Chapter 8: Further Applications of the Integral

Welcome to Last Minute Lecture.

This free chapter overview is designed to help students review and understand key concepts.

These summaries supplement not replaced the original textbook and may not be redistributed or resold.

For complete coverage, always consult the official text.

Usually when we talk about a calculus problem,

there's this expectation of pristine geometry,

like architectural drafting or something.

You have a clean equation, a perfect parabola, and you just find the exact area under that curve.

It feels safe.

It's contained right there on the chalkboard.

Yeah, it feels completely deterministic.

I mean, you plug in the numbers, you get a clean boxed -in area, and you just move on to the next problem.

But then you step outside and suddenly that pristine chalkboard feels, well, totally useless because today we are taking the definite integral out of the pure math classroom and throwing it into the messy chaotic real world.

Which is where it actually belongs, honestly.

Exactly.

Welcome to another deep dive from the Last Minute Lecture team.

I'm so thrilled you're here with us.

Today we're unpacking a massive stack of notes covering chapter eight of calculus,

early transcendentals.

Our mission is to see how the integral handles reality, from the unpredictability of human behavior to measuring curved surfaces, to the crushing weight of the ocean, and even balancing physical objects.

It's a huge shift from what you've probably studied so far.

It really is.

So to you listening right now, if you've ever stared at a page of dense calculus notation and wondered what it actually does outside of a classroom, stick around.

We're going to decode these theorems so you understand exactly why they work.

Let's jump right into the deep end.

How do we transition from measuring static geometric shapes to calculating the likelihood of random events?

Well, we have to reframe what we're actually measuring.

We start with what's called a continuous random variable.

It's typically denoted as a capital X.

Okay, capital X.

Right.

And think of X as the real world phenomenon you're tracking.

It could be, you know, the time you wait in a drive through, the lifespan of a light bulb, maybe the height of trees in a forest.

To figure out the probability of X happening, we use a probability density function or a PDF.

We write that as P of X.

And this function has to obey two absolutely non -negotiable rules to actually reflect reality.

Okay, the ground rules of chance.

What are they?

First, the function can never be negative.

Like it must be greater than or equal to zero everywhere.

Because you can't have a negative probability.

Exactly.

You simply cannot have a negative probability of an event happening.

That just doesn't exist.

Second, if you integrate that function over its entire domain, meaning you find the total area under the entire curve, that area must equal exactly one.

Right.

Because an area of one represents 100%.

If you're tracking wait times at a coffee shop,

every single customer will wait some amount of time between zero seconds and, well, infinity.

Yeah, hopefully not infinity, but mathematically, yes.

Right.

So there is 100 % chance that something in that domain happens.

That's the logic.

So if you want to find the probability that a customer waits between, say, 30 and 45 seconds, you don't need to look at the whole curve.

You just take the definite integral of your probability density function from a lower limit of 30 to an upper limit of 45.

The physical area of that specific slice gives you the exact probability of the event.

Okay, let's unpack this because there's a concept here that totally broke my brain at first.

We're calculating the probability for a window of time.

But what about a specific moment?

The notes have this conceptual insight that says the probability of a random variable taking on one exact specific value is zero.

Wait.

So the probability of an exact event happening is mathematically impossible.

I know.

It sounds crazy.

What's fascinating here is the distinction between continuous and discrete math.

Think about it in terms of the integral we just set up.

If you want the probability of a customer waiting exactly 30 seconds, not 30 .1, not 30 .001, but exactly 30 .30, you are integrating from 30 to 30.

Oh, and an interval with no width has zero area.

Precisely.

No width, no area, no probability.

Let me try an analogy on you.

It's like throwing a dart at a number line.

If I ask, you know, what's the probability you hit somewhere between the numbers one and two?

You'd say, sure, it's a wide target.

There's a good chance.

Yeah, easily.

But if I ask, what's the probability you hit the exact atomic center of the number 1 .5000000,

continuing into infinity.

The target is infinitely thin.

Like a physical dart can never hit an infinitely precise mathematical point.

That is a brilliant way to visualize it.

It's very similar to physical mass versus density.

The mass of an infinitely small point on a wire is zero, even if the wire itself is dense.

You need a physical segment, an interval to actually accumulate mass.

In calculus, you need an interval to accumulate probability.

Which makes me wonder about averages.

If we can only measure probability over an interval, how do we find the mean?

How do you find an average value from a curve that stretches out infinitely?

Well, you calculate the mean, which is represented by the Greek letter mu, by integrating the variable X multiplied by the probability function over the whole domain.

A great real world application of this is the exponential density function.

Right.

The formula with the constant E.

Yeah, it's used heavily in logistics to model waiting times between random events.

The formula is one over R times E raised to the power of negative T over R.

Okay, the drive, for example, from the text.

Let's say the historical average waiting time is 60 seconds.

So R is 60.

We can plug that 60 into the exponential formula.

Then to find the probability that the next customer waits between 30 and 45 seconds, we integrate that function from 30 to 45.

And the math is surprisingly straightforward.

You just find the antiderivative, plug in the downs, and the area under that curve turns out to be about

a 13 .4 % chance.

It's a very clean calculation.

But things get complicated when we look at the most famous probability curve of all, right?

The normal distribution,

the Gaussian distribution, the classic bell curve.

I mean, it's everywhere.

Tet scores, human heights, blood pressure.

It is ubiquitous in nature.

Its standard density function contains the term E to the power of negative X squared.

But there is a massive mathematical trap hidden inside that seemingly simple expression.

Wait, I've heard this equation is notoriously ugly to work with.

What's the trap?

The trap is that E to the negative X squared doesn't have an elementary antiderivative.

You're kidding.

No, seriously.

We have all these brilliant integration techniques, substitution, integration by parts.

But the most common distribution in nature is essentially immune to manual integration.

You literally cannot solve it with standard algebra.

That's a huge roadblock.

If the normal distribution governs the natural world and we can't integrate it by hand, how do statisticians get anything done?

This raises an important question about problem solving.

Mathematicians realized they couldn't fight the equation directly, so they built a workaround.

Theorem one in the text.

We use the standard normal cumulative distribution function.

Okay, that's a mouthful.

It is.

But since we can't integrate it on the fly, we rely on numerical approximation.

We take any specific normal distribution, say tree heights, and we convert our target values into a standard Z score.

You just take the value, subtract the mean, and divide by the standard deviation.

So we standardize it.

We take our specific messy real world bell curve and we shrink or stretch it to perfectly fit a universal standardized curve.

Exactly.

And once it's standardized, the heavy lifting is already done.

We just look up the exact probability area using a precomputed table or a calculator.

We completely bypass the impossible manual integration by just standardizing the geometry.

Wow.

Okay, so we've been using areas to measure invisible abstract concepts like chance and time.

Let's bring things back to the physical world for section 8 .2.

Arc length.

If you have a physical wavy line in space, say a drooping power line, how do we measure its literal physical length?

Because you can't exactly bend a straight wooden ruler around a curved cable.

You certainly can't.

So we do what calculus always does when faced with curves we approximate with straight lines and then we take it to the limit.

Right, the classic calculus move.

Always.

To find the arc length, which we call utes, we chop the smooth curve into microscopic straight line segments.

We essentially turn the curve into a jagged polygonal path.

Okay, I'm picturing zooming in until the curve looks like a staircase made of tiny slanted lines.

Yes.

And then we just use middle school geometry on those microscopic pieces.

Each tiny slanted line is the hypotenuse of a microscopic right triangle.

The horizontal leg of the triangle is our tiny change in x delta x.

The vertical leg is our tiny change in y delta y.

And we use the Pythagorean theorem.

A squared plus b squared equals c squared.

Exactly.

The length of that tiny segment is the square root of delta x squared plus delta y squared.

But how does that turn into an integral?

I mean, we can't just sit there adding up billions of tiny triangles manually.

The brilliant algebraic leap happens when we factor delta x out of that square root.

When you do that, you are left with a one plus delta y divided by delta x all squared inside the radical.

Delta y over delta x.

Wait, that's just the slope.

That's the derivative.

That's the magic.

By applying the mean value theorem as we shrink those segments towards zero, that fraction transforms entirely into the derivative of the function f prime of x.

It turns simple triangle math into an elegant Riemann sum.

When we push the number of segments to infinity, we get our formula.

Which is the integral of the square root of one plus the derivative squared.

You got it.

It's incredible how the Pythagorean theorem just scales up to infinity like that.

But I noticed in the notes that a lot of these arc length problems are, well, heavily curated.

Take the hanging cable example, mathematically known as a catenary.

The function that models that slump is the hyperbolic cosine.

Yeah, the math for that specific shape works out almost suspiciously well.

Right.

Because when you plug the derivative of hyperbolic cosine, which is hyperbolic sine, into our arc length formula, you get the square root of one plus hyperbolic sine squared.

But due to a very specific trigonometric identity, one plus hyperbolic sine squared perfectly equals hyperbolic cosine squared.

So the square root and the square just obliterate each other.

The radical vanishes, leaving you with trivial, easy integral.

It's like the universe wanted us to solve that specific problem.

Nature occasionally throws us a bone.

It does.

But for simpler looking curves, like a basic sine wave, the integral leaves you with the square root of one plus cosine squared.

And that radical doesn't vanish.

There is no exact algebraic formula for that, forcing us to fall back on numerical estimation tools like Simpson's rule.

Okay, here's where it gets really interesting.

We've traced the length of a curve.

But what if we add a dimension?

Surface area.

Instead of just tracing a curve, we take that curve and we spin it completely around an axis to create a 3D shape, like a vase or funnel.

How do we wrap our heads around the surface area of that?

We rely on the exact same logic we just built.

When you spin those tiny straight hypotenuse segments around an axis, they trace out slanted circular bands.

They're called frustums.

Imagine a bunch of rubber bands of slightly different sizes stacked together to form the shape.

So we just need the area of each little rubber band.

Which makes intuitive sense.

It's just the circumference of its circular path multiplied by the slanted width of the band itself.

Exactly.

The circumference is two times pi times the radius, and the radius is just the height of our function, f of x, and the slanted width of the band.

That is simply the arc length piece we derived a minute ago.

It's the square root part.

Right.

You multiply the circumference path, 2 pi f of x, by the arc length piece, integrate over the interval, and you have the exact surface area of the 3D shape.

That seems really straightforward until you run into Gabriel's horn.

We have to talk about this because it is a complete mind bender.

The text describes taking the graph of the fraction 1 over x, starting at 1 and going to infinity, and rotating it around the x -axis to get this infinite trumpet shape.

But the math proves that it has a finite volume, yet an infinite surface area.

Yeah, it's famously known as the painter's paradox.

How is it physically possible to hold a finite amount of space but have an infinite boundary?

It feels deeply contradictory.

But if we break down the mechanics of the formulas using improper integrals, the mystery dissolves.

Let's look at the volume first.

To find volume, you integrate pi times the radius squared.

Since our radius is 1 over x, squaring it gives us 1 over x squared.

Ah, and when you square a fraction, it shrinks extremely fast, like one -tenth becomes one -hundredth.

Right.

The cross -sections shrink to zero so rapidly as you move toward infinity that the total accumulated volume tapers off.

When you take the limit, it converges to exactly pi.

It's finite.

So theoretically, I could fill this infinite trumpet completely to the brim with exactly pi units of paint.

You could.

But the surface area formula doesn't square the radius.

It relies on the circumference, which just uses the radius itself, 1 over x.

Because 1 over x doesn't shrink nearly as fast as 1 over x squared, the surface area just keeps accumulating.

We use a comparison test here.

The surface area integral is greater than the integral of just 1 over x, which we know diverges to infinity.

So I can fill the inside of the horn with paint, but I don't have enough paint in the entire universe to paint the outside of it.

That's the paradox.

It shows us that when we push calculus to infinity, our everyday physical intuition breaks down.

The behavior of numbers at infinity doesn't always map to how we experience the physical world.

Wow.

Speaking of physical objects with massive surface areas, what happens if you take a gigantic surface and plunge it underwater?

Nature pushes back.

That brings us to Section 8 .3, Fluid Pressure and Force.

We enter the realm of hydrostatics.

Fluid pressure, p, is defined by a simple multiplier.

It's the density of the fluid, rho, times gravity, g, times the depth, h.

The deeper you go, the heavier the water above you, so the higher the pressure.

And the notes make a huge point about Pascal's principle here.

Pressure doesn't just push straight down from the weight of the water.

It pushes equally in all directions, perpendicular to any submerged surface.

Think about diving into the deep end of a pool.

Your ears feel that popping equally, no matter which way you tilt your head.

That's a great physical anchor.

But it creates a massive mathematical problem for civil engineers designing structures like dams.

If you have a giant vertical dam holding back a reservoir, the pressure at the top of the wall is relatively small.

But the pressure at the bottom is absolutely immense.

The pressure is constantly increasing at every single millimeter as you go down the wall.

So if pressure is a total force on the giant vertical dam, we can't just multiply one pressure value by the area of the wall.

No, we can't.

This is where we rely on the slicing strategy again.

Since we can't calculate the whole dam at once, we have to chop it up into pieces where the pressure is constant.

To do that, we slice the dam into infinitesimally thin horizontal strips, delta sha.

Why horizontal?

If we're looking at a big vertical wall, why not slice it vertically?

Because depth determines pressure.

If you take a vertical slice, the top of that slice is near the surface and the bottom is deep underwater.

The pressure is still changing along that piece.

But if you take a perfectly horizontal slice, a line from the left side of the dam to the right side, every single point on that line is at the exact same depth.

And if the depth is constant, the pressure is constant.

Oh, that is so clever.

On one specific millimeter thin horizontal slice, the water pressure is effectively just one uniform number.

So we can actually do the basic math.

Exactly.

We calculate the force on that tiny horizontal strip by multiplying its specific pressure by its area.

The area is simply the width of the dam at that depth, f of y, multiplied by its microscopic height, delta y.

We create an expression for the force on one strip and then we rhyme and sum all those horizontal strips together from the surface to the bottom.

We let the slices get infinitely thin, take the integral, and we've got a dynamic changing force across an entire structure.

Force equals rho times g times the integral of depth times width.

It's a masterclass in breaking a dynamic problem into static, solvable pieces.

The text even shows how this adapts to an inclined dam like a slanted reservoir wall.

The depth still dictates the pressure, but trigonometry introduces a sine factor to account for the longer slanted surface area.

The underlying logic of the integral scales perfectly to fit complex engineering problems.

It's literally holding back millions of tons of water with a single equation.

Incredible.

So we've got with these crushing uneven forces across a huge surface.

Let's pivot to our final challenge in 8 .4.

What if we want to take that exact same surface, that flat plate, and perfectly balance it on the tip of a pencil?

We're looking for the center of mass.

If we connect this to the bigger picture of physics, this is a problem that fascinated

Archimedes formulated the law of the lever over 2 ,000 years ago.

He realized that a seesaw balances when the mass multiplied by the distance on one side perfectly equals the mass multiplied by the distance on the other.

M1 times d1 equals m2 times d2.

Right.

A lighter person sitting farther out on a seesaw can perfectly balance a heavier person sitting very close to the center.

In physics, these mass times distance values are called moments.

Yes.

A moment measures the tendency of a mass to cause rotation around an axis.

Now Archimedes was dealing with single points of mass, like individual kids on a seesaw.

The challenge in calculus is scaling that up to a continuous 2D plate, a lamina, where the mass isn't in one spot, but spread out everywhere across the shade.

We have to assume the plate is made of the same material throughout, right?

A constant density, roll.

So how do we find that magical balance point, the centroid, with an exact x coordinate, x bar, and e coordinate, i bar?

We use the exact same horizontal and vertical slicing logic we used for the dam.

We divide the plate into tiny, measurable strips.

Let's say we want to find the moment with respect to the axis, which tells us how the shape wants to rotate left or right.

We integrate the distance of the slice from the axis, multiplied by the height of the slice, all multiplied by the density of the material.

So we sum up all those infinite little rotational forces.

But that just gives us the total tendency to rotate.

To find the actual physical coordinate of the balance point, we have to divide that total moment by the total mass of the entire plate.

And this leads to one of the most elegant mathematical cancellations you will ever see.

When you divide the moment integral by the mass integral, the density variable, rho, appears in both the numerator and the denominator.

Which means it completely cancels out.

Completely.

The balance point is dictated entirely by the geometry of the shape, not what it's made of.

You could cut a shape out of light cardboard and cut the exact same shape out of heavy lead.

As long as the density is uniform, they will balance perfectly on a pin in the exact same spot.

That is deeply satisfying.

The physical material just drops away, leaving only pure geometry.

But I have to admit, doing two massive integrals, one for the x coordinate and one for

every single time, feels, well, exhausting.

Are there any conceptual shortcuts?

There are two incredibly powerful tools we can lean on here.

First is the symmetry principle.

If a shape is perfectly symmetric like a circle or a perfectly mirrored parabola,

its centroid must lie somewhere on that line of symmetry.

You don't even have to calculate the balance for that axis, the geometry just gives it to you for free.

Always love a geometric freebie.

And the second shortcut.

The theorem of Pappas.

It is a brilliant piece of spatial reasoning that links 2D balance to 3D volume.

It states that if you want to find the volume of a solid of revolution,

say spinning a circle through space to make a donut shape,

a torus, you don't have to do a complex volume integral.

Oh, thank goodness.

Right.

You just take the 2D area of the starting shape and multiply it by the exact distance its centroid travels as it spins in a circle.

Volume equals area times 2 pi a bar.

Volume equals area times the circular path of the balance point.

Feels like a magic trick.

You bypass the complex 3D integration entirely by just knowing where the shape balances.

It's not magic.

It's just the incredible internal consistency of mathematics.

Everything is deeply connected.

And that really traces the whole through line of what we've covered today.

We started with the definite integral as just a rudimentary tool for finding flat areas under curves.

But we used it to accumulate tiny slices of chance to predict white times.

We used it to trace the physical lengths of hanging cables.

We summed up the crushing weight of ocean water on a dam, and we mathematically balanced a geometric shape on a pin point.

The definite integral is the ultimate tool for synthesis.

It takes the infinitely small, the microscopic changes in a system, and aggregates them into something profoundly significant.

To you listening, this is where calculus stops being just variables and notation on a page.

This is where it starts explaining the actual architecture of reality.

The messy, curved, unpredictable reality we live in.

It gives us the framework to tame the chaos.

It really does.

Which leaves me with a final thought for you to mull over as you close your notes today.

We've seen how integrals are simply the act of summing up infinitely small, seemingly insignificant slices to find absolute truth in physics and probability.

So, what other chaotic, overwhelming problems in your everyday life could be solved if you just learned to break them down into their absolute smallest possible moments?

A fantastic lens to view the world through.

Thank you so much for studying alongside the Last Minute Lecture team today.

Keep slicing up those complex problems, and we'll see you on the next deep dive.