Unit 4 Differentiation Tools for Approximation and Limits (AP Calculus BC)

Local Linearity and Approximation

What “local linearity” means

A core idea in calculus is that smooth curves look like straight lines when you zoom in far enough. If a function $f(x)$ is differentiable at a point $x=a$, then near $a$ the graph of $f$ is well-approximated by its tangent line at $a$. This is called local linearity.

You can think of it like this: a curved road is definitely not a straight line globally, but if you stand on a small enough segment of it, it looks nearly straight. Differentiability is the condition that guarantees this “straight-looking” behavior at a point.

This matters because straight lines are easy to compute with. If you can replace a complicated function with a linear function near a point, you can approximate values, estimate changes, and solve applied problems much more simply.

From tangent line to linearization

The tangent line to $y=f(x)$ at $x=a$ has slope $f'(a)$ and passes through $\left(a,f(a)\right)$. Its equation is

$L(x) = f(a) + f'(a)(x-a)$

This function $L(x)$ is called the **linearization** of $f(x)$ at $x=a$.

• $a$ is the center point where you know the function value and derivative.
• $f(a)$ anchors the line vertically.
• $f'(a)$ gives the slope.
• $x-a$ measures how far you’ve moved from the center.

The key approximation statement is:

$f(x) \approx L(x) \text{ for } x \text{ near } a$

The word “near” is crucial. Linearization is not meant to be accurate far away from $a$. The further you move from $a$, the more the curve’s bending (its concavity) will cause the linear model to drift from the true value.

Why linearization works (the derivative connection)

The derivative at $a$ is defined by a limit that compares the function to a line:

$f'(a)=\lim_{x \to a} \frac{f(x)-f(a)}{x-a}$

This says that as $x$ approaches $a$, the ratio of “vertical change” to “horizontal change” approaches a constant slope. Rearranging that idea informally gives:

$f(x)-f(a) \approx f'(a)(x-a)$

and then

$f(x) \approx f(a) + f'(a)(x-a)$

So linearization is basically the derivative definition turned into an approximation.

Differentials: approximating changes

Another common way to express local linearity is with differentials, which are especially useful in applied problems about small changes.

Let

$\Delta x = x-a$

and

$\Delta y = f(x)-f(a)$

Near $x=a$, the linear approximation predicts

$\Delta y \approx f'(a)\Delta x$

In differential notation, you write

$dy = f'(x)dx$

At the point $x=a$, a small input change $dx$ produces an approximate output change

$dy \approx f'(a)dx$

Here’s the conceptual difference:

• $\Delta y$ is the actual change in the function.
• $dy$ is the predicted change from the tangent line (the linear model).

When changes are small, $dy$ is typically a very good approximation to $\Delta y$.

Linear approximation in action: estimating function values

When you need to approximate $f(b)$ and you know a “nice” nearby point $a$ where the function and derivative are easier, you do:

1. Choose $a$ close to $b$ where you can compute $f(a)$ and $f'(a)$ exactly.
2. Build the linearization $L(x)=f(a)+f'(a)(x-a)$.
3. Use $f(b)\approx L(b)$.

Example 1: Approximate a square root

Approximate $\sqrt{4.1}$.

Let $f(x)=\sqrt{x}$. Choose $a=4$ because $\sqrt{4}=2$ is easy and $4.1$ is close to $4$.

Compute the derivative:

$f'(x)=\frac{1}{2\sqrt{x}}$

Evaluate at $a=4$:

$f(4)=2$

$f'(4)=\frac{1}{2\cdot 2}=\frac{1}{4}$

Linearization at $a=4$:

$L(x)=2+\frac{1}{4}(x-4)$

Now plug in $x=4.1$:

$L(4.1)=2+\frac{1}{4}(0.1)=2+0.025=2.025$

So

$\sqrt{4.1} \approx 2.025$

What can go wrong: if you chose $a=0$ (not even differentiable for $\sqrt{x}$ in the same way) or chose a point far away, the approximation would be poor or invalid.

Example 2: Using differentials to approximate a change

Suppose the radius of a sphere is $r=10$ cm and increases by $0.1$ cm. Approximate the change in volume.

Volume:

$V=\frac{4}{3}\pi r^3$

Differentiate with respect to $r$:

$\frac{dV}{dr}=4\pi r^2$

Use differentials with $dr=0.1$ at $r=10$:

$dV \approx 4\pi (10)^2(0.1)=4\pi \cdot 100 \cdot 0.1=40\pi$

So the volume increases by approximately $40\pi$ cubic cm.

Common misconception: some students plug $r=10.1$ into the derivative. The derivative gives a rate at a specific radius; the differential method is “rate at the starting point times small change.”

Error and when linearization overestimates or underestimates

Linearization is an approximation, so it has error. While AP Calculus often focuses more on setting up and using the approximation than on strict error bounds, you should understand how concavity affects the approximation.

• If $f''(x)>0$ near $a$, the function is **concave up** and the tangent line tends to lie **below** the curve near $a$. So $L(x)$ tends to **underestimate** $f(x)$.
• If $f''(x)<0$ near $a$, the function is **concave down** and the tangent line tends to lie **above** the curve near $a$. So $L(x)$ tends to **overestimate** $f(x)$.

Quick illustration

For $f(x)=\sqrt{x}$, you can compute

$f''(x)=-\frac{1}{4x^{3/2}}$

For $x>0$, this is negative, so $\sqrt{x}$ is concave down. That means the tangent line approximation tends to be slightly above the true value (an overestimate) near the point.

Notation connections you’ll see

Linearization, tangent line approximation, and differential approximation often appear with slightly different notation but the same idea.

Exam Focus

• Typical question patterns:
  • “Find the linearization of $f(x)$ at $x=a$ and use it to approximate $f(b)$.”
  • “Use differentials to approximate the change in a quantity (area, volume, error) given a small measurement change.”
  • “Determine whether the linear approximation overestimates or underestimates using concavity or the sign of $f''(a)$.”
• Common mistakes:
  • Using $L(x)=f(x)+f'(x)(x-a)$ (mixing variable and point). You must evaluate at $a$: use $f(a)$ and $f'(a)$.
  • Choosing $a$ not close to the target $b$, which can make the approximation inaccurate.
  • Confusing $\Delta y$ (actual change) with $dy$ (linear predicted change) and treating them as exactly equal for large changes.

L'Hôpital's Rule

The problem L'Hôpital’s Rule solves: indeterminate forms

In limits, you sometimes run into expressions that look like they should be simple, but direct substitution gives a form that does not determine the limit. The two main forms handled in AP Calculus with L'Hôpital’s Rule are:

• Indeterminate form $\frac{0}{0}$
• Indeterminate form $\frac{\infty}{\infty}$

These are called “indeterminate” because they don’t tell you the limit. For example, $\frac{0}{0}$ could come from ratios that approach anything depending on how fast numerator and denominator approach zero.

L'Hôpital’s Rule is a technique that often turns a difficult limit of a quotient into a simpler limit involving derivatives.

What L'Hôpital’s Rule says (and what it doesn’t)

Suppose you have a limit

$\lim_{x \to a} \frac{f(x)}{g(x)}$

and as $x \to a$ you get either $\frac{0}{0}$ or $\frac{\infty}{\infty}$. If $f$ and $g$ are differentiable near $a$ (with $g'(x)\neq 0$ near $a$) and the derivative limit exists (or is infinite), then

$\lim_{x \to a} \frac{f(x)}{g(x)} = \lim_{x \to a} \frac{f'(x)}{g'(x)}$

The important meaning is: for these indeterminate quotient forms, the limit of the ratio is the same as the limit of the ratio of derivatives, provided the conditions are met.

What it does not say:

• It does not apply if you get a determinate form like $\frac{5}{0}$ or $\frac{0}{7}$.
• It does not automatically apply to products like $0\cdot \infty$ or differences like $\infty-\infty$ unless you algebraically rewrite them into a quotient that gives $\frac{0}{0}$ or $\frac{\infty}{\infty}$.
• It is not a replacement for algebraic simplification. Often factoring or rationalizing is faster and cleaner.

Why it works (intuition you can trust)

A full proof uses more advanced theorems, but the intuition connects to local linearity.

Near a point, differentiable functions behave like their tangent lines:

$f(x) \approx f(a) + f'(a)(x-a)$

$g(x) \approx g(a) + g'(a)(x-a)$

If $f(a)=0$ and $g(a)=0$ (the $\frac{0}{0}$ case), then near $a$:

$f(x) \approx f'(a)(x-a)$

$g(x) \approx g'(a)(x-a)$

So the ratio behaves like

$\frac{f(x)}{g(x)} \approx \frac{f'(a)(x-a)}{g'(a)(x-a)} = \frac{f'(a)}{g'(a)}$

That’s the big idea: when both numerator and denominator go to zero, their leading linear behavior (their derivatives) controls the ratio.

How to use L'Hôpital’s Rule (process)

When you see a limit of a quotient:

1. Try direct substitution.
2. If you get $\frac{0}{0}$ or $\frac{\infty}{\infty}$, you may apply L'Hôpital’s Rule.
3. Differentiate numerator and denominator separately.
4. Evaluate the new limit.
5. If you still get an indeterminate form, you may be able to apply the rule again.

A key habit: every time you differentiate, re-check the form by substitution. Do not apply the rule “automatically” multiple times without checking.

Worked examples

Example 1: A classic $\frac{0}{0}$ limit

Evaluate

$\lim_{x \to 0} \frac{\sin x}{x}$

Direct substitution gives $\frac{0}{0}$, so L'Hôpital’s Rule applies.

Differentiate numerator and denominator:

$\frac{d}{dx}(\sin x)=\cos x$

$\frac{d}{dx}(x)=1$

So the limit becomes

$\lim_{x \to 0} \frac{\cos x}{1}$

Now substitute $x=0$:

$\cos 0 = 1$

Therefore,

$\lim_{x \to 0} \frac{\sin x}{x} = 1$

Note: In many courses you learn this limit using geometry or the Squeeze Theorem. On the AP exam, L'Hôpital’s Rule is allowed in BC and is often the fastest route, but it’s still essential you recognize the indeterminate form before using it.

Example 2: Exponential vs polynomial (an $\frac{\infty}{\infty}$ limit)

Evaluate

$\lim_{x \to \infty} \frac{x}{e^x}$

As $x \to \infty$, numerator $\to \infty$ and denominator $\to \infty$, so the form is $\frac{\infty}{\infty}$.

Apply L'Hôpital’s Rule:

$\lim_{x \to \infty} \frac{1}{e^x}$

As $x \to \infty$, $e^x \to \infty$, so $\frac{1}{e^x} \to 0$.

Therefore,

$\lim_{x \to \infty} \frac{x}{e^x} = 0$

This result reflects an important growth-rate fact: exponentials dominate polynomials as $x \to \infty$.

Example 3: Applying L'Hôpital’s Rule more than once

Evaluate

$\lim_{x \to 0} \frac{1-\cos x}{x^2}$

Direct substitution gives $\frac{0}{0}$.

First application:

• Numerator derivative: $\sin x$
• Denominator derivative: $2x$

So the limit becomes

$\lim_{x \to 0} \frac{\sin x}{2x}$

Substitution still gives $\frac{0}{0}$, so apply again.

Second application:

• Numerator derivative: $\cos x$
• Denominator derivative: $2$

Now evaluate:

$\lim_{x \to 0} \frac{\cos x}{2} = \frac{1}{2}$

So

$\lim_{x \to 0} \frac{1-\cos x}{x^2} = \frac{1}{2}$

Common pitfall: differentiating the entire fraction with the quotient rule. L'Hôpital’s Rule says differentiate top and bottom separately, not “take the derivative of the quotient.”

Rewriting other indeterminate forms into quotients

AP Calculus BC commonly expects you to recognize forms like $0\cdot \infty$ or $\infty-\infty$ and rewrite them into a quotient that becomes $\frac{0}{0}$ or $\frac{\infty}{\infty}$.

Product form $0\cdot \infty$

Example:

$\lim_{x \to 0^+} x\ln x$

As $x \to 0^+$, $x \to 0$ and $\ln x \to -\infty$. This is an indeterminate product form.

Rewrite as a quotient. A good strategy is to turn one factor into a denominator:

$x\ln x = \frac{\ln x}{1/x}$

Now check the form as $x \to 0^+$:

• $\ln x \to -\infty$
• $1/x \to \infty$

So it’s $\frac{-\infty}{\infty}$, which is an indeterminate form of type $\frac{\infty}{\infty}$.

Apply L'Hôpital’s Rule:

$\lim_{x \to 0^+} \frac{\frac{1}{x}}{-\frac{1}{x^2}}$

Simplify the fraction:

$\frac{\frac{1}{x}}{-\frac{1}{x^2}} = -x$

Now take the limit:

$\lim_{x \to 0^+} -x = 0$

So

$\lim_{x \to 0^+} x\ln x = 0$

What can go wrong: forgetting that $\ln x$ is only defined for $x>0$ in real-valued calculus, so the limit must be one-sided.

Difference form $\infty-\infty$

Example:

$\lim_{x \to \infty} \left(\sqrt{x^2+x}-x\right)$

Directly, both terms go to infinity, giving $\infty-\infty$.

A standard fix is to multiply by the conjugate:

$\sqrt{x^2+x}-x = \frac{(\sqrt{x^2+x}-x)(\sqrt{x^2+x}+x)}{\sqrt{x^2+x}+x}$

The numerator simplifies:

$(x^2+x)-x^2 = x$

So the expression becomes

$\frac{x}{\sqrt{x^2+x}+x}$

Now evaluate the limit as $x \to \infty$. This is a quotient with $\frac{\infty}{\infty}$ form, so L'Hôpital’s Rule is available, but here algebra is even cleaner: factor $x$ from the denominator.

$\sqrt{x^2+x} = x\sqrt{1+\frac{1}{x}}$

So

$\frac{x}{\sqrt{x^2+x}+x} = \frac{x}{x\sqrt{1+\frac{1}{x}}+x} = \frac{1}{\sqrt{1+\frac{1}{x}}+1}$

Now as $x \to \infty$, $\frac{1}{x} \to 0$, so

$\lim_{x \to \infty} \frac{1}{\sqrt{1+\frac{1}{x}}+1} = \frac{1}{1+1} = \frac{1}{2}$

So the original limit equals $\frac{1}{2}$.

Important lesson: L'Hôpital’s Rule is powerful, but algebraic rewriting (like conjugates) may be the intended path and often avoids messy derivatives.

L'Hôpital’s Rule and local linearity (a conceptual connection)

It’s not an accident that linearization and L'Hôpital’s Rule appear together. Both rely on the idea that near a point, a differentiable function is well-approximated by its linear part.

• In linearization, you approximate $f(x)$ itself by a line.
• In L'Hôpital’s Rule, you approximate the ratio $\frac{f(x)}{g(x)}$ near a point where both behave like small linear changes, letting derivatives control the limit.

This connection helps you remember that L'Hôpital’s Rule is fundamentally about “comparing how fast things go to zero or infinity,” which is exactly what derivatives measure.

When not to use L'Hôpital’s Rule

Students often overuse it. On the AP exam, you want correctness first, but efficiency matters too.

Avoid or pause before using L'Hôpital if:

• The expression can be simplified by factoring, canceling, or rationalizing.
• You do not yet have $\frac{0}{0}$ or $\frac{\infty}{\infty}$.
• Differentiating will make the expression more complicated (for example, products of trig and exponentials might be easier with known limits or algebra).

Also remember: L'Hôpital’s Rule applies to limits (as $x\to a$, $x\to \infty$, or one-sided limits) of a quotient, under appropriate differentiability conditions.

Exam Focus

• Typical question patterns:
  • “Evaluate a limit that yields $\frac{0}{0}$ or $\frac{\infty}{\infty}$; justify use of L'Hôpital’s Rule and compute.”
  • “Rewrite an expression with indeterminate form $0\cdot \infty$ or $\infty-\infty$ into a quotient, then apply L'Hôpital’s Rule or simplify.”
  • “Compare growth rates using limits (polynomial vs exponential, exponential vs logarithmic) where L'Hôpital’s Rule quickly reveals the result.”
• Common mistakes:
  • Applying L'Hôpital’s Rule when the form is not $\frac{0}{0}$ or $\frac{\infty}{\infty}$ (for example, when substitution gives $\frac{0}{5}$ or $\frac{5}{0}$).
  • Differentiating using the quotient rule instead of differentiating numerator and denominator separately.
  • Forgetting domain restrictions and limit direction (for example, using $x\to 0$ instead of $x\to 0^+$ for $\ln x$).