[Intermediate] — Using derivatives to answer questions: how fast is this changing given that, and where is a quantity largest.
Before you start
- You can apply all the derivative rules, especially the chain rule — derivative rules.
- You know the Extreme Value Theorem — continuity.
What you will be able to do after this lesson:
- Solve a related-rates problem by differentiating a relationship with respect to time.
- Find critical points and classify them as maxima, minima or neither.
- Solve an optimisation problem end to end, including checking the endpoints.
- Explain what the second derivative tells you that the first does not.
1. Related rates
The setup: two quantities are related by an equation, both change over time, you know one rate and want the other.
The move: differentiate the relationship with respect to time, using the chain rule on every variable — because each is itself a function of .
Worked example
A ladder 5 m long leans against a wall. Its base slides away at 0.5 m/s. How fast is the top sliding down when the base is 3 m from the wall?
Relationship (Pythagoras): .
Differentiate with respect to — note the chain rule on both terms:
Substitute the known instant. When : , and :
Negative means the top is descending, as it must be.
NOTE
Substitute the specific values only after differentiating. Putting into the relationship first gives , a statement about one frozen instant with no in it — differentiating that yields .
The relationship must hold for all when you differentiate it. This is the single commonest error in related rates, and it is worth checking every time.
2. Critical points
At a maximum or minimum of a smooth function, the tangent is horizontal, so .
A critical point is where or does not exist. Every interior extremum is a critical point — but not every critical point is an extremum.
yet is neither a maximum nor a minimum. The curve flattens and continues rising — an inflection point.
So finding critical points is step one, not the answer.
3. Classifying them
First derivative test — check the sign of either side:
| before | after | Conclusion |
|---|---|---|
| local maximum | ||
| local minimum | ||
| same sign | same sign | neither |
Second derivative test — usually faster:
- → curve is concave up → minimum
- → concave down → maximum
- → inconclusive, fall back to the first derivative test
What the second derivative means: tells you the direction; tells you how the direction is changing. In motion, is velocity and is acceleration — and you can be moving forwards while decelerating, which is precisely with .
4. Optimisation
The procedure:
- Write the quantity to be optimised as a function of one variable.
- Use any constraint to eliminate the other variables.
- Find critical points.
- Check the endpoints of the valid domain too.
- Compare all candidates and state the answer with units.
Worked example
A farmer has 100 m of fencing and wants the largest rectangular enclosure against a straight river, so only three sides need fencing.
Let be the two perpendicular sides and the side parallel to the river.
Constraint: , so .
Objective: .
Differentiate and solve: .
Classify: , so it is a maximum.
Check the domain: must lie in , and . So wins.
Answer: m, area 1250 m².
Note the shape of the answer — the side parallel to the river is twice each perpendicular side. That relationship holds for any amount of fencing, which is a more useful result than the number.
NOTE
Step 4 is the one people skip. The Extreme Value Theorem guarantees a maximum exists on a closed interval, but it does not promise it is at a critical point — it can sit at an endpoint.
A function decreasing throughout has its maximum at , where . Checking endpoints is not optional.
5. Practice — problems
- A spherical balloon is inflated at 100 cm³/s. How fast is the radius growing when cm? (Volume .)
- A 10 m ladder slides down a wall; the base moves out at 1 m/s. How fast does the top fall when the base is 6 m out?
- Find and classify the critical points of .
- Find the maximum and minimum of on . Include the endpoints.
- A box with a square base and no lid must hold 32 000 cm³. Find the dimensions minimising the material used.
- Explain the error: a student solving problem 2 substitutes into before differentiating, gets , and then differentiates to obtain . What went wrong, in one sentence?
Answers — open only after an attempt
- , so , giving cm/s.
- . At , : , so m/s.
- , so critical points at and . : at , → local maximum ; at , → local minimum .
- at . Candidates: , , , . Maximum 18 at (an endpoint), minimum at both and . The maximum is at an endpoint, not a critical point — which is exactly why step 4 exists.
- Base , height , volume so . Material . Differentiating: cm, cm.
- Substituting first turns a relationship that holds for all time into a statement about one frozen instant, which contains no variables left to change — so differentiating it says nothing. Differentiate first, substitute after.
Before moving on
- Solve a related-rates problem, differentiating before substituting.
- Find critical points and classify them with both tests.
- Solve an optimisation problem including the endpoint check.
- Explain what the second derivative adds to the first.
Recap: Related rates differentiate a relationship with respect to time, applying the chain rule to every variable — and the specific values must be substituted only after differentiating, or the relationship collapses to a statement about one instant. Extrema occur at critical points where or is undefined, but not every critical point is an extremum; the first and second derivative tests classify them. Optimisation means expressing the target as a function of one variable, finding critical points, and comparing them against the endpoints, which the Extreme Value Theorem guarantees are candidates too.
This is the end of Calculus 1.
Next: Calculus 2 — the other half of the subject, where you reverse differentiation to compute accumulated totals.
Related
- Derivative rules · Continuity
- ai-ml — optimisation is what training a model is
Software engineering
Infrastructure & security
Computer science
Hardware & engineering