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

Additional Mathematics Tutor Singapore | Optimisation, Maxima and Minima Word Problems

—

by

eduKate Asia
An open mathematics textbook and practice notebook sit beside stacked schoolbooks, pens and a calculator in a sunlit study space.

Additional Mathematics Tutor Singapore · Sec 3–4 SEC G2 K232 / G3 K341 · Maxima, minima and constrained optimisation

Did you know that the hardest part of an A-Math maximum-area question may happen before the first derivative? A student can differentiate 36x−2x² correctly, obtain x=9, and still lose marks by forgetting that x is a width, not the area, or by using a value that makes the other side of a rectangle negative. Optimisation is about choosing a usable answer from a mathematical model, not merely solving f′(x)=0.

When parents search for an Additional Mathematics tutor in Singapore because their child struggles with maxima and minima, I start by asking what the question actually requests. Is it a length, area, volume, cost, greatest height or a value of a parameter? What quantities are fixed, and which can change? A useful sketch and clear variables often do more than a page of rushed differentiation.

This guide teaches the common G2 and G3 Additional Mathematics optimisation foundations: completing the square for quadratic extrema, derivative-based stationary points, domain restrictions and interpretation. The 2027 SEC G2 K232 and G3 K341 syllabuses both specify maxima and minima applications. The worked problems are original learning examples, not predictions of future papers or guarantees of examination grades.

Explore maxima, minima and modelling problems

Quick answer: how do I solve A-Math optimisation questions?

Read the target and define a variable with units. Convert the fixed conditions into an equation, reduce the target to a function of one variable, and state the permitted domain. Use completing the square for suitable quadratics or differentiation where it gives a clear route. Solve for stationary candidates, classify them and check any relevant boundaries. Finally translate the selected input back into the quantity requested.

The final verification should include the original constraint and physical feasibility. A candidate maximum is not useful if it requires a negative side length, more material than available or a time outside the model’s stated interval. The algebra and the story must agree.

Quadratic extrema through completing the square

Consider A(x)=20x−x². Rewriting gives A(x)=−(x−10)²+100. Because a square is non-negative, the negative squared term is at most zero, so A has maximum 100 at x=10 on any domain containing ten.

This method is exact and requires no calculus. Both official 2027 G2 and G3 A-Math syllabuses include finding quadratic maxima/minima by completing the square. Students should understand the bound instead of memorising an isolated vertex formula.

If the same function were restricted to 0≤x≤5, the unrestricted candidate x=10 would not be available. The maximum over that interval occurs at x=5 and equals 75. The domain can change the answer even though the polynomial remains the same.

Differentiation gives stationary candidates

For A(x)=20x−x², the derivative is A′(x)=20−2x. Setting the derivative to zero gives x=10. Because A″(x)=−2, the stationary point is a local maximum; for this downward-opening quadratic it is also the unrestricted global maximum.

But f′(x)=0 alone is not a complete optimisation argument. It finds stationary candidates. The point may be a minimum, maximum or stationary inflection, and a restricted domain may have its greatest or smallest value at an endpoint instead.

A well-written solution explains why the candidate is of the required type and fits the question’s domain. If the target is an area, attach square units rather than reporting only a dimension.

Physical constraints must be written before differentiation

Suppose three sides of a rectangle use 36 metres of fencing, with a wall forming the fourth boundary. If the perpendicular widths are x, the remaining fenced side is 36−2x. Both must be positive: x>0 and 36−2x>0, so 0<x<18.

The area is A=x(36−2x)=36x−2x². The derivative A′=36−4x gives x=9, which is feasible; the remaining length is eighteen and the area is 162 square metres. The constraint check uses 9+9+18=36 metres of fencing.

If a mathematically obtained candidate were x=20, it would require a negative other side and could not describe the stated rectangle. A physical model adds restrictions beyond those of the polynomial alone.

Sometimes the requested quantity is not x

A question may ask for maximum area and both dimensions. If differentiation gives x=9, the answer is not finished: find the other side, 36−2(9)=18, and calculate 9×18=162 square metres.

Similarly, in a cost model, the optimising input may represent the number of items produced, while the question asks for the minimum cost. Substitute the optimal input into the original cost function and state the amount in the correct units. The input and output are two different mathematical quantities.

Local versus global extrema

A local maximum means nearby function values do not exceed it. A global maximum means no permitted value anywhere in the domain exceeds it. These are not interchangeable, especially for cubics with more than one stationary point or problems on restricted intervals.

For a polynomial on a closed interval, compare values at all interior stationary candidates and endpoints when the function is continuous on that interval. A higher local maximum might still be smaller than the value at an endpoint. The graph and domain determine the actual answer.

Why the second derivative is a test, not a promise

If f′(a)=0 and f″(a)<0, the usual second-derivative test identifies a local maximum. If f″(a)>0, a local minimum. If f″(a)=0, the test is inconclusive and a sign chart or other reasoning is needed.

Consider x³ and x⁴ at zero. Both have zero first and second derivatives at the origin, yet x³ has a stationary inflection and x⁴ a minimum. That pair helps students avoid the false rule that every zero second derivative proves an inflection.

Optimising a model is not forecasting the world

A simplified revenue function may assume all products sell at the stated price and demand follows a specified linear rule. The mathematical maximum is correct under that assumed model. It does not prove a real business should set that price without considering customers, supply, competition and uncertainty.

The same limitation applies to geometrical manufacturing examples. A sheet-of-card problem may ignore the thickness of the material and waste at folds. A tutor should teach the mathematics precisely while honestly distinguishing a modelled answer from an actual engineering design.

How a short sketch prevents a long wrong solution

For a fencing problem, draw the existing wall, show which three edges use purchased material and write their lengths. If the learner incorrectly includes the wall in the fencing total, the formula will describe a different problem even if all later calculus is correct.

For an area beneath a curve, mark the allowed x-interval and the relevant boundaries. For a storage-box problem, sketch how folding changes base dimensions and identify when the material would collapse. The diagram should make the constraints visible before the derivative is taken.

Forty original optimisation and extrema examples

Optimisation question 1: Two sides with a fixed sum

Question: Two positive lengths add to 20. Find their greatest possible product.

Worked answer: Let one be x, the other 20−x, so P=x(20−x)=−(x−10)²+100. Maximum product 100 occurs when both equal ten.

Common mistake: State 0<x<20; the equal split is feasible.

Check: 10×10=100.

Optimisation question 2: Rectangular area with perimeter 40

Question: A rectangle has perimeter 40 cm. Find maximal area.

Worked answer: If one side is x, the other is 20−x. Area A=x(20−x), maximum 100 cm² at x=10, a square.

Common mistake: The perimeter means sum of adjacent sides is 20, not 40.

Check: Four sides of length ten sum to 40.

Optimisation question 3: Rectangular area with perimeter 28

Question: A rectangle has perimeter 28 cm. Find maximal area.

Worked answer: Adjacent sides add to 14. Area x(14−x)=−(x−7)²+49, maximum 49 cm² at x=7.

Common mistake: Do not stop at optimising input x; state area.

Check: A 7-by-7 square has perimeter28.

Optimisation question 4: Three-sided fencing with 36 m

Question: A wall supplies one boundary and 36 m fence supplies three sides. Maximise enclosed area.

Worked answer: With widths x, length 36−2x, area 36x−2x². Derivative 36−4x=0 gives x=9, length18, maximum area162 m².

Common mistake: Only three sides use fencing.

Check: 9+9+18=36 and9×18=162.

Optimisation question 5: Three-sided fencing with 30 m

Question: Maximise rectangular area when three sides use 30 m of fencing.

Worked answer: Area x(30−2x) on0<x<15 has maximum at x=7.5, length15, area112.5 m².

Common mistake: The unfenced wall must not consume part of the 30 m allocation.

Check: 7.5+7.5+15=30.

Optimisation question 6: A riverbank with 100 m of fence

Question: Use 100 m of fencing for three sides of a rectangle along a river. Maximise area.

Worked answer: If the two widths are x, the fenced length is100−2x. Area A=100x−2x²; derivative 100−4x=0 gives x=25, length50, area1250 m².

Common mistake: Check 0<x<50 before accepting a critical point.

Check: 25+25+50=100.

Optimisation question 7: Two positive numbers summing to 30

Question: Find the maximum product of two positive numbers with sum30.

Worked answer: Product x(30−x)=−(x−15)²+225, so maximum225 when the numbers are both fifteen.

Common mistake: The completed-square maximum is the product, not the individual number fifteen.

Check: 15×15=225.

Optimisation question 8: Area under a sloping straight line

Question: A rectangle in the first quadrant has upper-right point on y=12−2x. Maximise its area.

Worked answer: Area A=x(12−2x) on0<x<6. Derivative 12−4x=0 gives x=3 and height6, maximum area18 square units.

Common mistake: The height is 12−2x, not twelve.

Check: At (3,6) the line equation holds.

Optimisation question 9: A profit quadratic

Question: A simplified profit model is P(x)=−2x²+20x−18. Find its maximum.

Worked answer: Derivative −4x+20=0 gives x=5. Substituting gives maximum P(5)=−50+100−18=32 model units.

Common mistake: The optimal input five is different from maximum output32.

Check: Second derivative −4 confirms a maximum.

Optimisation question 10: A cost minimum

Question: Minimise C(x)=x²−12x+50 over real x.

Worked answer: Complete the square: C=(x−6)²+14. Minimum cost is14 model units at x=6.

Common mistake: The fact that the minimum is positive is a feature of this illustrative model, not a real price forecast.

Check: The squared part is at least zero.

Optimisation question 11: Revenue from linear demand

Question: Demand q=100−2p, with p>0 and q>0. Maximise revenue R=pq.

Worked answer: R=100p−2p², so R′=100−4p. Set to zero: p=25, q=50 and revenue1250 units.

Common mistake: The chosen price must keep demand non-negative.

Check: 25×50=1250.

Optimisation question 12: Revenue from another demand curve

Question: Demand q=60−3p. Find revenue-maximising positive p.

Worked answer: R=p(60−3p). Derivative 60−6p=0 gives p=10, q=30, revenue300 units.

Common mistake: Do not confuse the price p with quantity q.

Check: At p=10 the demand relation gives30.

Optimisation question 13: Largest value of a quadratic

Question: Find maximum of f(x)=4x−x².

Worked answer: Complete square f=−(x−2)²+4. Maximum four occurs at x=2.

Common mistake: The maximum value is four, not the x-coordinate two.

Check: Derivative 4−2x is zero at two.

Optimisation question 14: Smallest quadratic value

Question: Find minimum of f(x)=x²−6x+8.

Worked answer: Complete square f=(x−3)²−1. Minimum −1 occurs at x=3.

Common mistake: A negative minimum is valid for an algebraic function.

Check: At x=3, original gives9−18+8=−1.

Optimisation question 15: A second quadratic minimum

Question: Find minimum of f(x)=x²−4x+10.

Worked answer: Complete square (x−2)²+6, so minimum six at x=2.

Common mistake: A completed-square constant shows the value after the squared term reaches zero.

Check: The derivative 2x−4 vanishes at two.

Optimisation question 16: A second quadratic maximum

Question: Find maximum of f(x)=−x²+8x−5.

Worked answer: Complete square f=−(x−4)²+11. Maximum eleven at x=4.

Common mistake: The maximum is an output, not the stationary x-value.

Check: Substituting four gives−16+32−5=11.

Optimisation question 17: Domain stops before stationary input

Question: For f=(x−4)²+1 on0≤x≤2, find its minimum.

Worked answer: The unrestricted minimum at x=4 is not allowed. The function decreases over the domain, so min f(2)=5.

Common mistake: Do not report the unattainable value one.

Check: Endpoint comparison gives f(0)=17, f(2)=5.

Optimisation question 18: Domain stops before a maximum

Question: For f=−(x−4)²+9 on0≤x≤2, find maximum.

Worked answer: The unrestricted peak nine at x=4 is outside the interval. The function increases on 0≤x≤2, so maximum f(2)=5.

Common mistake: Stationary points outside the domain are not feasible solutions.

Check: At x=0, f=−7, smaller than five.

Optimisation question 19: Constrained downward quadratic

Question: Maximise f=−x²+4x on0≤x≤1.

Worked answer: The unrestricted vertex is at x=2, beyond the domain. On the interval the function increases, so max f(1)=3.

Common mistake: Do not use global maximum four at x=2.

Check: Derivative −2x+4 remains positive on0≤x≤1.

Optimisation question 20: A rational cost-style function

Question: Minimise f=x+9/x for x>0.

Worked answer: Derivative f′=1−9/x²=0 gives x=3. Then f(3)=3+3=6, the minimum.

Common mistake: The domain x>0 matters; negative inputs would be a different problem.

Check: By AM–GM, x+9/x≥2√9=6.

Optimisation question 21: Another reciprocal minimum

Question: Minimise f=x+4/x for x>0.

Worked answer: Derivative 1−4/x²=0 gives x=2 within the domain; f(2)=4 is minimum.

Common mistake: Do not accept x=−2, which lies outside x>0.

Check: The function tends upward near zero and infinity.

Optimisation question 22: A cubic-plus-reciprocal model

Question: Minimise f=x²+16/x for x>0.

Worked answer: Derivative 2x−16/x²=0 leads to 2x³=16, so x=2. The minimum is f(2)=4+8=12.

Common mistake: The allowed positive domain eliminates nonpositive candidates.

Check: Second derivative 2+32/x³ is positive for x>0.

Optimisation question 23: Fixed-area rectangle, least perimeter

Question: A rectangle has area36 cm². Find least perimeter.

Worked answer: Sides x and36/x with x>0 give P=2x+72/x. Differentiate: P′=2−72/x²=0, so x=6, other side6, minimum perimeter24 cm.

Common mistake: A fixed area is not the same as a fixed perimeter.

Check: Four six-centimetre sides give24 cm.

Optimisation question 24: Fixed area100, minimum boundary

Question: A rectangle has area100 m². Find the minimum perimeter.

Worked answer: Sides x and100/x give P=2x+200/x. Setting P′=2−200/x²=0 gives x=10 and perimeter40 m.

Common mistake: Choose positive x because dimensions are lengths.

Check: A 10-by-10 square has area100 and perimeter40.

Optimisation question 25: An open box from a rectangular sheet

Question: A 30-by-20 cm sheet has equal square corners of side x cut out and folded into an open box. Maximise volume.

Worked answer: Volume V=x(30−2x)(20−2x) with0<x<10. Differentiate V=600x−100x²+4x³: V′=600−200x+12x². The feasible stationary point is x=(25−5√7)/3≈3.924 cm, producing volume about1056 cm³.

Common mistake: The larger algebraic stationary root is outside x<10 and cannot describe a box.

Check: Use the original dimensions to verify both remaining base sides are positive.

Optimisation question 26: A closed cylinder with fixed volume

Question: A closed cylinder has volume54π cm³. Find radius and height minimising total surface area.

Worked answer: Let radius r>0. Then h=54/r². Surface S=2πr²+2πrh=2πr²+108π/r. S′=4πr−108π/r²=0 gives r³=27, r=3 and h=6; minimum surface54π cm².

Common mistake: Differentiate total surface area, not the constant volume.

Check: At r3,h6, volume isπ·9·6=54π.

Optimisation question 27: A cubic with two stationary points

Question: Find local extrema of f=x³−6x²+9x+2.

Worked answer: Derivative 3(x−1)(x−3) gives x1 and3. Values f(1)=6 and f(3)=2. Since f″=6x−12, x1 is local maximum, x3 local minimum.

Common mistake: These are local extrema; on an unrestricted cubic there are no global extrema.

Check: Derivative changes sign at both roots.

Optimisation question 28: Endpoint comparison on the same cubic

Question: Find global extrema of f=x³−6x²+9x+2 on [0,4].

Worked answer: Check f(0)=2, f(1)=6, f(3)=2 and f(4)=6. Global maximum six occurs at x=1 and4; global minimum two at x=0 and3.

Common mistake: A local extremum need not be the only global winner on a restricted interval.

Check: Evaluating both endpoints reveals the ties.

Optimisation question 29: A cubic with close extrema

Question: Find local extrema of f=2x³−9x²+12x.

Worked answer: Derivative 6(x−1)(x−2) gives x1 and2. f(1)=5 and f(2)=4. Second derivative 12x−18 is negative at1, positive at2, so five is local maximum and four local minimum.

Common mistake: Do not assume the local minimum must be negative.

Check: Derivative signs change as expected.

Optimisation question 30: Another cubic with minimum then maximum

Question: Classify stationary points of f=−x³+3x²+1.

Worked answer: Derivative −3x(x−2) gives x=0 and2. f(0)=1 and f(2)=5. Second derivative −6x+6 is positive at0 and negative at2, so (0,1) is local minimum and (2,5) local maximum.

Common mistake: The negative cubic leading term changes end behaviour.

Check: The derivative sign changes −/+ at zero and +/− at two.

Optimisation question 31: A flat minimum with an inconclusive second derivative

Question: Classify x=0 for f=x⁴.

Worked answer: First derivative 4x³ changes from negative to positive at zero, so it is a minimum f(0)=0. But f″(0)=0 makes the simple second-derivative test inconclusive.

Common mistake: Second derivative zero does not imply inflection.

Check: x⁴≥0 throughout.

Optimisation question 32: A stationary inflection, not an extremum

Question: Classify x=0 for f=x³.

Worked answer: First derivative 3x² equals zero at0 but remains non-negative on both sides. The function keeps increasing, so zero is a stationary inflection.

Common mistake: f′=0 alone is not proof of a maximum or minimum.

Check: f(−1)<f(0)<f(1).

Optimisation question 33: A cubic prism capacity model

Question: Maximise V=x(12−x)² for 0<x<12.

Worked answer: Differentiate V′=(12−x)(12−3x) by product and chain rules. Stationary candidates x=4 and12, but12 is excluded. At x=4, volume V=4·8²=256 cubic units and is maximum inside the positive domain.

Common mistake: Check both factors and physical feasibility; x12 would collapse the base.

Check: Volume tends to zero at each boundary.

Optimisation question 34: A triangle with fixed hypotenuse

Question: A right triangle has legs x,y>0 and hypotenuse10. Maximise its area.

Worked answer: The condition is x²+y²=100, and area A=xy/2. For fixed x²+y², the product xy≤(x²+y²)/2=50. Maximum area25 occurs when x=y=5√2.

Common mistake: A right triangle’s area is half the product of the legs, not the full product.

Check: Leg squares 50+50=100 and area(50)/2=25.

Optimisation question 35: Closest point on a line

Question: Find the point on y=x closest to P=(2,0).

Worked answer: A general point is (x,x). Squared distance D²=(x−2)²+x²=2(x−1)²+2. It is minimised at x=1, giving closest point (1,1) and distance √2.

Common mistake: Minimising squared distance is valid because square root increases with non-negative input.

Check: Both coordinate differences are ±1 and the distance is √2.

Optimisation question 36: Highest rectangle under a parabola

Question: A rectangle in the first quadrant has upper-right corner on y=9−x². Maximise its area.

Worked answer: Area A=x(9−x²) on0<x<3. A′=9−3x²=0 gives x=√3 and height 9−3=6; maximum area is 6√3 square units.

Common mistake: The other stationary root −√3 is outside the first-quadrant domain.

Check: At the chosen x the top-right corner lies on the parabola.

Optimisation question 37: A constrained cost at an endpoint

Question: Minimise C=(x−5)²+2 for0≤x≤3.

Worked answer: The unrestricted minimum occurs at x=5 outside the domain. C decreases throughout0≤x≤3, so minimum C(3)=6.

Common mistake: Do not assume a model’s optimal input is feasible.

Check: C(0)=27, larger than six.

Optimisation question 38: Distinguish input and maximum output

Question: Find the maximising input and maximum value of f=−(x−3)²+7.

Worked answer: The function is largest when the squared part is zero at x=3. Maximum output is seven. State both values separately.

Common mistake: A student answering only three has given the location rather than the maximum value.

Check: Substitute x=3 to obtain seven.

Optimisation question 39: A rectangle with a fixed diagonal, optional extension

Question: A rectangle has diagonal10. Maximise its area.

Worked answer: Sides x,y>0 satisfy x²+y²=100. Area xy≤50, with equality at x=y=5√2. Thus maximum area is50 square units.

Common mistake: The rectangle’s area is xy, unlike a right triangle’s area xy/2.

Check: The diagonal of the resulting square has length √(50+50)=10.

Optimisation question 40: A positive interval quadratic minimum

Question: Minimise f=x²−2x+3 for x≥2.

Worked answer: Unrestricted vertex at x=1 is not in the domain. For x≥2, f′=2x−2>0, so function increases; minimum f(2)=3.

Common mistake: The selected domain changes which candidate can be used.

Check: Substitute two:4−4+3=3.

Optimisation question 41: A quadratic with a unique maximum

Question: Maximise A=12x−3x² for x>0.

Worked answer: Complete square A=−3(x−2)²+12. Maximum12 occurs at x=2, which is in the domain.

Common mistake: Do not confuse the coefficient −3 with a negative maximum.

Check: Derivative 12−6x vanishes at two.

Optimisation question 42: A product with endpoints

Question: Maximise P=x(8−x) for0≤x≤8.

Worked answer: Rewrite P=−(x−4)²+16. Maximum16 at x=4, while both endpoints give zero.

Common mistake: Include endpoint checks when using a closed permitted interval.

Check: The product is four times four at the maximum.

Optimisation question 43: Three measurements but only one variable

Question: A rectangle has length three times its width and perimeter32. Find area.

Worked answer: If width is x and length3x, 2(x+3x)=32 gives x=4 and length12. The area is48 square units. This is a constraint-solving task, not a free optimisation once the ratio is fixed.

Common mistake: Do not differentiate when the constraints already determine a unique rectangle.

Check: 2(4+12)=32.

Optimisation question 44: A stationary point that fails the domain

Question: Maximise f=−(x−2)²+4 on x≤0.

Worked answer: The unrestricted peak at x=2 is excluded. Over x≤0, f increases as x approaches zero, giving maximum f(0)=0.

Common mistake: The function’s mathematical peak may not belong to the feasible domain.

Check: At negative x the squared distance from two is larger.

Optimisation question 45: Read the model’s output units

Question: A radius r cm increases at0.5 cm/min. Find rate of circle area when r=3.

Worked answer: Area A=πr². dA/dt=2πr dr/dt=2π(3)(0.5)=3π cm²/min. This is a related-rates calculation rather than optimisation.

Common mistake: The rate is area per time, not an area alone.

Check: The positive rate fits an expanding radius.

How to tell which optimisation method is appropriate

A quadratic given in a form that can be completed to a square may reveal its extremum immediately. A product or rational expression with several changing factors may be clearer through differentiation. Neither method is universally superior; use the one that is correct and easiest to explain for the actual function.

If the problem contains a physical story, the decisive first step is often not calculus but writing the constraint. For three-sided fencing, this is 2x+y=L. For a box cut from a sheet, the permitted cut size must keep both base lengths positive.

A practical four-week optimisation revision sequence

Week 1 revisits quadratic graphs and completing the square. Week 2 uses geometric word problems with one-variable models and domains. Week 3 introduces derivative-based extrema and classification, including rational and cubic examples. Week 4 mixes constrained models, endpoint checks and final interpretation without topic-labelled hints.

After each lesson, ask for one changed question a few days later. A learner who can reconstruct the relationship from the words is developing transfer. A student who can only follow a tutor’s fully prepared diagram still needs a more independent attempt.

What a tutor should diagnose in marked work

One learner may choose the correct area formula but write a wrong perimeter constraint. Another may derive a sound model but differentiate incorrectly. Another may solve f′=0 accurately but forget to check the permitted interval. A final learner may obtain the right optimising x-value without reporting the requested cost, area or dimensions.

These errors need different instruction. A credible Additional Mathematics tutor should identify the first invalid line, explain the relevant mathematical relationship and test a changed question. Simply calling every mistake careless prevents useful feedback.

Frequently asked questions about A-Math optimisation

Is optimisation in the 2027 G2 and G3 Additional Mathematics syllabuses?

Yes. Both K232 G2 and K341 G3 include quadratic extrema and differentiation applications to maxima and minima. The overall calculus content differs, so choose the correct level.

Do I always need differentiation to find a maximum?

No. Completing the square gives many quadratic maxima or minima exactly and efficiently. Differentiation is useful for broader functions or when the question asks for rate-based reasoning.

Is every stationary point a maximum or minimum?

No. It may be a stationary inflection. First-derivative sign changes and a decisive second derivative can help classify the candidate.

Why should I check endpoints?

A problem with a restricted input interval may have its largest or smallest permitted value at an endpoint, even when an unrestricted stationary point exists elsewhere.

Can optimisation answers have surds?

Yes. Exact solutions of derivative or geometric equations may involve square or cube roots. Preserve the exact form when requested and apply numerical accuracy only as directed.

What if a stationary root makes a length negative?

Reject it if the model requires a positive length. The original physical domain controls which algebraic candidates are usable.

How can I check a maximum after differentiation?

Use the relevant second-derivative test, first-derivative signs, completed-square form or endpoint comparison. Also verify the dimensions satisfy the original constraint.

Does a maximum-area model guarantee a real engineering design?

No. A model’s result depends on its assumptions about shape, material and constraints. Real decisions may require additional measurements and factors.

Should a student memorise all forty worked examples?

No. Learn to name the target, build a one-variable model, state the domain and interpret a selected candidate. Changed independent examples are better evidence of learning.

Can optimisation be improved without private tuition?

Many students can progress with school teaching and deliberate independent correction. Tuition may help when the same modelling or interpretation error persists without targeted feedback.

The six-line optimisation template

Start with the target: area, volume, cost or profit. Write what the variable means and its unit. Translate the fixed resources into a constraint. Substitute until the objective depends on one variable. State the permitted domain. Then find and classify the relevant maximum or minimum and translate the result back into words.

This template is a decision process, not an extra formula to memorise. It stops a student from differentiating before establishing the correct function. When the question changes from four-sided fencing to three-sided fencing, the constraint line changes first. The later derivative is only as correct as that earlier model.

Choosing a variable that makes the algebra shorter

In a rectangular enclosure with two equal widths, choose x for the width rather than giving each side a separate unrelated variable. Then a fixed fencing total L immediately gives the remaining side L−2x. The one-variable area A=x(L−2x) is ready for completing the square or differentiation.

A learner might instead choose the long side y; that can still work, but the resulting substitution should be made clearly. The purpose is not to force one symbol on everyone. It is to reduce a multi-variable story into an objective that can be analysed correctly.

A quick diagram can expose an invalid optimisation

Before solving the open-box problem, sketch the original sheet and label the corner cut x. Both base dimensions shrink by 2x, not x, because a strip is removed from each side. Thus a 30-by-20 sheet produces base sides 30−2x and20−2x after folding.

The shorter original dimension limits x to less than ten. If a derivative produces a larger stationary candidate, the sketch explains why it must be rejected: the folded base would have non-positive length. A physical constraint is not optional just because a cubic equation yields a neat root.

Why local optima and boundary values must be compared

A cubic function on a closed interval can have several stationary points and endpoint values. The greatest among those candidates is the global maximum over the interval; a local maximum is only greater than nearby values. These may not be the same result.

A useful self-test is f(x)=x³−6x²+9x+2 on0≤x≤4. Critical values occur at x1 and3, but endpoints0 and4 must also be evaluated. The maximum six occurs at both1 and4; the minimum two at both0 and3. Checking just the stationary points would miss the endpoint ties.

When completing the square saves a page

For a quadratic A(x)=−2x²+36x, rewrite it as −2(x−9)²+162. The maximum is immediately 162 at x=9. Differentiation would produce exactly the same candidate, but the completed-square form also shows the global bound of the parabola.

The method choice should fit the question. A student who knows how to complete the square can often finish a quadratic extremum efficiently, while derivative methods become especially useful for rational functions, cubics and models involving changing rates.

Optimise a positive reciprocal function with a second check

For x>0, f(x)=x+9/x has derivative 1−9/x². Its only permitted stationary candidate is x=3, giving minimum six. But a student can also use the inequality between arithmetic and geometric means: x+9/x≥2√(x·9/x)=6, with equality at x=3.

The alternative proof helps verify the calculus result and shows why the minimum is global over x>0. The advanced inequality is optional enrichment if not part of the student’s assessment methods; the derivative and sign reasoning remain an accessible route.

Domain endpoints may be excluded but still informative

A problem may have 0<x<18 rather than a closed interval. The limiting cases x→0 and x→18 collapse a side of the fenced rectangle and have area approaching zero, but those collapsed shapes are not valid positive-area rectangles.

That behaviour still helps explain why an interior positive stationary maximum is plausible. The student should not report an excluded endpoint as an achieved minimum if the strict physical domain forbids it. Distinguish an infimum approached near a boundary from a minimum actually attained.

Turn an error log into a follow-up lesson

If a learner correctly sets A=x(36−2x) but writes A′=36−2x, the first invalid line is differentiation of −2x². A short power-rule correction is appropriate. If they write A=x(36−x), the derivative may be accurate, but the first error is the fencing model.

The tutor should repair the real cause, give a changed example and later test it without a visible model solution. That is more efficient than repeating the original question until its answer is remembered. A corrected strategy should survive changes in numbers, diagrams and question wording.

A self-check before submitting an optimisation answer

Ask whether the function represents the requested quantity. Check the domain before solving. After finding a candidate, confirm the extremum classification and compare endpoints where relevant. Finally recompute every dimension from the original constraint and state the unit of the answer.

When a result seems unreasonable, estimate. A rectangle with total fencing36 across three sides cannot have a width of50. A proposed area should also have square units, not metres. Such checks cannot replace mathematical proof, but they can warn the learner before a preventable wrong answer is finalised.

What progress should look like after several weeks

Initially a student might need the tutor to draw every rectangle or choose every variable. After instruction they should be able to sketch the constraints themselves, construct a one-variable objective, choose a valid extremum method and explain why the selected result belongs to the permitted domain.

A short unseen test with different dimensions is more informative than a polished original worksheet. Ask whether repeated mistakes have disappeared and whether the child can describe the method aloud without prompts. That is a meaningful measure alongside school assessment marks.

The relevance to JC and polytechnic Mathematics

Optimisation introduces the idea of making the best feasible decision under stated constraints. Later courses in mathematical sciences, economics, computing and engineering develop this through broader functions and sophisticated models. Secondary school examples are a starting point, not a guarantee of professional competence.

For course planning, consult the actual post-secondary institution’s entry requirements. A-Math can support quantitative readiness, but admission depends on the relevant qualification and rules. The educational value here is learning to model, calculate and verify rather than claiming that one topic determines a future career.

Official syllabuses and connected study

The formal requirements are in the 2027 SEC G2 Additional Mathematics K232 syllabus and 2027 SEC G3 Additional Mathematics K341 syllabus. For prerequisite calculus use tangents, normals and stationary points and product, quotient and chain-rule differentiation.

For mathematical translation, see the modelling and constraints guide and the quadratic-functions guide. The immutable Clementi Mathematics tutorial reference sets a standard of close feedback and independent explanation.

Choose the next lesson according to your current mathematical difficulty. The articles connect G3 logarithms, trigonometric graphs and shared calculus applications, but they do not replace the official syllabus or the student’s school teaching sequence.

The correct answer is the feasible one

I want students to finish an optimisation problem able to explain not just where a derivative is zero but why the chosen design or value satisfies the original constraint and is actually optimal in the permitted range. That is mathematical reasoning they can carry into later courses and everyday decisions.

For Secondary 3/4 A-Math tuition enquiries, contact eduKate Singapore or enquire by WhatsApp. Our teaching base is at 8 Fourth Avenue, Singapore 268674, near Sixth Avenue MRT. Confirm class suitability and current availability directly. Properly taught kids shine a bright light into the future.