We’re going to continue to think about different applications of definite integrals: what they can measure and how we can construct these integral formulas. In this section, we’re going to add two more formulas for two more measurements. Before we get far into this discussion, we want to center the important parts of our discussion.
Sure, it is worth noting that, in this section, we’ll add a 1-dimensional measurement of size to our list of things an integral can measure. We have talked about a 2-dimensional measure of size (area) and a 3-dimensional measure of size (volume), but we’ll add length to the list now! We’ll also add a 2-dimensional extension of perimeter to the list when we talk about surface area. That’s cool!
But, more importantly, we’re going to see how we can construct an integral formula from a Riemann sum, and we’re going to get some experience constructing a Riemann sum to measure the thing we care about. In our study of integrals, it might not actually be that important to know how to calculate the specific kinds of volumes or lengths that we’re talking about. But we can get some experience with using some formulas from a pretty comfortable field (geometry) to get some experience with the slice-and-sum process. And this process is a useful one to know! We want to see that a definite integral is more than just an area under a curve, and we want to be able to look at an integral formula for some measurement or calculation and see some of the parts of that formula that could be familiar.
SubsectionIntegrals for Evaluating the Length of a Curve
When we talk about arc length, we might think of the length of some portion of a circle. Here, we’ll use it to refer to the length of some more general curve. We’ll graph a function and think about how long the curve of the graph is from some point to another point.
This might be a reminder of something we already knew, but let’s make sure we are certain: when we calculate distances, we’re really just using the Pythagorean Theorem! We can square the vertical distance between the points and the horizontal distance between the points, and then the length of the straight line connecting two points is:
If we think about the slice-and-sum technique, then we’ll want to visualize the \(k\)th slice of whatever we’re trying to measure. In this case, that means we’ll divide the curve up into equally-wide slices and calculate the length of each subsection of the curve. We’ll make a recognizable assumption: we’ll assume that the curve is actually a straight line between the end points, and calculate that length.
In order to calculate \(\ell_k\text{,}\) the straight-line length connecting the end points of the \(k\)th subinterval, we can use the Pythagorean Theorem or distance formula (from Activity 6.5.1).
We’re using \(\Delta y_k\) to denote the vertical distance between the end points of the \(k\)th subinterval because we expect these to differ for each subinterval. We don’t need to do this for \(\Delta x\text{,}\) since we’ve been slicing things into equally-wide subintervals this whole time.
Before we can do anything, we need to try to manipulate this sum so that it is in the form of a Riemann sum. What does this mean? What are some of the things required for the Riemann sum structure that we don’t have here? Feel free to look back at Definition 5.2.3 Riemann Sum to remind yourself what elements are needed for a Riemann sum.
Notice, first, that we need a function evaluated at any single input on the subinterval: \(f(x_k^*)\text{.}\) In our version, we have a function evaluated twice at very specific inputs:
We also need to have this function multiplied by \(\Delta x\text{.}\) In our sum, we have \(\Delta x\) as a part of the function itself, under the square root. We’ll want to move this \(\Delta x\) outside of the root. Let’s start there.
This looks better! We have \(\Delta x\) floating around at the end of our sum, ready to turn into \(dx\) once we apply the limit as \(n\to\infty\text{.}\)
The inside of our root, though, is still a bit messed up. We would like a single function of \(x_k^*\text{,}\) any \(x\)-value from the \(k\)th subinterval. Instead, we have a function involving the two \(x\)-values of the end points and we still have \(\Delta x\) involved in this part!
But we can notice something about \(\frac{\Delta y_k}{\Delta x}\text{:}\) it really is the slope of the straight line! Can we use a function to represent this? We absolutely can approximate this slope using a function: the derivative!
If we pick some point, \((x_k^*, f(x_k^*))\text{,}\) on the \(k\)th subinterval, then we can approximate \(\frac{\Delta y_k}{\Delta x}\) with \(f'(x_k^*)\text{.}\) This is a fine approximation of this slope (and the Mean Value Theorem guarantees that there is a point on the subinterval where \(f'(x_k^*) = \frac{\Delta y_k}{\Delta x}\) exactly), but the real magic will happen when \(n\to\infty\text{.}\) The definition of the Derivative at a Point will make sure that these slopes are equal in the limit!
If \(f(x)\) is continuous on the interval \([a,b]\) and differentiable on \((a,b)\text{,}\) then the length of the curve \(y=f(x)\) from \(x=a\) to \(x=b\) is:
SubsectionIntegrals for Evaluating the Surface Area of a Solid
Moving from the length of some curve towards calculating the surface area of some solid of revolution won’t be hard: we’ll use the length formula in our procedure!
Instead of forming a rectangle for the \(k\)th slice, we’ll do the same thing that we did for arc length: we’ll connect the end points of the \(k\)th subinterval. This will create a trapezoid.
When we revolve the curve \(y=f(x)\) around the \(x\)-axis, we can see not just the solid created by the curve, but the solid representing this \(k\)th slice.
In order for us to find the surface area of this \(k\)th slice, we’ll think about how "far" the diagonal line revolves. This is based on the circumference of each circular end of our slice, which means we have two radii to consider: the function outputs at both endpoints of the \(k\)th subinterval:
Instead of averaging the large and small radii from the end-points, we’ll just select the one function output to represent this "average" radius. In the limit as \(n\to\infty\text{,}\) things will work out, since this randomly selected radius will become exactly equal to the average radius in the limit since \(\Delta x\to 0\text{.}\)
Let \(f(x)\) be a continuous function with \(f(x)\geq 0\) on the interval \([a,b]\) and differentiable on \((a,b)\text{.}\) If the region bounded by \(f(x)\) and the \(x\)-axis from \(x=a\) to \(x=b\) is revolved around the \(x\)-axis, then the surface area of the resulting solid is:
\begin{equation*}
A = 2\pi\int_{x=a}^{x=b} f(x)\sqrt{1+(f'(x))^2}\;dx\text{.}
\end{equation*}
The formula for arc length of the function \(f(x)\) on the interval \([a,b]\) is \(\displaystyle L = \int_{x=a}^{x=b} \sqrt{1+f'(x)^2}\;dx\text{.}\) Explain how this definition is built, using the slice-and-sum method. Make sure to explain how we the Pythagorean Theorem is involved.
In the integral formula \(\displaystyle A = \int_{x=a}^{x=b} 2\pi f(x) \sqrt{1+f'(x)^2} \; dx\text{,}\) what does \(f(x)\) represent? What about \(2\pi\text{?}\)
For each of the following curves and intervals, set up the surface area of the solid formed when the curve is revolved around the \(x\)-axis. Do not evaluate the integral.