eduKate Singapore

VIEW THIS AS

Auto mode follows the Route Engine until you choose a viewpoint.

YOU ARE HERE

ROUTE CHECK

CONNECTED TO

WHAT NEXT

Use the canonical route for this room, or HELP if you are unsure.

Next useful routeCanonical hubHELP

What is Additional Mathematics | The Rectangle That Needs the Shortest Fence

—

by

eduKate Asia
Secondary Mathematics exercises showing the importance of clear written working.

A fixed-area rectangle teaches how constraints, derivatives, endpoints and alternate inequalities establish a global minimum perimeter.

Explore the perimeter optimisation

What is Additional Mathematics when a design must meet one fixed requirement while another quantity should be as small as possible? A rectangular plot offers a clear example.

Imagine a hypothetical flat rectangle with area exactly 144 square metres. We can choose its length and width, but the product of those dimensions must remain 144. If the rectangle needs a simple boundary on all four sides, which dimensions use the least total boundary length?

The answer is a square, 12 metres by 12 metres, whose perimeter is 48 metres. Yet that result is not something students should merely assert because squares look tidy. We can prove it using a constraint equation, differentiation, the derivative sign and an independent inequality.

This is an original mathematical teaching model. It does not account for land access, materials, setbacks, actual fencing regulations, gates or prices. It connects the modelling and maxima/minima applications of differentiation in Singapore’s 2027 G3 Additional Mathematics syllabus K341.

Start with one fixed condition

Let x represent the rectangle’s width in metres.

Let y represent its length in metres.

A rectangle’s area is:

A = xy.

The area is fixed at 144 m², so:

xy = 144.

We have two variables but only one constraint.

We can choose many positive pairs that satisfy it.

For example:

x = 8 and y = 18.

Or:

x = 9 and y = 16.

Or:

x = 12 and y = 12.

Every pair has product 144.

The design question is not which pairs have the correct area. It asks which permitted pair has the smallest perimeter.

That is an optimisation problem.

Build the perimeter formula

A rectangle has two sides of length x and two sides of length y.

Therefore:

P = 2x + 2y.

This is the total boundary length in metres.

However, P still contains two unknown quantities.

To differentiate with respect to x, we first express y in terms of x using the fixed-area constraint.

From:

xy = 144,

divide by x, where x > 0:

y = 144/x.

Substitute into the perimeter:

P(x) = 2x + 2(144/x).

Thus:

P(x) = 2x + 288/x.

Now the perimeter is a function of one variable.

That is the crucial modelling step before calculus begins.

State the meaningful domain

A rectangular width cannot be zero or negative in this model.

Therefore:

x > 0.

The corresponding length is:

y = 144/x,

which is also positive for x > 0.

So every positive x produces a mathematically permitted rectangle of area 144.

There is no finite upper bound on x in this simplified model.

If x becomes very large, the corresponding y becomes very small, but it remains positive.

We must not substitute x = 0 into P(x) because division by zero is undefined.

The domain is part of the model and will matter when we decide whether a stationary point gives a valid minimum.

A table suggests the answer

Calculate several possible rectangles.

Width xLength y = 144/xPerimeter P
4 m36 m80 m
6 m24 m60 m
8 m18 m52 m
9 m16 m50 m
12 m12 m48 m
16 m9 m50 m
18 m8 m52 m
24 m6 m60 m
Exact rectangular dimensions and perimeter calculations from a fixed-area mathematical model.

The square appears to have the smallest perimeter in our chosen table.

But the table checks only eight widths.

It does not prove that some value such as x = 11.7 or x = 12.3 cannot produce a slightly smaller perimeter.

To prove a minimum over all x > 0, we need reasoning that covers the entire permitted interval.

Differentiation provides one such method.

Differentiate the perimeter

Start with:

P(x) = 2x + 288/x.

Rewrite the reciprocal with a negative power:

P(x) = 2x + 288x⁻¹.

Differentiate term by term.

The derivative of 2x is two.

The derivative of 288x⁻¹ is:

288(−1)x⁻² = −288x⁻².

Therefore:

P′(x) = 2 − 288/x².

The derivative measures how the perimeter changes when we increase the width while adjusting the length to preserve area.

It is not the derivative of area. The area is fixed at 144 throughout the model.

Solve the stationary-point equation

A stationary point occurs where:

P′(x) = 0.

So:

2 − 288/x² = 0.

Rearrange:

2 = 288/x².

Multiply both sides by x², which is positive:

2x² = 288.

Divide by two:

x² = 144.

Algebra gives:

x = ±12.

But our domain requires x > 0.

Therefore only:

x = 12 metres

is permitted.

The negative solution is a root of the squared equation, not a valid rectangular width.

This is a simple example of why an optimisation solution must be filtered through the original domain.

Recover the other dimension

The area constraint gives:

y = 144/x.

At x = 12:

y = 144/12 = 12.

Therefore the corresponding rectangle is:

12 m by 12 m.

It is a square.

Its perimeter is:

P = 2(12) + 2(12).

So:

P = 48 metres.

We have found a stationary rectangle, but one more question remains.

Is that stationary point truly a minimum rather than a maximum or another type of stationary point?

We must classify it.

Use the first derivative sign

Our derivative is:

P′(x) = 2 − 288/x².

For 0 < x < 12:

x² < 144.

Therefore:

288/x² > 2.

So:

P′(x) < 0.

The perimeter decreases as width increases through values below twelve.

For x > 12:

x² > 144.

Therefore:

288/x² < 2.

So:

P′(x) > 0.

The perimeter increases as width grows beyond twelve.

Thus P decreases until x = 12 and increases afterwards.

x = 12 is a minimum.

This sign analysis proves the turning behaviour across the whole positive domain.

Interpret the derivative in ordinary language

For a very narrow rectangle, increasing the width allows the length to shrink considerably, reducing the total perimeter.

When the width reaches twelve, the reduction in the two long sides balances the increase in the two widening sides.

Beyond twelve, making the rectangle wider begins to cost more boundary length than it saves in the shrinking length.

The derivative records that trade-off mathematically.

P′ is negative when increasing width helps.

It is zero at the balanced choice.

It is positive when further width makes the perimeter larger.

This is why the stationary point has a meaningful design interpretation.

Confirm with the second derivative

Differentiate again:

P′(x) = 2 − 288x⁻².

The second derivative is:

P″(x) = −288(−2)x⁻³.

Therefore:

P″(x) = 576/x³.

For x > 0:

P″(x) > 0.

In particular:

P″(12) = 576/12³ = 576/1728 = 1/3 > 0.

A positive second derivative at the stationary point confirms a local minimum.

Because the first derivative has the decrease-then-increase sign pattern across the entire domain, this is also the global minimum for the model.

The two methods agree.

Check what happens at the edges of the domain

As x approaches zero from the positive side:

288/x becomes very large.

Therefore P(x) becomes very large.

As x becomes very large:

2x becomes very large.

Therefore P(x) again becomes very large.

These limiting observations are consistent with a global minimum occurring at some positive width in between.

We do not need to know advanced limit notation to understand the idea.

An extremely narrow rectangle uses much boundary because it must be very long to maintain the same area.

An extremely wide rectangle has the opposite difficulty.

The balanced square avoids both extremes.

An independent proof without calculus

There is another way to verify the minimum.

Start with:

P = 2x + 288/x,

where x > 0.

Half the perimeter is:

P/2 = x + 144/x.

The two positive numbers x and 144/x have product:

x(144/x) = 144.

By the arithmetic–geometric mean inequality:

(x + 144/x)/2 ≥ √144 = 12.

Therefore:

x + 144/x ≥ 24.

Multiply by two:

P ≥ 48.

Equality occurs when the two positive numbers are equal:

x = 144/x.

Then:

x² = 144,

so x = 12.

Thus the exact minimum is 48, achieved by the 12 m by 12 m square.

The inequality proof agrees with calculus.

Why the inequality method is useful

The derivative method explains how the perimeter changes when x changes.

The inequality method gives a direct lower bound that no admissible rectangle can beat.

Both methods establish the same result under the same positive-domain condition.

Students should not assume that a solution is more trustworthy merely because it uses a derivative.

A second mathematically valid method is valuable as an independent verification.

However, when a G3 Additional Mathematics question explicitly asks students to use differentiation, the derivative calculation and classification should still be shown.

The alternative proof supports understanding rather than replacing a required method.

A numerical check near the optimum

Choose x = 11.

Then:

y = 144/11.

The perimeter is:

P(11) = 22 + 288/11.

Since 288/11 ≈ 26.1818:

P(11) ≈ 48.1818 metres.

Now choose x = 13.

Then:

P(13) = 26 + 288/13.

Since 288/13 ≈ 22.1538:

P(13) ≈ 48.1538 metres.

Both exceed 48.

The differences are small, which explains why a low-resolution numerical table might fail to identify the exact optimum.

The symbolic proof does not depend on the spacing of sampled widths.

The graph of perimeter has one valley

For x > 0:

P(x) = 2x + 288/x.

The curve decreases until x = 12.

It has a horizontal tangent at the point:

(12, 48).

Then it rises.

Its second derivative is positive throughout the positive domain, so the graph bends upward everywhere in that interval.

This is a typical optimisation shape: one smooth valley and one global minimum.

A good graph sketch should show the proper domain and should not join the curve through x = 0, where the function is undefined.

The drawing should confirm the algebra rather than substitute for the analysis.

The meaning of the gradient at x = 8

Our derivative is:

P′(x) = 2 − 288/x².

At x = 8:

P′(8) = 2 − 288/64.

Since 288/64 = 4.5:

P′(8) = −2.5.

The negative derivative means that a small increase in width near eight metres reduces the perimeter.

For an increase Δx near x = 8:

ΔP ≈ −2.5Δx.

For example, a very small width increase of 0.1 m predicts a perimeter reduction of about 0.25 m.

That is a local approximation, not an exact change over a large interval.

The derivative answers a sensitivity question as well as the optimisation question.

The gradient at x = 16

At x = 16:

P′(16) = 2 − 288/256.

Calculate:

288/256 = 9/8 = 1.125.

Therefore:

P′(16) = 2 − 1.125 = 0.875.

The positive derivative means increasing width slightly from sixteen metres increases the perimeter.

Compare this with the negative gradient at eight metres.

The sign change between them is consistent with the minimum at twelve.

A student who finds P′(12) = 0 but never examines the sign or second derivative has not fully established the optimisation claim.

Generalise to any fixed area

Let the required area be a positive number A.

Then:

xy = A.

So:

y = A/x,

for x > 0.

The perimeter is:

P(x) = 2x + 2A/x.

Differentiate:

P′(x) = 2 − 2A/x².

Set the derivative equal to zero:

2 − 2A/x² = 0.

Thus:

x² = A.

Because x > 0:

x = √A.

Then:

y = A/√A = √A.

Therefore the minimum-perimeter rectangle with fixed positive area A is a square.

The minimum perimeter is:

P_min = 4√A.

Our original example used A = 144, giving 4 × 12 = 48.

Why a scale change affects length and area differently

Suppose the fixed area is multiplied by four.

The original area 144 becomes:

576 square metres.

The minimum-perimeter square then has side:

√576 = 24 metres.

Its minimum perimeter is:

4(24) = 96 metres.

The area increased by a factor of four.

The minimum perimeter increased by a factor of two.

That is because area scales with the square of a length scale, while perimeter scales linearly.

This connects optimisation with similarity and dimension analysis.

A student should not assume that quadrupling an area automatically quadruples the required boundary length.

A fixed-area rectangle need not be physically practical

The mathematical minimum occurs at a square.

But a real piece of land may have constraints such as terrain, boundary setbacks, entrances, drainage, neighbouring buildings or a requirement that one side be long enough for access.

Any such constraint may change which designs are allowed.

Our model ignores all those details.

A mathematically minimal perimeter is not automatically the cheapest or safest real construction.

The value of mathematical modelling lies partly in knowing what was held constant and what was omitted.

The answer applies exactly to the ideal rectangle and simple boundary assumption defined in the question.

Add a width restriction

Suppose a hypothetical site permits widths only between eight and ten metres:

8 ≤ x ≤ 10.

The area must still be 144 m².

The unrestricted optimum x = 12 is now outside the allowed interval.

So it is not a valid answer to the revised problem.

The derivative satisfies:

P′(x) < 0

for every x in [8, 10], because the entire interval lies below twelve.

Therefore P decreases throughout the permitted interval.

The smallest perimeter occurs at the upper endpoint:

x = 10.

Then:

y = 144/10 = 14.4 metres.

The perimeter is:

P = 2(10) + 2(14.4) = 48.8 metres.

The domain restriction has changed the optimal design.

Why the endpoint must be checked

In the unrestricted problem, the minimum occurs at a stationary point.

In the restricted interval [8, 10], no stationary point lies within the allowed range.

Therefore the minimum occurs at a boundary.

This is a general optimisation lesson: the best admissible value may occur at an endpoint, even if the mathematical function has a better unrestricted stationary value elsewhere.

A student should not automatically report the same stationary x from an earlier problem after the domain changes.

State the permitted interval, find critical values inside it and compare endpoints when necessary.

The word “allowed” is as important as the derivative calculation.

Check the restricted result numerically

At x = 8:

P(8) = 16 + 288/8 = 16 + 36 = 52 metres.

At x = 9:

P(9) = 18 + 288/9 = 18 + 32 = 50 metres.

At x = 10:

P(10) = 20 + 28.8 = 48.8 metres.

The perimeter decreases as width increases through the permitted interval.

That supports the derivative sign result.

But the exact reason the endpoint x = 10 is optimal is that P′ is negative throughout [8, 10].

The table illustrates that conclusion; it is not a replacement for it.

A variation: only three sides need a boundary

Now imagine the rectangle borders an ideal straight wall, and only the other three sides require a separate boundary.

Let x be the width perpendicular to the wall and y the length along the wall.

Keep the area fixed:

xy = 144.

The three-sided boundary length is:

F = 2x + y.

Use y = 144/x:

F(x) = 2x + 144/x.

Differentiate:

F′(x) = 2 − 144/x².

Set F′(x) = 0:

2x² = 144.

Thus:

x² = 72.

For positive x:

x = 6√2 metres.

Then:

y = 144/(6√2) = 12√2 metres.

The optimal length is twice the optimal width in this ideal three-sided model.

Check the three-sided minimum

For:

F(x) = 2x + 144/x,

the second derivative is:

F″(x) = 288/x³.

It is positive for x > 0.

Therefore the stationary point is a minimum.

Substitute x = 6√2 and y = 12√2:

F = 2(6√2) + 12√2.

So:

F_min = 24√2 metres.

This is approximately 33.94 metres.

The four-sided optimum was a square, but the three-sided optimum is not.

Why?

The cost function changed. One long side no longer contributes to the boundary requirement.

The correct result depends on what the problem actually counts.

Do not confuse the three-sided and four-sided models

For a fully enclosed rectangle:

P = 2x + 2y.

For a rectangle along a wall, with the wall side not counted:

F = 2x + y.

They are different objective functions.

Both may share the same area constraint xy = 144.

But the different objectives produce different optimal shapes.

This is one reason translating a word problem into algebra is often harder than differentiating the resulting formula.

A derivative cannot repair an incorrectly modelled perimeter.

The student should draw and label the required boundary segments before writing the cost function.

Why unit checking matters

The area constraint is:

xy = 144 m².

Both x and y have units of metres.

The perimeter is a length, so its final unit is metres.

The derivative P′ = dP/dx measures metres of perimeter per metre of width, making its numerical value dimensionless when both lengths use metres.

A student who reports 48 m² for perimeter has mixed area and boundary length.

Likewise, reporting 144 metres as the area requirement omits the squared unit.

Units are a useful final check on whether the formula represents the question correctly.

Common mistake: differentiating before eliminating y

The perimeter P = 2x + 2y contains two changing dimensions.

If the constraint xy = 144 has not yet been used, we do not know how y changes when x changes.

The safe beginner’s method is to write:

y = 144/x,

then substitute into the perimeter.

Only after expressing P as a function of x do we differentiate.

This makes the fixed-area condition explicit throughout the calculation.

Common mistake: forgetting the second pair of sides

A four-sided rectangle has two widths and two lengths.

So:

P = 2x + 2y.

Using x + y would calculate only half the perimeter.

That constant factor would not change the optimal x in this simple case, but it would give the wrong minimum perimeter and any associated quantities.

The model should be correct even if a particular mistake happens not to move the stationary point.

Common mistake: differentiating 288/x incorrectly

Write:

288/x = 288x⁻¹.

The derivative is:

−288x⁻² = −288/x².

It is not +288/x², and it is not 288.

The negative power explains the sign.

This is a direct application of differentiation of rational powers.

Common mistake: keeping the negative square root

From x² = 144:

x = ±12 algebraically.

But a rectangle width must be positive.

Only x = 12 belongs to the domain.

Filtering the root is part of answering the optimisation problem.

Common mistake: calling a stationary point a minimum without proof

A zero derivative identifies a stationary candidate.

It does not automatically classify the point.

Use a first derivative sign change or an appropriate second derivative test.

Here P″(12) > 0, and the derivative changes from negative to positive.

Both support a minimum.

Common mistake: ignoring a new width restriction

The unrestricted optimum is x = 12.

But if a revised problem permits only 8 ≤ x ≤ 10, that answer is invalid.

The permitted minimum occurs at x = 10.

The domain can change the answer even when the algebraic formula is unchanged.

A six-minute independent practice set

1. Express the length y of a 144 m² rectangle in terms of its positive width x. 2. Write its four-sided perimeter P(x). 3. Differentiate P. 4. Find every algebraic solution of P′(x) = 0. 5. Select the permitted stationary width. 6. Find the corresponding length and minimum perimeter. 7. Use the second derivative to classify the stationary point. 8. Explain why a small numerical table is not a global proof. 9. Find the minimum if the width is restricted to [8, 10]. 10. Form the three-sided boundary function for a rectangle against a wall. 11. Determine its optimal width and length. 12. State the general minimum four-sided perimeter for area A > 0.

Answers: y = 144/x; P = 2x + 288/x; P′ = 2 − 288/x²; ±12; x = 12; y = 12 and P = 48 m; P″(12) = 1/3 > 0; a table tests finitely many widths; x = 10 with P = 48.8 m; F = 2x + 144/x; x = 6√2 and y = 12√2; and P_min = 4√A.

The key skill is building the correct one-variable model, not merely remembering that a square is optimal in one particular situation.

What parents can ask without doing calculus

Ask your child why a rectangle with the same area can require a different amount of boundary material.

Ask why increasing the width forces the length to shrink.

Ask why a square minimises the four-sided perimeter in the unrestricted model.

Then ask why the best dimensions change when one side is supplied by a wall or the allowed width is restricted.

These questions reveal whether the learner understands constraints, optimisation and the role of assumptions.

The derivative calculation is important, but the choice of what to minimise and which designs are allowed comes first.

How a focused small-group tutorial can help

One learner may correctly differentiate but construct the wrong perimeter formula.

Another may model the perimeter correctly but forget to eliminate one variable using the area constraint.

Another may find x = ±12 and keep the negative root.

A fourth may report a stationary point without proving it is a minimum.

Another may handle the unrestricted problem but miss an interval endpoint when the permitted domain changes.

A careful tutor should identify the earliest incorrect step and test understanding using a changed fixed area or boundary configuration.

The aim is independent mathematical modelling and a justified final answer.

Where this belongs in Singapore G3 Additional Mathematics

The official 2027 Singapore–Cambridge Secondary Education Certificate G3 Additional Mathematics K341 syllabus includes derivative rules, stationary points, increasing/decreasing functions, second derivative tests and differentiation applications to maxima and minima under Calculus C1.

The rectangle problem uses those prescribed concepts in a model with a clear objective and constraint.

The three-sided variation and inequality proof are supplementary ways to understand the result, not assertions about real fencing practice or particular examination questions.

The 2027 G3 Additional Mathematics code is K341, with GCE O-Level Additional Mathematics 4049 given by SEAB as the reference for 2026 and earlier. The G2 K232 syllabus is separate.

Frequently asked questions

Why is the optimal fixed-area rectangle a square?

For a full four-sided perimeter, the derivative of 2x + 2A/x vanishes at x = √A, making both dimensions equal. The first and second derivative tests establish a minimum.

Can a wide rectangle have the same area as a narrow rectangle?

Yes. Their widths and lengths may differ, provided the product remains the required area.

What is the minimum perimeter for area 144 m²?

It is 48 metres, achieved by the 12 m × 12 m square.

What does the derivative tell us?

It tells us how the required perimeter changes when the width changes under the fixed-area constraint.

Is P′(x) = 0 enough to prove a minimum?

Not by itself. Classify the stationary point using derivative signs, a valid second derivative test or another general proof.

Why reject x = −12?

A negative width is outside the physical and mathematical domain of positive rectangle dimensions.

Does the square remain optimal if only three sides need a boundary?

No. The objective changes to 2x + A/x, producing a different optimal length-to-width ratio.

Can an endpoint be the best design?

Yes. When the permitted interval excludes the unrestricted stationary point, an endpoint can provide the constrained minimum.

Is optimisation included in G3 Additional Mathematics?

Yes. Applying differentiation to maxima and minima is explicitly prescribed under Calculus C1 of the 2027 K341 syllabus.

For another constrained model, read the three-sided garden fence that explains a quadratic maximum.

For the difference between a turning point and a stationary inflexion, see the parameter dial that changes a cubic’s stationary points.

For a derivative of a changing ratio, read the changing fraction and quotient rule.

For general explanations of Additional Mathematics, visit What Is Additional Mathematics? and How Additional Mathematics Works.

The immutable Clementi Mathematics small-group teaching reference shows why algebraic foundations, clear assumptions and close correction matter when students tackle multi-step optimisation problems.

Official source and mathematical limits

The curriculum reference is SEAB’s 2027 G3 Additional Mathematics K341 syllabus (PDF), Calculus C1, together with the official 2027 G3 subject listing.

The rectangle is an ideal geometric model with positive side lengths and fixed area. The boundary lengths are mathematical quantities, not estimates of real fencing costs, material requirements or regulatory compliance.

The enduring lesson is that optimisation begins with a correct objective and a valid constraint. Differentiation then finds and tests the best admissible choice.