Why Bound a Computed Volume?
A definite integral can give the exact volume of a solid, but an answer is more convincing when it also passes an independent check. Bounds can show that a computed volume is too large or too small, even when the integral is difficult. They also give a useful estimate when an exact calculation is not needed.
For a solid sliced perpendicular to the \(x\)-axis, let \(A(x)\) denote the area of its cross section at position \(x\). Over an interval from \(x=a\) to \(x=b\), the volume is found by accumulating those cross-sectional areas:
The main idea in this tutorial is simple: if every cross-sectional area is at least some lower bound and at most some upper bound, the total volume is bounded by the corresponding accumulated areas.
Bounds from Cross-Sectional Areas
The constants \(m\) and \(M\) are lower and upper bounds for area, not for a radius or side length. If the bounds are given for a radius or side length, first convert them to area using the appropriate cross-sectional formula.
For example, a disk of radius \(r\) has area \(\pi r^2\), while a square of side length \(s\) has area \(s^2\). A bound on \(r\) or \(s\) therefore produces an area bound before it produces a volume bound.
Why the Bounds Work
The volume formula treats each thin slice as having area \(A(x)\) and thickness \(dx\). On any small interval of width \(\Delta x\), if every cross-sectional area lies between \(L\) and \(U\), the slice’s volume contribution lies between approximately \(L\Delta x\) and \(U\Delta x\). Adding such contributions over the interval gives lower and upper sums. As the widths of the subintervals shrink, these sums approach the corresponding definite integrals.
For every \(x\), use \(L(x)\leq A(x)\leq U(x)\).
On a subinterval of width \(\Delta x\), the lower and upper volume contributions are bounded by the area bounds times \(\Delta x\).
The resulting lower and upper sums converge to the integrals of \(L\) and \(U\), while the sums for the actual areas converge to \(V\). The inequality is preserved in the limit.
This is the comparison property of the definite integral: integrating pointwise inequalities preserves their order. No assumption that the slices line up to form a particular cylinder or prism is needed.
Worked Example: Disk Areas Between Two Constant Bounds
Worked Example: Disk Areas Between Two Constant Bounds
A solid extends from \(x=0\) cm to \(x=2\) cm. Its cross sections perpendicular to the \(x\)-axis are disks whose radii are \(r(x)=1+x\) cm. Find the exact volume and compare it with bounds from the smallest and largest cross-sectional areas.
The cross-sectional area is
On \(0\leq x\leq2\), the radius increases from \(1\) cm to \(3\) cm. Therefore the area is at least \(\pi\ \text{cm}^2\) and at most \(9\pi\ \text{cm}^2\). The interval length is \(2\) cm, so
Now compute the exact volume using the cross-sectional area:
Check the comparison: \(2\pi\leq 26\pi/3\) because \(2\leq26/3\), and \(26\pi/3\leq18\pi\) because \(26/3\leq18\). The exact answer lies inside the bounds.
The constant-area calculations are the volumes of cylinders with the same length and with cross-sectional areas \(\pi\) and \(9\pi\). They give valid lower and upper volume comparisons. They do not, by themselves, show that either cylinder physically fits inside or encloses the solid: knowing the areas of the disks does not specify where their centers are from one slice to another.
Simple Solid Estimates Are Checks, Not Always Bounds
A convenient quick estimate is to multiply the interval length by a representative cross-sectional area. This is the volume of a constant-cross-section solid, such as a prism or cylinder, with that chosen area and length. It can be a useful estimate, but it is not automatically an upper or lower bound: that conclusion requires a pointwise area comparison.
Worked Example: A Midpoint-Area Estimate for Square Slices
A solid has square cross sections perpendicular to the \(x\)-axis for \(0\leq x\leq1\) m. Each square has side length \(s(x)=1+x\) m. Find the exact volume, a simple midpoint-area estimate, and constant bounds.
The area is \(A(x)=(1+x)^2\ \text{m}^2\). Since \(1\leq1+x\leq2\) on the interval, squaring these nonnegative quantities gives \(1\leq A(x)\leq4\), with area measured in square metres. Thus
The exact integral is
At the midpoint \(x=1/2\) m, the side length is \(1+1/2=3/2\) m, so the area there is \((3/2)^2=9/4\ \text{m}^2\). Using this representative area gives the estimate
The midpoint estimate is close to the exact value: \(7/3-9/4=28/12-27/12=1/12\ \text{m}^3\). It is not an upper or lower bound established by the midpoint calculation alone. The exact value does satisfy the constant bounds: \(1\leq7/3\leq4\). As a final check, differentiating the antiderivative \(x+x^2+x^3/3\) gives \(1+2x+x^2=(1+x)^2\), the original area function.
Refining Bounds with Smaller Intervals
Constant bounds over a long interval can be wide. A standard improvement is to split the interval into smaller pieces and bound the cross-sectional area separately on each piece. When the area changes continuously and monotonically, its endpoint values give the minimum and maximum on each subinterval. Adding the subinterval bounds gives a tighter overall bracket.
Worked Example: Tightening Bounds by Splitting the Interval
For \(0\leq x\leq1\) cm, suppose the cross-sectional area is \(A(x)=\pi(4-x^2)\ \text{cm}^2\). First find the exact volume. Then bound it using one interval and using two equal subintervals.
The area decreases on \([0,1]\): if \(x\) increases from \(0\) to \(1\), then \(x^2\) increases from \(0\) to \(1\), so \(4-x^2\) decreases from \(4\) to \(3\). The one-interval bounds are therefore \(3\pi\leq A(x)\leq4\pi\), in square centimetres. Since the interval has length \(1\) cm,
The exact value is
Now split at \(x=1/2\) cm. On \([0,1/2]\), the area ranges from \(4\pi\) down to \(\pi(4-(1/2)^2)=15\pi/4\). This subinterval has width \(1/2\) cm, so its volume contribution is between \((1/2)(15\pi/4)=15\pi/8\) and \((1/2)(4\pi)=2\pi\), in cubic centimetres.
On \([1/2,1]\), the area ranges from \(15\pi/4\) down to \(3\pi\). Its contribution is between \((1/2)(3\pi)=3\pi/2\) and \((1/2)(15\pi/4)=15\pi/8\). Adding the lower bounds and the upper bounds separately gives
Check the exact result against the refined bounds using a common denominator: \(27\pi/8=81\pi/24\), \(11\pi/3=88\pi/24\), and \(31\pi/8=93\pi/24\). Hence \(81\pi/24\leq88\pi/24\leq93\pi/24\), as required. The refined bracket is narrower than the original \(3\pi\leq V\leq4\pi\): its width is \(31\pi/8-27\pi/8=\pi/2\), compared with \(4\pi-3\pi=\pi\).
A Common Mistake: Confusing Area Bounds with Containment
Suppose every cross section has area between \(m\) and \(M\). It is valid to conclude that the volume lies between \(m(b-a)\) and \(M(b-a)\). But this does not necessarily mean that a fixed cylinder of cross-sectional area \(m\) sits inside the solid, or that a fixed cylinder of area \(M\) surrounds it.
Area alone does not locate a cross section in space. For instance, disks of a fixed radius could have centers that shift as \(x\) changes. The individual disk areas remain unchanged, but the collection of disks need not form a solid that contains a particular fixed cylinder. The correction is to make only the conclusion justified by the information: the volume lies between the integrals of the lower and upper area bounds. A physical containment claim requires additional geometric information, such as the disks having a common axis.
Checking an Answer Systematically
After computing a volume, use several independent checks. Each catches a different kind of error.
- Check the area formula. Confirm that the cross-sectional shape and dimensions give the stated \(A(x)\).
- Check the bounds. Verify that the computed volume lies between valid lower and upper volume estimates.
- Check the units. Area has square units, and integrating with respect to a length gives cubic units.
- Check the scale. If all slices have area at most \(M\), a claimed volume larger than \(M(b-a)\) cannot be correct.
- Check the integral calculation. Differentiate the proposed antiderivative to recover the integrand, then evaluate both endpoints carefully.
A volume estimate is strongest when its assumptions are explicit. A midpoint-area calculation can be informative without being a rigorous bound. A bound becomes rigorous when the area comparison is valid throughout the interval. If the first bracket is too broad, subdividing the interval can improve it.
Check Your Understanding
Use the cross-sectional area comparisons and the distinction between volume bounds and geometric containment to answer these questions.
- If \(5\leq A(x)\leq8\) on an interval of length \(3\), what lower and upper bounds follow for the volume?
- Why do bounds on disk areas alone not establish that a fixed cylinder physically fits inside the solid?
- For square cross sections with side length \(s(x)\), what expression gives the cross-sectional area?
- Why is interval length multiplied by the midpoint area not automatically a rigorous upper or lower bound?
- What happens to area-based volume bounds when a decreasing area function is bounded separately on smaller subintervals?
