< Previous Section  Home  Next Section > 
Suppose that h is a function with a rate of change function $r_h$. The function $r_h$ is a function, so we can inquire about its rate of change function, $r_{(r_h)}$.
Suppose that $r_h$ has a rate of change at the moment $x=k$. Then $r_{(r_h)}(k)$ can give us, in principle, the same kind of information about the behavior of $r_h(x)$ around $x=k$ as $r_h(k)$ gives us about the behavior $h(k)$ around $x=k$.
When $r_{(r_h)}(x)$ is positive at $x=k$, $r_h$ is increasing over an interval containing the value of k; when $r_{(r_h)}(x)$ is negative at $x=k$, $r_h$ is decreasing over an interval containing the value of k. When $r_{(r_h)}(x)=0$, or $r_{(r_h)}(x)$ is undefined at $x=k$, then $r_h$ has all the possibilities we investigated for $r_h$ in relation to h.
Rate Notation 
Differential
Notation 
Prime
Notation 
$$r_k(x)=2e^x\left(\sin(x)+\cos(x)\right)$$ $$r_{\left(r_k\right)}(x)=2e^x\left((\sin x + \cos x)+(\cos(x)\sin(x))\right)$$ 
$$\frac{d}{dx}\left(2\sin(x)e^x\right)=2e^x\left(\sin(x)+\cos(x)\right)$$ $$\frac{d^2}{dx^2}\left(2\sin(x)e^x\right)=2e^x\left((\sin x + \cos x)+(\cos(x)\sin(x))\right)$$ 
$$k'(x)=2e^x\left(\sin(x)+\cos(x)\right)$$ $$k''(x)=2e^x\left((\sin x + \cos x)+(\cos(x)\sin(x))\right)$$ 
Reflection 7.2.1. Confirm for yourself that $\dfrac{d}{dx}\left(2\sin(x)e^x\right)=2e^x\left(\sin(x)+\cos(x)\right)$ and that $\dfrac{d^2}{dx^2}\left(2\sin(x)e^x\right)=2e^x\left((\sin x + \cos x)+(\cos(x)\sin(x))\right)$
The advantage of rate notation for higher order rate of change functions is that it reminds you that you are thinking about a function that gives the rate of change of another function at each moment in the function's domain. A disadvantage of rate notation is that it becomes unwieldy for rate of change functions of order 3 or higher.
The advantages of differential notation are that (1) you are reminded explicitly of the defining formula that you are working with, and (2) it generalizes easily to rate of change functions of any order:
$$\frac{d^n}{dx^n}2\sin(x)e^x=\frac{d}{dx}\left(\frac{d^{n1}}{dx^{n1}}\left(2\sin(x)e^x\right)\right)$$
tells us that we are looking at the rate of change function for the $(n1)^{th}$order rate of change function of $2\sin(x)e^x$. A disadvantage of differential notation is that, when applied to a function's defining formula, it is easy for you to think that you are only acting symbolically on a formula, forgetting that you are finding the exact rate of change function for a function that has this formula is a rule of assignment.
Prime notation has the advantage that you can generalize $k'(x)$ to higher orders, e.g. $k^{(n)}(x)$ as indicating that you are speaking of the $n^{th}$order rate of change function for k. But it is also easy to forget that you are talking about a function that gives the rate of change of another function at each moment of the function's domain.
(Practice finding $n^{th}$order rate of change functions and representing them using the three different notation schemes.)
Second and thirdorder rate of change functions have direct applications in the sciences, in economics, and in characterizing behaviors of functions in general.
Acceleration as a quantity is often associated most closely with mechanics. If an object is "speeding up", a natural question is, "At what rate is it speeding up?". For example, suppose that an object at rest "speeds up" at a constant rate to a velocity of 15 m/sec over a period of 5 seconds. At what rate did it speed up? It increased its velocity at the rate of 3 m/sec every second, or 3 (m/sec)/sec.
Acceleration at a moment is the rate at which a rate of change function changes at that moment. 
Mechanical force is defined in terms of an accelerated mass, or $F=ma$, where m is a measure of mass, typically in kg, and a is the object's acceleration, typically in (m/sec)/sec. A Newton is a force of 1 kg accelerated at 1 (m/sec)/sec.
We can represent forces using the language of rate of change functions. If $s(t)$ represents a number of meters traveled in t seconds, then instead of $F=ma$ we could write $F=m\dfrac{d^2}{dt^2}s(t)$, as $F=r_{(r_s)}(t)$, or as $F=m \cdot s''(t)$. They all say the same thing  that force is the product of a mass and the rate at which its velocity changes.
We also often speak of economic quantities in terms of acceleration: "How fast is inflation changing at this moment?" Inflation is a rate of change of the cost of goods. To ask how fast it is changing is to ask about how the price of goods is accelerating.
Imagine that are a water skier, waiting in the water for the boat to pull the rope taught and accelerate to top speed (see Figure 7.2.1). Your arms will hurt if the boat accelerates too rapidly. That "hurt" is due to the boat trying to accelerate your body to top speed in too short a period of time. The rate at which you accelerate is called jerk. Let $s(t)$ represent how far the boat pulls you from a stationary position in t seconds. Then the jerk that you experience t seconds after starting is represented by $\dfrac{d}{dt}\left(\dfrac{d^2}{dt^2}s(t)\right)=\dfrac{d^3}{dt^3}s(t)$.
Figure 7.2.2. A dragster at the beginning of its race. The record for 1/4 mile (0.40 km) is 3.58 seconds with a top speed of 386 mi/hr (621.2 km/hr).
Suppose that f is a function of x and that it has a rate of change at every moment in an interval I. In Section 7.1 you learned that you can tell whether the function is increasing or decreasing at the moment $x=k$ in I by examining the sign of its rate of change function $r_f(x)$ at $x=k$:
Unfortunately, when $r_f(x)$ is not zero, we have no information about the way that f is increasing or decreasing. Figure 7.2.3 illustrates this ambiguity. Both functions f and g are increasing over the interval $x>0$. However, while f increases, the value of $r_f$ also increases, whereas as g increases, the value of $r_g$ decreases.
Play the animation in Figure 7.2.4. How would you describe the behavior as k varies of the red graph that is tangent to the graph of $y=f(x)$ at $x=k$ (Figure 7.2.4, Left)? Many people describe the red graph as having a rate of change that gets smaller (less steep), equals zero, and then gets larger. However, if you look at the value of $r_f(k)$ as k varies, you will see that it is always getting larger. This is because a sequence of numbers like 10, 9, 8, ..., 1, 0, 1, 2, 3, ... is increasing. Each term is larger than the one preceding it.
Reflection 7.2.2. Describe the behavior of the graph in the right side of Figure 7.2.4. How does its rate of change vary as the value of k varies.
Reflection 7.2.3. Determine $\dfrac{d^2}{dx^2}$ of both f and g as given in Figure 7.2.4. What do their values tell you about $r_f$ and $r_g$? About f and g?
Suppose that f is a function, and that $r_f(k)$=0. If $r_f(x)$ is continuous and changes sign around $x=k$, then f is said to have a point of inflection at $x=k$. 
Another way to put this is that f has a point of inflection at $x=k$ if $r_f(x)$ changes from increasing to decreasing or from decreasing to increasing around $x=k$. A consequence of $r_f(x)$ changing from increasing to decreasing or decreasing to increasing is that $f''(k)=0$ and $f''(x)$ changes sign around $x=k$.
The function q defined as $q(x)=(x2)^3+2$ has a point of inflection at $x=2$: $q'(x)=3(x2)^2$, so $q'(2)=0$; $q''(x)=6(x2)$, so $q''(2)=0$; finally $q''(x)$ changes sign from negative to positive around $x=2$, therefore $q'(x)$ changes from decreasing to increasing, making a point of inflection at $x=2$. The graph in Figure 7.2.5 illustrates this.
Higher order rate of change functions are used in defining physical quantities like force, acceleration, and jerk. They are also useful for characterizing the behavior of functions by characterizing the behavior of their rates of change functions. In the case of behaviors of functions in general.
The table given below summaries the information you can glean by looking at values of first and secondorder rate of change functions in combination.
$f''(k)>0$ 
$f''(k)<0$ 
$f''(k)=0$ 

$f'(k)>0$ 
f(x) increasing around $x=k$ $r_f(x)$ increasing around $x=k$ 
f(x) increasing around $x=k$ $r_f(x)$ decreasing around $x=k$ 
f(x) increasing around $x=k$ Ambiguous regarding what $r_f(x)$ is doing around $x=k$ 
$f'(k)<0$ 
$f(x)$ decreasing around $x=k$ $r_f(x)$ increasing around $x=k$ 
$f(x)$ decreasing around $x=k$ $r_f(x)$ decreasing around $x=k$ 
$f(x)$ decreasing around $x=k$ Ambiguous regarding what $r_f(x)$ is doing around $x=k$ 
$f'(k)=0$ 
Local minimum at $x=k$ 
Local maximum at $x=k$ 
Ambiguous. Requires further investigation. Possible that $\left(k,f(k)\right)$ is a point of inflection.

Table 7.2.1
$f(x)=(\ln x)\cos(x^2)$
$h(x)=g(f(x))$, where g and f are defined as above.
$k(x)=\log(\cos(x))$
The graph of $y=k(t)$ for a function k is displayed below. Over what intervals, approximately, is $k'$ increasing? Decreasing?
The graph of a function h is displayed below. Complete the tables below, giving approximate values where $h'$ and $h''$ are zero, and giving approximate intervals over which $h'$ and $h''$ are increasing or decreasing.
h' is zero at these values of its independent variable 
h' is increasing over these intervals  h' is decreasing over these intervals 
h'' is zero at these values of its independent variable 
h'' is increasing over these intervals  h'' is decreasing over these intervals 
Two of these functions have another of the three as an accumulation function. Which ones have another as an accumulation function, and which function is its accumulation function?
Masakazu visited Disneyworld to enjoy a new ride, called the Rocket Sled. In the Rocket Sled, Masakazu was strapped into a car seat, facing forward, with a sixpoint harness across his chest and shoulders. The car accelerated forward and backward along a long, straight rail. The graph below shows Masakazu's velocity in relation to the number of seconds since the Rocket Sled began moving.
During what intervals of time did Masakazu feel pressure on his back? Explain.
When did Masakazu feel the greatest pressure on his chest? The greatest pressure on his back? Explain.