Modelling with functions

Modelling with functions

Spec 2.11 · Edexcel A-Level Maths: Revision notes

Studying this course? Keep your progress in one place.

Key facts

  • A mathematical model uses a function to describe a real situation; match the function to the behaviour.
  • Exponential growth N=AektN=Ae^{kt} has rate proportional to size; decay T=a+be−ktT=a+be^{-kt} falls towards a limit aa.
  • Trigonometric models d=a+bsin⁡(ct)d=a+b\sin(ct) repeat: mean aa, amplitude bb, period 360∘c\dfrac{360^\circ}{c}.
  • Isolate the exponential before taking natural logs.
  • Always evaluate a model: name an assumption, say what goes wrong and suggest a refinement.

Choosing a model

Match the behaviour (growth, decay, cycles or inverse proportion) to the function, and link each constant to its meaning.

A mathematical model uses a function to describe a real situation. The amplitude of a+bsin⁡(ct)a+b\sin(ct) is bb, and its period is 360∘c\dfrac{360^\circ}{c} (or 2πc\dfrac{2\pi}{c}). Inverse proportion P=kVP=\dfrac kV models pressure and volume at constant temperature.

Exponential growth

  • N=AektN=Ae^{kt}, k>0k>0
  • Rate proportional to size

Exponential decay

  • T=a+be−ktT=a+be^{-kt}
  • Falls towards a limit aa

Trigonometric

  • d=a+bsin⁡(ct)d=a+b\sin(ct)
  • Repeating cycles

Reciprocal

  • P=kVP=\dfrac kV
  • Inverse proportion
1234562468101214161820xyy = a e^(kx)
Exponential growth: a is the starting value, k the growth rate

Which function best models the temperature of a cup of tea cooling towards room temperature?

Exponential models

Isolate the exponential, then take natural logs to solve for time.

For N=500e0.4tN=500e^{0.4t}, dNdt=200e0.4t=0.4N\dfrac{dN}{dt}=200e^{0.4t}=0.4N, so the rate of change is proportional to the size. At t=5t=5 it is 200e2≈1480200e^2\approx1480 bacteria per hour.

For cooling T=20+70e−0.05tT=20+70e^{-0.05t} the initial value is T(0)=90T(0)=90 and the limit is 2020.

10203040506020406080100xyT = 20(0, 90)(16.9, 50)T = 20 + 70e^(−0.05t)
Cooling curve T=20+70e−0.05tT=20+70e^{-0.05t} falls towards 2020

Worked example: solving for time

When does the temperature reach 5050 for T=20+70e−0.05tT=20+70e^{-0.05t}?

Worked example: target size

When does N=500e0.4tN=500e^{0.4t} reach 1000010000?

For N=500e0.4tN=500e^{0.4t}, what is dNdt\frac{dN}{dt} in terms of NN?

Trigonometric and reciprocal models

Use the smallest positive angle for the first time, then add a full cycle or use symmetry for later times.

For tides d=6+2.5sin⁡(30t)∘d=6+2.5\sin(30t)^\circ the maximum is 8.58.5 m, the minimum 3.53.5 m and the period 36030=12\dfrac{360}{30}=12 hours.

For P=kVP=\dfrac kV, find k=PVk=PV from one pair of values, then substitute.

24681012141618202224246810xymean 6(3, 8.5)(9, 3.5)(7, 4.75)d = 6 + 2.5 sin(30t)°
Tide height over 24 hours: mean 6 m, amplitude 2.5 m, period 12 hours

Worked example: when is the tide 4.75 m?

Solve d=4.75d=4.75 for d=6+2.5sin⁡(30t)∘d=6+2.5\sin(30t)^\circ.

Worked example: reciprocal model

P=120P=120 kPa when V=250V=250. Find PP when V=80V=80.

In d=6+2.5sin⁡(30t)∘d=6+2.5\sin(30t)^\circ, what is the period?

Limitations and refinements

Name the assumption, say what happens to the prediction, and state a specific improvement.

Exponential growth predicts unlimited growth: 500e0.4t500e^{0.4t} gives about 8×1078\times10^7 bacteria at 3030 hours. A limitation is a condition where the model gives unrealistic results, and a refinement is a change that fits reality better.

To compare with data, find a percentage difference: 15001500 against a measured 12001200 is 25%25\% too high.

24681012141618202224500010000150002000025000xylimit 20000(0, 500)logisticexponential
Exponential growth (dashed) is unbounded; the logistic model levels off at 20000

Exponential growth

Assumption:
Unlimited growth
Problem:
Impossible with finite food
Refinement:
Logistic model levels off

Cooling

Assumption:
Surroundings stay at 20 °C
Problem:
Room temperature may change
Refinement:
Let the limit vary

P = k/V

Assumption:
Constant temperature, ideal gas
Problem:
Over-predicts at high pressure
Refinement:
Use P(V − b) = k

A model predicts 15001500 but the measured value is 12001200. What is the percentage error relative to the measured value?

Try an exam question

The temperature of a drink, TT degrees Celsius, tt minutes after it is poured is modelled by T=20+70e−0.05tT=20+70e^{-0.05t}. Find the initial temperature, and find the time taken to cool to 50 ∘C50\,^\circ\text{C}, giving your answer to 1 decimal place.

[4 marks]

That's the notes covered.

Carry on to the next subtopic.

Try a question

What does the model predict for the initial temperature of the tea?

1 mark, marked against the mark scheme

Revision notes