
This is a first-order linear ODE: dv/dt + (1/τ)v = g.
The goal of the integrating-factor method is to multiply the equation by a function μ(t) so the left side becomes a product derivative: (d/dt)(μ v) = μdv/dt + μ'v. We want μ' = μ/τ, so choose μ(t) = e^(t/τ).
Start from dv/dt + (k/m)v = g. Multiply by the integrating factor e^((k/m)t) = e^(t/τ): e^(t/τ)dv/dt + (1/τ)e^(t/τ)v = g e^(t/τ). The left side is a product derivative: (d/dt)(ve^(t/τ)) = g e^(t/τ). Integrate from 0 to t and use v(0) = v₀: ve^(t/τ)-v₀ = gτ(e^(t/τ)-1). Solve for v(t): v(t) = gτ + (v₀-gτ)e^(-t/τ) = vₜ + (v₀-vₜ)e^(-t/τ).
Equivalent “time-constant” shortcut: rewrite the ODE as dv/dt = (vₜ-v)/τ. Let u = vₜ-v (the gap to terminal speed). Then du/dt = -dv/dt = -u/τ so u(t) = u(0)e^(-t/τ) = (vₜ-v₀)e^(-t/τ). Therefore v(t) = vₜ-(vₜ-v₀)e^(-t/τ) = vₜ + (v₀-vₜ)e^(-t/τ).
From rest (v₀ = 0): v(t) = vₜ(1-e^(-t/τ)).
This is an exponential approach to terminal speed:
- At t = 0, v = 0 and the initial slope is .dv/dt|ₜ₌₀ = g (drag is initially zero).
- The “gap to terminal speed” shrinks exponentially: vₜ-v(t) = vₜ e^(-t/τ).
Interpretation: τ sets the approach rate to terminal speed. For a release from rest: v(τ) = vₜ(1-e⁻¹) ≈ 0.63 vₜ.
More generally, from rest the time to reach a fraction p of terminal speed is: v = pvₜ ⇒ t = -τ ln(1-p). (e.g. p = 0.95 gives t ≈ 3.00 τ.)
Rule of thumb (from rest):
- t = τ: v ≈ 0.63 vₜ
- t = 2τ: v ≈ 0.86 vₜ
- t = 3τ: v ≈ 0.95 vₜ
- t = 5τ: v ≈ 0.99 vₜ
How parameters affect the approach:
- Larger m increases both vₜ and τ (heavier objects fall faster, but “relax” more slowly in time).
- Larger k decreases both vₜ and τ (stronger drag lowers the terminal speed and is reached sooner).
Optional: displacement y(t) for linear drag
If y is the downward coordinate and y(0) = y₀, integrate v(t) to get y(t) = y₀ + vₜ t + (v₀-vₜ)τ(1-e^(-t/τ)). In particular, from rest (v₀ = 0): y(t) = y₀ + vₜ[t-τ(1-e^(-t/τ))].
Checks (non-optional)
- t → 0: e^(-t/τ) ≈ 1-t/τ so v ≈ vₜ(t/τ) = gt (free fall initially).
- t → ∞: e^(-t/τ) → 0 so v → vₜ (terminal speed).
2) Quadratic drag: F_drag = -b v|v|
This model is typical for larger objects moving faster through air (e.g. a skydiver).
Model and parameter meanings
For signed v (downward positive), use: mdv/dt = mg-b v|v|.
If the motion stays downward (v ≥ 0), this reduces to: mdv/dt = mg-bv².
A common microscopic form is: F_drag = (1/2)ρ C_D A v², so in this notation: b = (1/2)ρ C_D A, where ρ is fluid density, A is cross-sectional area, and C_D is the (dimensionless) drag coefficient.
Units check: since bv² is a force, [b] = N/(m²s⁻²) = kg m⁻¹.
Terminal speed
Set dv/dt = 0: mg-bvₜ² = 0 ⇒ vₜ = square root of (mg/b) = square root of (2mg/(ρ C_D A)) .
Solution for v(t) (from rest)
With v(0) = 0: v(t) = vₜ tanh(gt/vₜ).
Intuition: the net downward force is mg-bv², so as v grows the resistive term grows rapidly, reducing the acceleration until it tends to zero at v = vₜ.
If you differentiate the solution you can see the acceleration decay explicitly: a(t) = dv/dt = g sech² (gt/vₜ).
The dimensionless time variable here is gt/vₜ, so a natural time scale is τ_q = vₜ/g. At t = τ_q: v(τ_q) = vₜ tanh(1) ≈ 0.76 vₜ.
More generally, from rest: v = pvₜ ⇒ p = tanh(gt/vₜ) ⇒ t = (vₜ/g)artanh(p).
Optional derivation (short)
Separate variables (assuming downward-only motion so v ≥ 0 and the equation is m dv/dt = mg-bv²): dv/(g-(b/m)v²) = dt.
Define the terminal speed: vₜ² = mg/b ⇒ g-(b/m)v² = g(1-v²/vₜ²). Then: dv/(g(1-v²/vₜ²)) = dt.
Substitute u = v/vₜ so dv = vₜ du: ∫ du/(1-u²) = ∫ (g dt)/vₜ.
The left integral is an inverse hyperbolic tangent: artanh(u) = gt/vₜ + C.
Apply the initial condition v(0) = 0 (so u(0) = 0) to get C = 0: artanh (v/vₜ) = gt/vₜ.
Finally apply tanh to both sides: v/vₜ = tanh(gt/vₜ) ⇒ v(t) = vₜ tanh(gt/vₜ).
Optional: displacement y(t) for quadratic drag (from rest)
From rest, integrate v(t) = vₜ tanh(gt/vₜ): y(t) = y₀ + vₜ²/g ln cosh(gt/vₜ). (Check: for small t, ln cosh x ≈ x²/2 so y-y₀ ≈ (1/2)gt² initially.)
Checks
- t → 0: tanh(x) ≈ x so v ≈ gt (free fall initially).
- t → ∞: tanh(x) → 1 so v → vₜ.
- Scaling: if you double A (same m), then vₜ decreases by 1/square root of 2.
Worked examples
Example: linear drag time constant
Problem
- Given: m = 0.020 kg, k = 0.080 kg s⁻¹, g = 9.80 m s⁻², released from rest
- Find: τ, vₜ, and v(0.50 s)
Approach
- Use the linear-drag model v(t) = vₜ(1-e^(-t/τ)) with τ = m/k and vₜ = mg/k.
Working
- Time constant: τ = m/k = 0.020/0.080 = 0.25 s.
- Terminal speed: vₜ = mg/k = 0.020(9.80)/0.080 = 2.45 m s⁻¹.
- Speed at t = 0.50 s: v(0.50) = vₜ(1-e^(-t/τ)) = 2.45(1-e^(-0.50/0.25)) = 2.12 m s⁻¹.
Answer + checks
- τ = 0.25 s, vₜ = 2.45 m s⁻¹, v(0.50 s) = 2.12 m s⁻¹
- v(0.50) < vₜ and increases toward vₜ as expected
Example: quadratic-drag terminal speed estimate
Problem
- Given: m = 80 kg, ρ = 1.2 kg m⁻³, C_D = 1.0, A = 0.70 m², g = 9.80 m s⁻²
- Find: terminal speed vₜ
Approach
- Use vₜ = square root of (2mg/(ρ C_D A)) from the quadratic-drag model.
Working vₜ = square root of (2(80)(9.80)/(1.2)(1.0)(0.70)) = square root of 1867 = 43.2 m s⁻¹.
Answer + checks
- vₜ ≈ 43 m s⁻¹ (≈ 155 km h⁻¹)
- Units: inside the square root is m²s⁻² so vₜ is m s⁻¹
Practice set (with hints + answers)
- Linear drag: show that v(t) = vₜ(1-e^(-t/τ)) implies a(t) = g e^(-t/τ).
- Linear drag: an object reaches 0.90 vₜ after time t. Express t in terms of τ.
- Linear drag: starting from rest, derive the displacement y(t) = y₀ + vₜ[t-τ(1-e^(-t/τ))] and show the small-t limit is y-y₀ ≈ (1/2)gt².
- Quadratic drag: verify that v(t) = vₜ tanh(gt/vₜ) satisfies m dv/dt = mg-bv².
- Quadratic drag: show that from rest y(t) = y₀ + vₜ²/g ln cosh(gt/vₜ).
- Sign matters: explain why the 1D quadratic drag law should be F_drag = -b v|v| rather than -bv² if the object can move upward (v < 0).
- Buoyancy: a falling object experiences buoyancy B upward (constant). Write the modified terminal speed for (i) linear drag, (ii) quadratic drag.
- Data/estimation (linear drag): suppose a best-fit curve gives vₜ = 3.00 m s⁻¹ and τ = 0.306 s. Use the linear-drag model to estimate g and comment.
Hints
- Use 0.90 = 1-e^(-t/τ).
- Differentiate tanh and use sech² = 1- tanh².
- Use ∫ tanh(ax) dx = (1/a) ln cosh(ax).
- Use “drag opposes motion”: if v < 0 then F_drag > 0 in the downward-positive convention.
- Replace mg by the net constant downward force mg-B (assume mg > B).
- Use vₜ/τ = g for linear drag.
Answers
- a = dv/dt = (vₜ/τ)e^(-t/τ) = ge^(-t/τ) since vₜ/τ = (mg/k)/(m/k) = g.
- e^(-t/τ) = 0.10 ⇒ t = τ ln 10 ≈ 2.30 τ.
- y(t) = y₀ + ∫₀^t vₜ(1-e^(-t'/τ)) dt' = y₀ + vₜ[t-τ(1-e^(-t/τ))]. Small t: e^(-t/τ) ≈ 1-t/τ + (1/2)(t/τ)² gives y-y₀ ≈ (1/2)(vₜ/τ)t² = (1/2)gt².
- dv/dt = g sech² (gt/vₜ), 1- tanh² = sech² so mg-bv² = mg(1-v²/vₜ²) = mg sech²(gt/vₜ) = mdv/dt.
- y(t) = y₀ + ∫₀^t vₜ tanh(gt'/vₜ) dt' = y₀ + vₜ²/g ln cosh(gt/vₜ).
- If v < 0 (object moving upward), drag must act downward, i.e. F_drag > 0 in a downward-positive convention. But -bv² is always negative, so it cannot represent drag for upward motion. Using -b v|v| fixes the sign automatically.
- Linear: vₜ = (mg-B)/k. Quadratic: vₜ = square root of ((mg-B)/b) = square root of ((2(mg-B))/(ρ C_D A)) .
- g = vₜ/τ = 3.00/0.306 = 9.80 m s⁻² (consistent with near-Earth gravity).
Quiz (MCQ)
Quiz — Resistive Forces (UY1)Notes and extensions (optional)
Buoyancy (constant upthrust)
If buoyancy B is not negligible, include it as an upward force. With downward-positive convention, the constant-force part becomes mg-B. For example, linear drag becomes: mdv/dt = (mg-B)-kv, so the terminal speed is vₜ = (mg-B)/k.
For quadratic drag with downward-only motion: mdv/dt = (mg-B)-bv², vₜ = square root of ((mg-B)/b) .
Summary + next steps
- Drag always opposes velocity; keep signs explicit to avoid errors.
- Linear drag gives exponential approach to vₜ = mg/k with time constant τ = m/k.
- Quadratic drag gives hyperbolic-tangent approach to vₜ = square root of (mg/b).
- Terminal speed is where net force (and acceleration) becomes zero.
Next: Uniform Circular Motion & Non-uniform Circular Motion Back To Mechanics (UY1)