Lesson 8 of 8 · 22 min
Building optimisation models
Turn a worded problem into a function of one variable using a constraint, state its feasible domain, then optimise it and justify the answer.
What you will be able to do
- Draw a diagram, name the variables, and write the quantity to be optimised.
- Use a constraint to write that quantity in terms of a single variable, and state the feasible domain.
- Optimise the model, justify the nature of the optimum, and answer in context.
Often nobody hands you the function. You are told about a box, a fence, a can or a pipeline, and you must build the model yourself. This is the part of optimisation that earns most of the marks in complex unfamiliar questions, and the part that needs the most care.
Building an optimisation model
Draw and label a diagram
Name every length or quantity that can change.
Write the quantity to be optimised
For example, length × width × height. It may have two variables at first.
Use the constraint
A fixed perimeter, area, volume or amount of material links the variables. Use it to eliminate all but one.
State the feasible domain
Every length must be positive. This often gives an interval like .
Optimise and justify
Differentiate, solve, reject values outside the domain, and use the second derivative test.
Answer in context
Give the optimal dimensions and the optimal value, with units.
Cut a square of side cm from each corner of a 24 cm square sheet, then fold up the sides to make an open box.
Worked example
The largest open box from a square sheet
An open box is made from a 24 cm by 24 cm square of cardboard by cutting a square of side cm from each corner and folding up the sides.
(a) Show that the volume of the box is cm³.
(b) Determine the maximum volume, and the value of that gives it.
Step 1: (a) Write the volume
The base is a square of side and the height is , so
Step 2: State the domain
and , so .
Step 3: (b) Differentiate and solve
when or . Reject : it is outside the domain (the base would have zero width).
Step 4: Justify the maximum
, so at it is : a maximum.
Step 5: Find the maximum volume
cm³
Try it yourself first, then check the working one step at a time.
Answer
The maximum volume is 1024 cm³, when squares of side 4 cm are cut from the corners. The box is then 16 cm by 16 cm by 4 cm.
Method
Volume = base area × height, in terms of .
Domain: every dimension positive.
Solve , reject values outside the domain, justify with .
Answer with the volume and the dimensions.
Watch out for
Using for the base: a square comes off both ends of each edge.
Keeping , which gives a box with no base.
In the box problem above, what is the feasible domain of ?
Constraints do the work. In the box problem the sheet size fixed everything. More often there are two variables and a separate condition, such as a fixed length of fencing or a fixed volume, that links them.
A farmer has 120 m of fencing to make a rectangular paddock beside a straight river. The river forms one side, so no fence is needed there. The two sides perpendicular to the river are each m long. Determine the maximum area of the paddock.
Trigonometric models appear when an angle can vary. The variable is then the angle, in radians, and the domain is an interval of angles.
Worked example
The best angle for a rain gutter
A rain gutter is made from a strip of metal 30 cm wide by bending up 10 cm on each side. The base is 10 cm wide and each side makes an angle with the horizontal, where , so the cross-section is a trapezium with positive top width.
(a) Show that the area of the cross-section is cm².
(b) Determine the angle that lets the gutter carry the most water, and the maximum cross-sectional area.
Step 1: (a) Find the height and the top width
Each slanted side is 10 cm at angle , so the height is and each side reaches beyond the base horizontally. The top width is .
Step 2: Use the trapezium area formula
Step 3: (b) Differentiate
Expand first: . Using the product rule on the second term,
Step 4: Write in terms of cos θ and factorise
With ,
Writing everything in terms of turns the equation into a quadratic in .
Step 5: Solve dA/dθ = 0 in the domain
gives . gives , outside the domain (the gutter would be flat and folded back).
Step 6: Justify the maximum
At : , so this is a maximum.
Step 7: Find the maximum area
cm²
Try it yourself first, then check the working one step at a time.
Answer
The gutter carries the most water when (60°), giving a maximum cross-sectional area of cm².
Method
Use trigonometry to write the lengths in terms of the angle.
Differentiate, then use to get a quadratic in .
Keep solutions in the domain and justify with the second derivative.
Watch out for
Working in degrees. Calculus with trigonometric functions needs radians.
Assuming the best angle is (vertical sides). Sloping the sides outwards widens the top, which helps until .
Some problems need technology to finish. Minimising the metal in a can gives an equation like ; the method is the same, but you finish with a calculator in Paper 2.
Worked example
The can that uses the least metal
A closed cylindrical can must hold 500 cm³. Determine the radius and height that use the least metal, and the minimum surface area, to two decimal places. What do you notice about the shape of the can?
Step 1: Write the quantity and the constraint
Surface area: (two ends and the curved side).
Volume constraint: , so .
Step 2: Write S in terms of r
Step 3: Differentiate and solve
Step 4: Justify the minimum
for all , so this is a minimum.
Step 5: Find the height and the area
cm and cm².
Step 6: Notice the shape
From , , so . The best can is as tall as it is wide.
Try it yourself first, then check the working one step at a time.
Answer
The least metal is used when cm and cm, giving a minimum surface area of about 348.73 cm². The height equals the diameter.
Method
Write the quantity to minimise and the constraint.
Use the constraint to eliminate .
Solve , justify with , and evaluate.
Watch out for
Forgetting one of the two ends in a closed can (or including a lid on an open one).
Rounding early: use the stored value to find and .
Common mistake
Keeping two variables
You cannot differentiate with respect to while is also changing. Use the constraint to replace first, so the model has only one variable.
Exam tip
Show that, then use it
Many exam questions say "Show that the area is " and then ask you to optimise it. Show every step of the set-up. If you cannot, still use the given model for the later parts: those marks are separate.
Summary
Diagram, variables, quantity to optimise, constraint, one-variable model, domain.
The feasible domain comes from every length (or angle) having to make sense.
Reject solutions of that are outside the domain, with a reason.
Justify with the second derivative test, and check endpoints when the domain is closed.
Give the optimal dimensions and the optimal value, with units, in a sentence.
Finished the lesson?
Try the 2 quick checks above and the lesson completes itself, or mark it complete now.