I learned that the volume of a sphere is $\frac{4}{3}\pi r^3$, but why? The $\pi$ kind of makes sense because its round like a circle, and the $r^3$ because it's 3-D, but $\frac{4}{3}$ is so random! How could somebody guess something like this for the formula?
|
|
In addition to the methods of calculus, Pappus, and Archimedes already mentioned, Cavalieri's Principle can be useful for these kinds of problems. Suppose you have two solid figures lined up next to each other, each fitting between the same two parallel planes. (E.g., two stacks of pennies lying on the table, of the same height). Then, consider cutting the two solids by a plane parallel to the given two and in between them. If the cross-sectional area thus formed is the same for each of the solids for any such plane, the volumes of the solids are the same. If you're willing to accept that you know the volume of a cone is 1/3 that of the cylinder with the same base and height, you can use Cavalieri, comparing a hemisphere to a cylinder with an inscribed cone, to get the volume of the sphere. This diagram (from Wikipedia) illustrates the construction: look here Consider a cylinder of radius $R$ and height $R$, with, inside it, an inverted cone, with base of radius $R$ coinciding with the top of the cylinder, and again height $R$. Put next to it a hemisphere of radius $R$. Now consider the cross section of each at height $y$ above the base. For the cylinder/cone system, the area of the cross-section is $\pi (R^2-y^2)$. It's the same for the hemisphere cross-section, as you can see by doing the Pythagorean theorem with any vector from the sphere center to a point on the sphere at height y to get the radius of the cross section (which is circular). Since the cylinder/cone and hemisphere have the same height, by Cavalieri's Principle the volumes of the two are equal. The cylinder volume is $\pi R^3$, the cone is a third that, so the hemisphere volume is $\frac{2}{3} \pi R^3$. Thus the sphere of radius $R$ has volume $\frac{4}{3} \pi R^3$. |
|||||||||
|
|
The volume of a sphere with radius $a$ may be found by evaluating the triple integral $\displaystyle V = \iiint_{\ S}\mathrm dx\,\mathrm dy\,\mathrm dz,$ where $S$ is the volume enclosed by the sphere $x^2+y^2+z^2=a^2$. Changing variables to spherical polar coordinates, we obtain $\displaystyle V = \int_{0}^{2\pi}\mathrm d\phi\int_{0}^{\ \pi}\mathrm d\theta\int_{0}^{\ a}r^2\sin\theta \mathrm dr = \int_{0}^{2\pi}\mathrm d\phi\int_{0}^{\pi}\sin\theta \mathrm d\theta\int_{0}^{\ a}r^2\mathrm dr = \frac{4\pi a^3}{3},$ as expected. |
||||
|
|
A complete answer using the disk method would be the following. If you revolve $ y = \sqrt{ r^2 - x^2 } $ about the x-axis (and form a solid) you get the volume of a sphere.
Form a disk with height $f(x)$, and find its area.
The area of the red disk above is $ \pi r^2 $, or $ \pi f^2(x) $, or we could say $ \pi \sqrt{ r^2 - x^2 }^2 = \pi (r^2 - x^2) $ at any point x between x=-r and x=+r. To find the volume of a sphere of radius r then, just add up the areas of infinitesimally thin disks on as x goes from -r to +r. To work it out: $ \int_{-r}^{r}{ \pi \sqrt{ r^2 - x^2 }^2 dx } $ $ = 2 \pi \int_0^r{ r^2 - x^2 dx } $ $ = 2 \pi ( r^2x - \frac{1}{3}x^3 )|_0^r $ $ = 2 \pi ( r^3 - \frac{1}{3}r^3 ) $ $ = \frac{4\pi}{3}r^3 $ |
||||
|
|
|
This was first "guessed at" by Archimedes, by what fraction of the volume of a cylinder a sphere is. That is, a sphere that is contained within that cylinder. From that, he figured out $4/3$. You can do it these days with the tools of calculus. One way would to use the Disk Method, over the graph of a semicircle. Another would be to use spherical coordinates and calculate the 3D Integral. All of these methods should give you $\frac{4}{3} \pi r^3$ Note that it's not just $r^3$ because it's "$3D$"; but also because it's the only way the dimensional analysis would work; that is, the units would have to be in $\text{distance}^3$. But that might be what you said, but just clarifying. |
||||
|
|
|
Pappus's centroid theorem (second theorem) says that the volume of a solid formed by revolving a region about an axis is the product of the area of the region and the distance traveled by the centroid of the region when it is revolved. A sphere can be formed by revolving a semicircle about is diameter edge. The area of the semicircle is $\frac{1}{2}\pi r^2$. The centroid of the semicircle can be found by intersecting two lines that both divide the area of the semicircle into two equal parts. One such line is perpendicular to the diameter edge through the center of the semicircle (this is a line of symmetry of the semicircle). Another such line is parallel to the diameter edge, $\frac{4r}{3\pi}$ away from it (verification of this is left as an exercise for the reader). When revolved about the diameter edge of the semicircle, the centroid travels $2\pi\cdot\frac{4r}{3\pi} = \frac{8}{3}\cdot r$, so the volume of the sphere is $\frac{1}{2}\pi r^2\cdot\frac{8}{3}\cdot r = \frac{4}{3}\pi r^3$. |
||||
|
|
|
To give you a larger perspective, the volume of an $n-1$-sphere in $\mathbb{R}^n$ is given by $C_n r^n$, where $r$ is the radius and $C_n=\frac{\pi^{n/2}}{\Gamma(\frac{n}{2} + 1)}$. For even $n$, this reduces to $C_n =\frac{\pi^{n/2}}{\left(\frac{n}{2}\right)!}$, and for odd $n$, it becomes $C_n =2^{(n+1)/2}\frac{\pi^{(n-1)/2}}{n!!}$, where $n!!$ denotes the double factorial. Thus in the case $n=3$, the constant is $2^2 \frac{\pi^1}{3!!} = \frac{4 \pi}{3}$. The point of this is to show you that the general formula also involves factors in the denominator, and that the formula for $n=3$ is not 'random' but rather fits into a general pattern. |
|||
|
|
|
Integration. $$\int 4\pi\cdot r^2 dr = 4/3 \pi r^3.$$ |
|||||||||||||||||
|
|
To expand on what anosov_diffeomorphism said: Suppose we agree that the surface area of a sphere is 4 π r2 (If we do not agree on this, that would be a great question to ask on this site, and then someone can edit this answer to link to it!). Now if I have a sphere of radius r, and I increase the radius by a tiny amount, dr, then the new, expanded sphere has a volume that is bigger, by the volume of the thin spherical shell that was just added. If dr is small enough, then it doesn't much matter whether we look at the inner or outer radius of that shell, and we can just say that the shell's volume is approximately 4 π r2 dr. So the difference in volume between two spheres of radius $r_1$ and $r_2$ is $\int_{r_1}^{r_2}$ 4 π r2 dr, which is 4/3 π r^3 |
|||||
|
|
I wanted to post this as a comment to the answer by Justin L. but it is my first posting here so I couldn't. Archimedes was so proud of his proof that the volume of a sphere is two-thirds of the volume of a cylinder of the same height and diameter that he asked that a sculpture illustrating this would be placed at his tomb. The tomb was later found by the Roman orator Cicero who describes it as follows: "I managed to track down his grave. The Syracusians knew nothing about it, and indeed denied that any such thing existed. But there it was, completely surrounded and hidden by bushes of brambles and thorns. I remembered having heard of some simple lines of verse which had been inscribed on his tomb, referring to a sphere and cylinder modelled in stone on top of the grave. And so I took a good look round all the numerous tombs that stand beside the Agrigentine Gate. Finally I noted a little column just visible above the scrub: it was surmounted by a sphere and a cylinder." The American painter Benjamin West imagined the scene in his 1797 painting "Cicero Discovering the Tomb of Archimedes". |
||||
|
|
|
In the diagram below on the left, we will assume that the the size of $\triangle PQR$ is infinitesimal compared to $\triangle PNO$, and thus, the green arc and the segment $\overline{PR}$ are essentially equal in length. Note that both $\angle NPO$ and $\angle QPR$ are complementary to $\angle OPQ$, and therefore, equal. Thus, the right triangles $\triangle PNO$ and $\triangle PQR$ are similar (corresponding sides are colored similarly). Since $\triangle PNO$ and $\triangle PQR$ are similar, $$ \overline{NP}\cdot\overline{PR}=\overline{OP}\cdot\overline{PQ}\tag{1} $$ The inner circumference of the green annulus on the sphere to the right is $2\pi\overline{NP}$ and its outer circumference is $2\pi\left(\overline{NP}+\overline{QR}\right)$, while its width is $\overline{PR}$. Therefore, its area is between $2\pi\overline{NP}\cdot\overline{PR}$ and $2\pi\left(\overline{NP}+\overline{QR}\right)\cdot\overline{PR}$. The circumference of the red strip on the cylinder to the right is $2\pi\overline{OP}$, while its width is $\overline{PQ}$. Therefore, its area is $2\pi\overline{OP}\cdot\overline{PQ}$. According to $(1)$, the difference between the area of the green annulus on the sphere and the red strip on the cylinder is less than $2\pi\overline{QR}\cdot\overline{PR}$. Summing $\overline{QR}$ as we walk down the sphere gives $2\overline{OP}$ (one as $\overline{NP}$ increases and one as it decreases), thus, the difference between the area of the sphere and the area of the cylinder is less than $4\pi\overline{OP}\cdot\max\overline{PR}$, which can be made to vanish by making $\overline{PR}$ as small as we wish. Thus, the area of the sphere is equal to the area of the cylinder. Since the circumference of the cylinder is $2\pi r$ and its height is $2r$, we get that the area of the cylinder, and therefore the sphere, is $4\pi r^2$. $\hspace{4mm}$ To compute the volume of the sphere, we just sum the surface area of all spherical shells times their thickness: $$ V=\int_0^r4\pi t^2\,\mathrm{d}t = \frac43\pi r^3\tag{2} $$ |
||||
|
|
|
If we already knew that the surface of the sphere is $4\pi r^2$, then we could go like this: We could divide the surface into small polygons (you can imagine them infinitesimaly small). The area of these polygons approximate the area of the surface. Now the volume of the sphere would be approximated by the pyramids obtained from the polygons, each of them having volume $(1/3)*\text{base}*\text{height}$. Since height is approximately $r$ (as we make the polygons smaller) and the bases together give the area of $4\pi r^2$, we arrive to $V=\frac43 \pi r^2$. (Although I have to admit that I liked this better as a way of deriving the formula for the surface area from the volume.) |
|||||||||
|
|
To add to the approach of @bobobobo, it can be shown that we can construct the volume $V_n$ of an $n$ dimensional sphere from the volume $V_{n-1}$ of an $n-1$ dimensional sphere using the following recurrence relation: $$V_n = 2r V_{n-1} \int_0^{\frac{\pi}{2}}\cos^n\theta d\theta$$ using this relation starting with $V_0 =1$ we get: $$V_1 = 2r V_0 = 2r$$ $$V_2 = \frac{\pi}{2}r V_1 = \pi r^2$$ $$V_3 = \frac{4}{3}r V_2 = \frac{4}{3}\pi r^3$$ continuing... $$V_4 = \frac{3\pi}{8}r V_3 = \frac{1}{2}\pi^2 r^4$$ |
|||
|
|
|
I add a Chinese version for fun. The ancient Chinese had another way to calculate this volume.
the principle is same as Cavalieri's Principle. the difference is to use two pipe to packing the sphere. The Chinese name is "mouhe square cover".so when you look at each cutting face, you will see there is square to a circle,(more picture see http://phdfishman.blogspot.com/2010/02/blog-post_07.html, that means the volume ratio of sphere and the square cover is $ πr^2/(2r)^2= π/4.\ $. now the question become to calculate the square cover(white) volume. it is also very difficult. but when you look at a cubic(red) pack the square cover(white),then the extra area(a big square minus the square of the square cover), it is same as a 4 mall square(blue).when the cutting face moving, the blue square will form 4 small Pyramids,total 8 pyramids. now we can know the volume of the red cubic minus the volume of the square cover(white),which is $ 8 *\frac{1}{3} r^2*r=\frac{8}{3} r^3 \ $,then we get the the square cover volume is $ (2r)^3-\frac{8}{3}r^3=\frac{16}{3}r^3 $ finally the sphere volume is $ \frac{π}{4}*\frac{16}{3}r^3=\frac{4π}{3}r^3 $ now you can see 3 is from the pyramids and 4 is from the cubic! it is not a random.
I add another picture to show the math reason . |
||||
|
|



