Navigation

Applications of Integration

2.2 Averages

29 problems · hints, answers and solutions shown beside each one

Recall that we are using logx\log x to denote the logarithm of xx with base ee. In other courses it is often denoted lnx\ln x.

1Stage 1Conceptual

Do you understand the idea? Usually little or no calculation.

Q1Stage 1

Below is the graph of a function y=f(x)y=f(x). Its average value on the interval [0,5][0,5] is AA. Draw a rectangle on the graph with area 05f(x)dx\int_0^5 f(x)\,\dee{x}.

Figure from prob_s2.2, line 2

Figure from prob_s2.2, line 2

Hint

See Definition 2.2.2 in the CLP-2 text and the discussion following it for the link between area under the curve and averages.

Answer

The most straightforward of many possible answers is shown.

Figure from prob_s2.2, line 2

Figure from prob_s2.2, line 2

Full solution

Since the average of f(x)f(x) on the interval [0,5][0,5] is AA, using Definition 2.2.2 in the CLP-2 text,

A=1505f(x)dx5A=05f(x)dx\begin{align*} A&=\frac{1}{5}\int_0^5 f(x)\,\dee{x}\\ 5A&=\int_0^5 f(x)\,\dee{x} \end{align*}

So, a rectangle with width 5 and height AA has area 05f(x)dx\int_0^5 f(x)\,\dee{x}.

That is: if we replace f(x)f(x) with the constant function g(x)=Ag(x)=A, then on the interval [0,5][0,5], the area under the curve is unchanged.

Figure from prob_s2.2, line 2

Figure from prob_s2.2, line 2

(There are many rectangles with area 5A5A; we drew the one we consider to be the most straightforward in this context.)

Q2Stage 1

Suppose a car travels for 5 hours in a straight line, with an average velocity of 100 kph. How far did the car travel?

Hint

Average velocity is discussed in Example 2.2.5 of the CLP-2 text. You don't need an integral for this.

Answer

500 km

Full solution

Average velocity, as discussed in Example 2.2.5 of the CLP-2 text, is change in position divided by change in time. So, the change in position (i.e. distance travelled) is (100 km/h)(5 h) = 500 km.

Q3Stage 1

A force F(x)F(x) acts on an object from position x=ax=a metres to position x=bx=b metres, for a total of WW joules of work. What was the average force on the object?

Hint

Much like Problem 2, you don't need to do any integration here.

Answer

Wba\dfrac{W}{b-a} N

Full solution

The work done is

W=abF(x)dxW=\int_a^b F(x)\,\dee{x}

so the average value of F(x)F(x) is

1baabF(x)dx=1ba(W).\frac{1}{b-a}\int_a^b F(x)\,\dee{x} =\frac{1}{b-a}(W).

We can quickly check our units: since WW is in joules (that is, newton-metres), and bab-a is in metres, so Wba\frac{W}{b-a} is in newtons.

Q4Stage 1

Suppose we want to approximate the average value of the function f(x)f(x) on the interval [a,b][a,b]. To do this, we cut the interval [a,b][a,b] into nn pieces, then take nn samples by finding the function's output at the left endpoint of each piece, starting with aa. Then, we average those nn samples. (In the example below, n=4n=4.)

Figure from prob_s2.2, line 2

Figure from prob_s2.2, line 2

  1. Using nn samples, what is the distance between two consecutive sample points xix_i and xi+1x_{i+1}?

  2. Assuming n4n \geq 4, what is the xx-coordinate of the fourth sample?

  3. Assuming n4n \geq 4, what is the yy-value of the fourth sample?

  4. Write the approximation of the average value of f(x)f(x) over the interval [a,b][a,b] using sigma notation.

Hint

Part (a) is asking the length of the pieces we've cut our interval into. Part (c) should be given in terms of ff. Our final answer in (d) will resemble a Riemann sum, but without some extra manipulation it won't be in exactly the form of a Riemann sum we're used to.

Answer

(a) ban\dfrac{b-a}{n} (b) a+3bana+3\dfrac{b-a}{n} (c) f(a+3ban)f\left(a+3\dfrac{b-a}{n} \right) (d) 1ni=1nf(a+(i1)ban)\dfrac{1}{n}\sum\limits_{i=1}^n f\left(a+(i-1)\frac{b-a}{n}\right)

Full solution
  1. The entire interval has length bab-a, and we're cutting it into nn pieces, so the length of one piece (and hence the distance between two consecutive samples) is ban\frac{b-a}{n}.

  2. The first sample, as given in the question statement, is taken at x=ax=a. The second sample, then, is at x=a+banx=a+\frac{b-a}{n}, this third is at x=1+2banx=1+2\frac{b-a}{n}, and the fourth is at a+3bana+3\frac{b-a}{n}.

  3. The yy-value of the fourth sample is simply f(a+3ban)f\left(a+3\frac{b-a}{n}\right). Note this is the number we use in our average, not the xx-value.

  4. Our samples are f(a)f(a), f(a+ban)f\left(a+\frac{b-a}{n}\right), f(a+2ban)f\left(a+2\frac{b-a}{n}\right), f(a+3ban)f\left(a+3\frac{b-a}{n}\right), etc. Since there are nn of them, we divide their sum by nn. So, the average is:

     f(a)+f(a+ban)+f(a+2ban)++f(a+(n1)ban)n=1n[f(a)+f(a+ban)f(a+2ban)++f(a+(n1)ban)]=1ni=1nf(a+(i1)ban)\begin{align*}&~\frac{f(a)+f\left(a+\frac{b-a}{n}\right)+ f\left(a+2\frac{b-a}{n}\right)+\cdots + f\left(a+(n-1)\frac{b-a}{n}\right)}{n}\\ &=\frac{1}{n}\left[f(a)+f\left(a+\frac{b-a}{n}\right)f\left(a+2\frac{b-a}{n}\right)+\cdots + f\left(a+(n-1)\frac{b-a}{n}\right)\right]\\ &= \frac{1}{n}\sum_{i=1}^nf\left(a+(i-1)\frac{b-a}{n}\right)\end{align*}

    Remark: if we multiply and divide by bab-a, we see this expression is equivalent to a left Riemann sum, divided by the length of our interval.

    =1bai=1nf(a+(i1)ban)ban=1bai=1nf(a+(i1)Δx)Δx\begin{align*}&= \frac{1}{b-a}\sum_{i=1}^nf\left(a+(i-1)\frac{b-a}{n}\right)\frac{b-a}{n}\\ &= \frac{1}{b-a}\sum_{i=1}^nf\left(a+(i-1)\De x\right)\De x\end{align*}

    As nn gets larger and larger, using the definition of a definite integral, this expression gets closer and closer to 1baabf(x)dx\frac{1}{b-a}\int_a^bf(x)\,\dee{x}. This is one way of justifying our definition of an average of a function on an interval.

Q5Stage 1

Suppose f(x)f(x) and g(x)g(x) are functions that are defined for all numbers in the interval [0,10][0,10].

  1. If f(x)g(x)f(x) \leq g(x) for all xx in [0,10][0,10], then is the average value of f(x)f(x) is less than or equal to the average value of g(x)g(x) on the interval [0,10][0,10], or is there not enough information to tell?

  2. Suppose f(x)g(x)f(x) \leq g(x) for all xx in [0.01,10][0.01,10]. Is the average value of f(x)f(x) less than or equal to the average value of g(x)g(x) over the interval [0,10][0,10], or is there not enough information to tell?

Hint

For (b), the value of f(0)f(0) could be much, much larger than g(0)g(0).

Answer

(a) yes (b) not enough information

Full solution
  1. Yes, the average of f(x)f(x) is less than or equal to the average of g(x)g(x) on [0,10][0,10]. The reason is that, if f(x)g(x)f(x) \leq g(x) for all xx in [0,10][0,10], then:

    110010f(x)dx110010g(x)dx.\frac{1}{10}\int_0^{10}f(x)\,\dee{x} \leq \frac{1}{10}\int_0^{10}g(x)\,\dee{x}.
  2. There is not enough information to tell. It's certainly possible: for instance, take f(x)=0f(x)=0 and g(x)=1g(x)=1 for all xx in [0,10][0,10]. Then f(x)g(x)f(x) \leq g(x) and the average of f(x)f(x) is 0, which is less than 1, the average of g(x)g(x).

    However, consider f(x)={100 if 0x0.010 else f(x) = \begin{cases} 100 & \text{ if } 0 \le x \le 0.01\\ 0 & \text{ else } \end{cases} and g(x)=0g(x) = 0. Then f(x)g(x)f(x) \le g(x) for all xx in [0.01,10][0.01,10], but the average of f(x)f(x) is 0.10.1, while the average of g(x)g(x) is 0.

Q6Stage 1

Suppose ff is an odd function, defined for all real numbers. What is the average of ff on the interval [10,10][-10,10]?

Hint

The answer is something very simple.

Answer

0

Full solution

Recall the definition of an odd function: f(x)=f(x)f(-x)=-f(x). Since the domain of integration is symmetric, the signed area on one side of the yy-axis “cancels out" the signed area on the other–this is Theorem 1.2.11 in the CLP-2 text.

1201010f(x)dx=120(0)=0\begin{align*}\frac{1}{20}\int_{-10}^{10}f(x)\,\dee{x}&=\frac{1}{20}(0)=0 \end{align*}

2Stage 2Procedural

Practising the skill itself, until applying it is automatic.

Q7Stage 2Past exam · 2016Q3

Find the average value of f(x)=sin(5x)+1f(x) = \sin(5x)+1 over the interval π/2xπ/2-\pi/2 \le x \le \pi/2.

Hint

Apply the definition of “average value” in Section 2.2 of the

CLP-2 text.

Answer

11

Full solution

By definition, the average value is

1ππ/2π/2(sin(5x)+1)dx\begin{align*} \frac{1}{\pi}\int_{-\pi/2}^{\pi/2} \big( \sin(5x)+1 \big) \,\dee{x} \end{align*}

We now observe that sin(5x)\sin(5x) is an odd function, and hence its integral over the symmetric interval [π2,π2][-\frac\pi2,\frac\pi2] equals zero. So the average value of f(x)f(x) on this interval is 11:

1ππ/2π/2(sin(5x)+1)dx=1ππ/2π/2sin(5x)dx+1ππ/2π/21dx=1ππ/2π/21dx=1\begin{align*} \frac{1}{\pi}\int_{-\pi/2}^{\pi/2} \big( \sin(5x)+1 \big) \,\dee{x}&= \frac{1}{\pi}\int_{-\pi/2}^{\pi/2} \sin(5x) \,\dee{x}+ \frac{1}{\pi}\int_{-\pi/2}^{\pi/2}1 \,\dee{x}\\ &= \frac{1}{\pi}\int_{-\pi/2}^{\pi/2}1 \,\dee{x}=1 \end{align*}

Alternatively, using the fundamental theorem of calculus, the average equals:

1π[cos(5x)5+x]π/2π/2=1π{[cos(5π/2)5+π2][cos(5π/2)5+π2]}=ππ=1\begin{align*} \frac{1}{\pi}\bigg[\frac{-\cos (5x) }{5} +x \bigg]_{-\pi/2}^{\pi/2} = \frac{1}{\pi}\bigg\{\bigg[\frac{-\cos (5\pi/2) }{5} +\frac{\pi}{2}\bigg] -\bigg[ \frac{-\cos (-5\pi/2) }{5} + \frac{-\pi}{2} \bigg]\bigg\} = \frac{\pi }{\pi}=1 \end{align*}
Q8Stage 2Past exam · 2015A

Find the average value of the function y=x2logxy= x^2\log x on the interval 1xe1 \le x\le e.

Hint

You can antidifferentiate x2logxx^2\log x using integration by parts.

Answer

1e1[29e3+19]\displaystyle\frac{1}{e-1}\Big[\frac{2}{9}e^3+\frac{1}{9}\Big]

Full solution

By definition, the average is

1e11ex2logx dx\begin{align*}\frac{1}{e-1}\int_1^e x^2\log x\ \dee{x} &\end{align*}

To antidifferentiate, we use integration by parts with u=logxu=\log x and dv=x2dx\dee{v}=x^2\,\dee{x}, hence du=1xdx\dee{u}=\frac{1}{x}\,\dee{x} and v=13x3v=\frac{1}{3}x^3.

1e11ex2logx dx=1e1([13x3logx]1e1e13x2dx)=1e1[x33logxx39]x=1x=e=1e1[e33e39+19]=1e1[29e3+19]\begin{align*}\frac{1}{e-1}\int_1^e x^2\log x\ \dee{x}&= \frac{1}{e-1}\left(\left[\frac{1}{3}x^3\log x\right]_1^e - \int_1^e \frac{1}{3}x^2\,\dee{x} \right) \\&= \frac{1}{e-1}\bigg[\frac{x^3}{3}\log x-\frac{x^3}{9}\bigg]_{x=1}^{x=e}\\ &=\frac{1}{e-1}\bigg[\frac{e^3}{3}-\frac{e^3}{9}+\frac{1}{9}\bigg]\\ &=\frac{1}{e-1}\bigg[\frac{2}{9}e^3+\frac{1}{9}\bigg]\end{align*}
Q9Stage 2Past exam · 2016A

Find the average value of the function f(x)=3cos3x+2cos2xf(x) = 3\cos^3x + 2\cos^2x on the interval 0xπ20\le x\le\frac\pi2.

Hint

You can antidifferentiate an odd power of cosine with a substitution; for an even power of cosine, use the identity cos2x=12(1+cos(2x))\cos^2 x = \frac{1}{2}\big(1+\cos(2x)\big).

Answer

4π+1\dfrac{4}{\pi}+1

Full solution

By definition, the average value in question equals

1π/200π/2(3cos3x+2cos2x)dx=2π(0π/23cos3xdx+0π/22cos2xdx)\begin{equation*} \frac1{\pi/2-0} \int_0^{\pi/2} ( 3\cos^3x + 2\cos^2x ) \,\dee{x} = \frac2\pi \bigg( \int_0^{\pi/2} 3\cos^3x \,\dee{x} + \int_0^{\pi/2} 2\cos^2x \,\dee{x} \bigg) \end{equation*}

For the first integral we use the substitution u=sinxu = \sin x, du=cosxdx\dee{u}=\cos x\,\dee{x}, cos2x=1sin2x=1u2\cos^2 x=1-\sin^2 x = 1-u^2. Note that the endpoints x=0x=0 and x=π2x=\frac\pi2 become u=0u=0 and u=1u=1, respectively.

0π/23cos3xdx=0π/23cos2xcosxdx=013(1u2)du=(3uu3)01=2.\begin{align*} \int_0^{\pi/2} 3 \cos^3x \,\dee{x} &= \int_0^{\pi/2} 3 \cos^2 x \cos x \,\dee{x} \\ &= \int_0^{1} 3(1-u^2) \,\dee{u} \\ &= (3u-u^3)\Big|_0^1 = 2. \end{align*}

For the second integral we use the trigonometric identity cos2xdx=1+cos(2x)2\cos^2x\,\dee{x} = \frac{1 + \cos(2x)}{2}.

20π/2cos2xdx=0π/2(1+cos(2x))dx=[x+12sin(2x)]0π/2=π2\begin{align*} 2 \int_0^{\pi/2} \cos^2x\,\dee{x} &= \int_0^{\pi/2} \big(1 + \cos(2x) \big) \,\dee{x} \\ &= \bigg[ x + \frac{1}{2}\sin(2x) \bigg]_0^{\pi/2} = \frac\pi2 \end{align*}

Therefore, the average value in question is

2π(0π/23cos3xdx+0π/22cos2xdx)=2π(2+π2)=4π+1.\begin{equation*} \frac2\pi \bigg( \int_0^{\pi/2} 3\cos^3x \,\dee{x} + \int_0^{\pi/2} 2\cos^2x \,\dee{x} \bigg) = \frac2\pi \bigg( 2 + \frac\pi2 \bigg) = \frac{4}{\pi}+1. \end{equation*}
Q10Stage 2Past exam · 2012A

Let kk be a positive constant. Find the average value of the function f(x)=sin(kx)f(x) = \sin(kx) on the interval 0xπ/k0\le x\le \pi/k.

Hint

If you're not sure how to antidifferentiate, try the substitution u=kxu=kx, du=kdx\dee{u}=k\dee{x}, keeping in mind that kk is a constant. Interestingly, your final answer won't depend on kk.

Answer

2π\dfrac{2}{\pi}

Full solution

By definition, the average value in question equals

Ave=1π/k00π/ksin(kx)dx\begin{equation*} \text{Ave}=\frac1{\pi/k-0} \int_0^{\pi/k} \sin(kx) \,\dee{x} \end{equation*}

To evaluate the integral, we use the substitution u=kxu = kx, du=kdx\dee{u}=k\,\dee{x}. Note that the endpoints x=0x=0 and x=π/kx=\pi/k become u=0u=0 and u=πu=\pi, respectively. So

Ave=kπ0πsin(u)duk=1π[cos(u)]0π=2π\begin{align*} \text{Ave}=\frac{k}{\pi} \int_0^\pi \sin(u) \,\frac{\dee{u}}{k} =\frac{1}{\pi}\Big[-\cos(u)\Big]_0^\pi =\frac{2}{\pi} \end{align*}

Remark: the average does not depend on kk. To see why this is, note that sin(kx)\sin(kx) runs between 1-1 and 11 as xx changes. When x=0x=0, kx=0kx=0, and when x=π/kx=\pi/k, kx=πkx=\pi. So, our function sin(kx)\sin (kx) runs exactly from sin0=0\sin 0 =0 to sin(π/2)=1\sin(\pi/2)=1, then back down to sinπ=0\sin\pi=0.

Figure from prob_s2.2, line 2

Figure from prob_s2.2, line 2

Q11Stage 2Past exam · 1997A

The temperature in Celsius in a 3 m long rod at a point xx metres from the left end of the rod is given by the function T(x)=8016x2T(x)=\frac{80}{16-x^2}. Determine the average temperature in the rod.

Hint

The method of partial fractions can help you antidifferentiate.

Answer

103log7\dfrac{10}{3}\log 7 degrees Celsius

Full solution

By definition, the average temperature is

1303T(x)dx=13038016x2dx\frac{1}{3}\int_0^3 T(x)\,\dee{x} =\frac{1}{3}\int_0^3 \frac{80}{16-x^2}\,\dee{x}

We don't see an obvious substitution, but integrand is a rational function. The degree of the numerator is strictly less than the degree of the denominator, so we factor the denominator and use a partial fraction decomposition.

8016x2=80(4x)(4+x)=A4x+B4+x80=A(4+x)+B(4x)\begin{align*}\frac{80}{16-x^2}&=\frac{80}{(4-x)(4+x)} = \frac{A}{4-x}+\frac{B}{4+x}\\ 80&=A(4+x)+B(4-x)\end{align*}

Setting x=4x=4, we see 80=8A80=8A, so A=10A=10. Setting x=4x=-4, we see 80=8B80=8B, so B=10B=10.

13038016x2dx=130380(4x)(4+x)dx=1303[104x+104+x]dx=1303[10x4+104+x]dx=103[logx4+logx+4]03=103logx+4x403=103[log7log1]=103log7 degrees Celsius\begin{align*}\frac{1}{3}\int_0^3 \frac{80}{16-x^2}\,\dee{x}&=\frac{1}{3}\int_0^3 \frac{80}{(4-x)(4+x)}\,\dee{x}=\frac{1}{3}\int_0^3 \Big[\frac{\textcolor{red}{10}}{4-x}+\frac{\textcolor{red}{10}}{4+x}\Big]\,\dee{x}\\ &=\frac{1}{3}\int_0^3 \Big[-\frac{10}{x-4}+\frac{10}{4+x}\Big]\,\dee{x}\\ &=\frac{10}{3}\Big[-\log|x-4|+\log|x+4|\Big]_0^3 \\ &=\frac{10}{3}\log\Big|\frac{x+4}{x-4}\Big|\bigg|_0^3 =\frac{10}{3}[\log 7-\log 1] \\ &=\frac{10}{3}\log 7\quad\text{ degrees Celsius}\end{align*}
Q12Stage 2Past exam · 1997D

What is the average value of the function f(x)=logxxf(x)=\dfrac{\log x}{x} on the interval [1,e][1,e]?

Hint

Try the substitution u=logxu=\log x, du=1xdx\dee{u}=\frac{1}{x}\,\dee{x}.

Answer

12(e1)\dfrac{1}{2(e-1)}

Full solution

By definition, the average value is

1e11elogxx dx\begin{align*}\frac{1}{e-1}\int_1^e\frac{\log x}{x}\ \dee{x}&\end{align*}

To integrate, we use the substitution u=logxu=\log x, du=1xdx\dee{u}=\frac{1}{x}\,\dee{x}. Then the limits of integration become 0 and 1, respectively.

1e11elogxx dx=1e101u du=1e1[u22]01=12(e1)\begin{align*}\frac{1}{e-1}\int_1^e\frac{\log x}{x}\ \dee{x}& =\frac{1}{e-1}\int_0^1u\ \dee{u} =\frac{1}{e-1}\bigg[\frac{u^2}{2}\bigg]_0^1 =\frac{1}{2(e-1)}\end{align*}
Q13Stage 2Past exam · 2000D

Find the average value of f(x)=cos2(x)f(x)=\cos^2(x) over 0x2π0\le x\le 2\pi.

Hint

Remember cos2x=12(1+cos(2x))\cos^2 x = \frac{1}{2}\big(1+\cos(2x)\big).

Answer

12\dfrac12

Full solution

By definition, the average value is:

12π02πcos2xdx=12π1202π(cos(2x)+1)dx=14π[sin(2x)2+x]02π=14π2π=12\begin{align*} \frac{1}{2\pi}\int_0^{2\pi}\cos^2x\,\dee{x} =\frac{1}{2\pi}\cdot\frac{1}{2}\int_0^{2\pi}\big(\cos(2x)+1\big)\,\dee{x} =\frac{1}{4\pi}\Big[\frac{\sin(2x)}{2}+x\Big]_0^{2\pi} =\frac{1}{4\pi}\cdot 2\pi =\frac{1}{2} \end{align*}
Q14Stage 2

The carbon dioxide concentration in the air at a particular location over one year is approximated by C(t)=400+50cos(t12π)+200cos(t4380π)C(t) = 400+50\cos\left(\frac{t}{12}\pi\right)+200\cos\left(\frac{t}{4380}\pi\right) parts per million, where tt is measured in hours.

  1. What is the average carbon dioxide concentration for that location for that year?

  2. What is the average over the first day?

  3. Suppose measurements were only made at noon every day: that is, when t=12+24nt=12+24n, where nn is any whole number between 0 and 364. Then the daily variation would cease: 50cos((12+24n)12π)=50cos(π+2πn)=50cosπ=5050\cos\left(\frac{(12+24n)}{12}\pi\right) = 50\cos\left(\pi+2\pi n\right) = 50\cos\pi=-50. So, the approximation for the concentration of carbon dioxide in the atmosphere might be given as

    N(t)=350+200cos(t4380π) ppmN(t) = 350 +200\cos\left(\frac{t}{4380}\pi\right)\quad\text{ ppm}

    What is the relative error in the yearly average concentration of carbon dioxide involved in using N(t)N(t), instead of C(t)C(t)?

You may assume a day has exactly 24 hours, and a year has exactly 8760 hours.

Hint

Notice the term 50cos(t12π)50\cos\left(\frac{t}{12}\pi\right) has a period of 24 hours, while the term 200cos(t4380π)200\cos\left(\frac{t}{4380}\pi\right) has a period of one year.

If nn is an approximation of cc, then the relative error of nn is ncc\frac{|n-c|}{c}.

Answer

(a) 400 ppm (b) 599.99\approx 599.99 ppm (c) 0.125, or 12.5%

Full solution

Before we start answering questions, let's look at our function a little more carefully. The term 50cos(t12π)50\cos\left(\frac{t}{12}\pi\right) has a period of 24 hours, while the term 200cos(t4380π)200\cos\left(\frac{t}{4380}\pi\right) has a period of one year. So, the former term describes a standard daily variation, while the latter gives a seasonal variation over the year.

(a) Using the definition of an average, the average concentration over one year (t=0t=0 to 87608760) is:

1876008760(400+50cos(t12π)+200cos(t4380π))dt=1876008760400dt+50876008760cos(t12π)dt+200876008760cos(t4380π)dt=400+5876[12πsin(t12π)]08760+5219[4380πsin(t4380π)]08760\begin{align*}&\frac{1}{8760}\int_0^{8760}\left(400+\textcolor{red}{50\cos\left(\frac{t}{12}\pi\right)}+\textcolor{blue}{200\cos\left(\frac{t}{4380}\pi\right)}\right)\,\dee{t}\\ &=\frac{1}{8760}\int_0^{8760}400\,\dee{t} + \textcolor{red}{\frac{50}{8760}\int_0^{8760}\cos\left(\frac{t}{12}\pi\right)\,\dee{t}} + \textcolor{blue}{\frac{200}{8760}\int_0^{8760}\cos\left(\frac{t}{4380}\pi\right)\,\dee{t}}\\ &= 400 +\textcolor{red}{ \frac{5}{876}\left[\frac{12}{\pi}\sin\left(\frac{t}{12}\pi\right)\right]_0^{8760}}+ \textcolor{blue}{\frac{5}{219}\left[\frac{4380}{\pi}\sin\left(\frac{t}{4380}\pi\right)\right]_0^{8760}}\end{align*}

Since 876012=730\frac{8760}{12} = 730, which is even, sin(876012π)=sin(0)=0\sin\left(\frac{8760}{12}\pi\right)=\sin(0)=0. Also, sin(87604380π)=sin(2π)=0\sin\left(\frac{8760}{4380}\pi\right)=\sin(2\pi)=0.

=400+5876(0)+5219(0)=400 ppm\begin{align*}&=400+\textcolor{red}{\frac{5}{876}(0}) + \textcolor{blue}{\frac{5}{219}(0)}\\ &=400\quad\text{ ppm}\end{align*}

Remark: for the portions of the integral in red and blue, we also could have noticed that the integrand goes through a whole (integer) number of periods. For every period, the net signed area between the curve and the xx-axis is zero, so we could have seen from the very beginning these terms would contribute 0 to the final average.

(b) Using the definition of an average, the average concentration over the first day (t=0t=0 to t=24t=24) is:

124024(400+50cos(t12π)+200cos(t4380π))dt=124024400dt+5024024cos(t12π)dt+20024024cos(t4380π)dt\begin{align*}&\frac{1}{24}\int_0^{24} \left(400+\textcolor{red}{50\cos\left(\frac{t}{12}\pi\right)}+\textcolor{blue}{200\cos\left(\frac{t}{4380}\pi\right)}\right)\,\dee{t}\\ &=\frac{1}{24}\int_0^{24}400\,\dee{t} + \textcolor{red}{\frac{50}{24}\int_0^{24}\cos\left(\frac{t}{12}\pi\right)\,\dee{t}}+ \textcolor{blue}{\frac{200}{24}\int_0^{24}\cos\left(\frac{t}{4380}\pi\right)\,\dee{t}}\end{align*}

Note t=0t=0 to t=24t=24 is one complete period for the integrand in red, so the red integral will evaluate to zero. However, t=0t=0 to t=24t=24 is less than one cycle for the integrand in blue, so we expect this will contribute some non-zero quantity to the average.

=400+0+20024[4380πsin(t4380π)]024=400+2534380πsin(244380π)=400+2534380πsin(2365π)400+199.99=599.99 ppm\begin{align*}&=400+\textcolor{red}{0} + \textcolor{blue}{\frac{200}{24}\left[\frac{4380}{\pi}\sin\left(\frac{t}{4380}\pi\right)\right]_0^{24}}\\ &=400+\color{blue}\frac{25}{3}\cdot\frac{4380}{\pi}\sin\left(\frac{24}{4380}\pi\right)\\ &=400+\color{blue}\frac{25}{3}\cdot\frac{4380}{\pi}\sin\left(\frac{2}{365}\pi\right)\\ &\approx 400+\color{blue}199.99\\ &=599.99 \quad\text{ ppm}\end{align*}

Remark: C(0)=400+50+200C(0) = 400+\textcolor{red}{50}+\textcolor{blue}{200}. The red term comes from the daily variation, and over the first day this will have an average of 0. The blue term comes from the seasonal variation, which changes dramatically over the course of an entire year but won't change very much over the course of a single day. So, it is reasonable that the average concentration over the first day should be close to (but not exactly) 400+200400+200 ppm.

(c) The average of N(t)N(t) over [0,8760][0,8760] is:

1876008760(350+200cos(t4380π))dt=350+2008760[4380πsin(t4380π)]08760=350+2008760[4380πsin(87604380π)]=350+100πsin(2π)=350\begin{align*} \frac{1}{8760}\int_0^{8760} \left ( 350+200\cos\left(\frac{t}{4380}\pi\right)\right)\,\dee{t}&=350 + \frac{200}{8760}\bigg[\frac{4380}{\pi}\sin\left(\frac{t}{4380}\pi\right)\bigg]_0^{8760} \\&=350+\frac{200}{8760}\left[\frac{4380}{\pi}\sin\left(\frac{8760}{4380}\pi\right)\right]\\ &=350+\frac{100}{\pi}\sin\left(2\pi\right)\\&=350 \end{align*}

Since the average of C(t)C(t) was 400, this gives us an absolute error of 400350=50|400-350|=50 ppm, for a relative error of

50400=0.125,\frac{50}{400} = 0.125,

or 12.5%.

That is: sampling at the same time every day, rather than throughout the day, lead to an error of 12.5% in the yearly average concentration of carbon dioxide.

Q15Stage 2

Let SS be the solid formed by rotating the parabola y=x2y=x^2 from x=0x=0 to x=2x=2 about the xx-axis.

  1. What is the average area of the circular cross-sections of SS? Call this value AA.

  2. What is the volume of SS?

  3. What is the volume of a cylinder with circular cross-sectional area AA and length 2?

Hint

A cross section of SS at location xx is a circle with radius x2x^2, so area πx4\pi x^4. Part (a) is asking for the average of this function on [0,2][0,2].

Answer

(a) 16π5\dfrac{16\pi}{5} (b) 32π5\dfrac{32\pi}{5} (c) 32π5\dfrac{32\pi}{5}

Full solution
  1. The cross-section of SS at xx is a circle with radius x2x^2, so area πx4\pi x^4. The average of these values, 0x20 \leq x \leq 2, is

    A=12002πx4dx=12[π5x5]02=16π5A = \frac{1}{2-0}\int_0^2 \pi x^4\,\dee{x} = \frac{1}{2}\left[\frac{\pi}{5}x^5\right]_0^2=\frac{16\pi}{5}
  2. To find the volume of SS, imagine cutting it into thin circular disks of radius x2x^2 and thickness dx\dee{x}. The volume of one such disk is πx4dx\pi x^4\,\dee{x}, so the volume of SS is

    02πx4dx=[π5x5]02=32π5\int_0^2 \pi x^4\,\dee{x} = \left[\frac{\pi}{5}x^5\right]_0^2 = \frac{32\pi}{5}
  3. The volume of a cylinder is the product of its base area with its length. A cylinder with circular cross-sections of area 16π5\frac{16\pi}{5} and length 2 has volume 32π5\frac{32\pi}{5}.

    Remark: this is the same as the volume of SS, so the average cross-sectional area of SS tells us the cross-sectional area of a cylinder with the same length and volume as SS. Compare this to Question 1, where we saw the average value of a function gave the height of a rectangle with the same area as the function over the given interval.

For Questions 16 through 18, let the root mean square of f(x)f(x) on [a,b][a,b] be 1baabf2(x)dx\displaystyle\sqrt{\frac{1}{b-a}\int_a^b f^2(x)\,\dee{x}}. This is the formula used in Example 2.2.6 in the CLP-2 text.

Q16Stage 2

Let f(x)=xf(x) = x.

  1. Calculate the average of f(x)f(x) over [3,3][-3,3].

  2. Calculate the root mean square of f(x)f(x) over [3,3][-3,3].

Hint

(a) can be done without calculation

Answer

(a) 0 (b) 3\sqrt3

Full solution
  1. We can see without calculation that the average will be zero, since f(x)=xf(x)=x is an odd function and [3,3][-3,3] is a symmetric interval. Alternately, we can use the definition of an average to calculate

    1633xdx=[112x2]33=112(99)=0\frac{1}{6}\int_{-3}^3 x\,\dee{x} = \left[\frac{1}{12}x^2\right]_{-3}^3 = \frac{1}{12}(9-9)=0
  2. Using the definition provided for root mean square:

    RMS=1633x2dx=[118x3]33=27182718=3\begin{align*} \text{RMS}&=\sqrt{\frac{1}{6}\int_{-3}^3 x^2\,\dee{x}} = \sqrt{\left[\frac{1}{18}x^3\right]_{-3}^3} = \sqrt{\frac{27}{18}-\frac{-27}{18}} = \sqrt{3} \end{align*}
Q17Stage 2

Calculate the root mean square of f(x)=tanxf(x) = \tan x over [π4,π4]\left[-\frac{\pi}{4},\frac{\pi}{4}\right].

Hint

tan2x=sec2x1\tan^2 x = \sec^2 x - 1

Answer

4π10.52\displaystyle\sqrt{\frac{4}{\pi} - 1}\approx 0.52

Full solution

Using the definition provided,

RMS=2ππ/4π/4tan2xdx=2ππ/4π/4(sec2x1)dx=2π[tanxx]π/4π/4=2π[(1π4)(1+π4)]=2π(2π2)=4π10.52\begin{align*} \text{RMS}&=\sqrt{\frac{2}{\pi}\int_{-\pi/4}^{\pi/4}\tan^2x\,\dee{x}} = \sqrt{\frac{2}{\pi}\int_{-\pi/4}^{\pi/4}\left(\sec^2x-1\right)\,\dee{x}}\\ &=\sqrt{\frac{2}{\pi}\Big[\tan x - x \Big]_{-\pi/4}^{\pi/4}} = \sqrt{\frac{2}{\pi}\left[\left(1 - \frac{\pi}{4} \right)-\left(-1 + \frac{\pi}{4} \right)\right]}\\ &=\sqrt{\frac{2}{\pi}\left(2-\frac{\pi}{2}\right)} = \sqrt{\frac{4}{\pi} - 1} \approx 0.52 \end{align*}
Q18Stage 2

A force acts on a spring, and the spring stretches and contracts. The distance beyond its natural length at time tt is f(t)=sin(tπ)f(t) = \sin\left(t\pi\right) cm, where tt is measured in seconds. The spring constant is 3 N/cm.

  1. What is the force exerted by the spring at time tt, if it obeys Hooke's law?

  2. Find the average of the force exerted by the spring from t=0t=0 to t=6t=6.

  3. Find the root mean square of the force exerted by the spring from t=0t=0 to t=6t=6.

Hint

Remember force is the product of the spring constant with the distance it's stretched past its natural length. The units given in the question are not exactly standard, but they are compatible with each other.

You can find part (b) without any calculation. For (c), remember sin2x=12(1cos(2x))\sin^2 x = \frac{1}{2}\big(1-\cos(2x)\big).

Answer

(a) F(t)=3f(t)=3sin(tπ)F(t) = 3f(t) = 3\sin\left(t\pi\right) N (b) 0 (c) 322.12\dfrac{3}{\sqrt{2}} \approx 2.12

Full solution
  1. Using Hooke's law, when the spring is stretched (or compressed) f(t)f(t) metres past its natural length, the force exerted is kf(t)kf(t), where kk is the spring constant. In this case, the force is

    F(x)=(3 N/cm)(f(t) cm)=3sin(tπ) NF(x) = (3\text{ N/cm})(f(t)\text{ cm}) = 3 \sin\left(t\pi\right)\text{ N}
  2. Our interval encompasses three full periods of sine, so the average will be zero.

    Alternately, we can compute, using the definition of an average:

    Avg=16063sin(tπ)dt=16[3πcos(tπ)]06=12π[cos0cos(6π)]=0\begin{align*} \text{Avg}&=\frac{1}{6}\int_0^6 3\sin(t\pi)\,\dee{t}= \frac{1}{6}\left[-\frac{3}{\pi}\cos(t\pi)\right]_0^6=\frac{1}{2\pi}\left[\cos 0 - \cos (6\pi)\right]=0 \end{align*}

    This it doesn't tell us very much about the “normal" amount of force from the spring during our time period. It only tells us that force in one direction at is “cancelled out" by force in the opposite direction at another time.

  3. Using the definition given for root mean square,

    RMS=1606(3sin(tπ))2dt=3206sin2(tπ)dt=3406(1cos(2tπ))dt=34[t12πsin(2tπ)]06=34[612πsin(12π)0]=34(6)=322.12\begin{align*} \text{RMS}&=\sqrt{\frac{1}{6}\int_0^6 \left(3\sin(t\pi)\right)^2\,\dee{t}} = \sqrt{\frac{3}{2}\int_0^6 \sin^2(t\pi)\,\dee{t}}\\& = \sqrt{\frac{3}{4}\int_0^6 \left(1-\cos(2t\pi)\right)\,\dee{t}}\\ &=\sqrt{\frac{3}{4}\left[t - \frac{1}{2\pi}\sin(2t\pi)\right]_0^6}\\ &=\sqrt{\frac{3}{4}\left[6 - \frac{1}{2\pi}\sin(12\pi)-0\right]} \\&=\sqrt{\frac{3}{4}(6)}=\frac{3}{\sqrt2}\approx 2.12 \end{align*}

3Stage 3Application

Further than practice: several ideas at once, or an unfamiliar situation.

Q19Stage 3Past exam · 1997A

A car travels two hours without stopping. The driver records the car's speed every 20 minutes, as indicated in the table below:

time in hours01/32/314/35/32
speed in km/hr50708055608040
  1. Use the trapezoidal rule to estimate the total distance traveled in the two hours.

  2. Use the answer to part (a) to estimate the average speed of the car during this period.

Hint

The trapezoidal rule is found in Section 1.11.2 of the

CLP-2 text.

Answer

(a) 130 km130\text{ km} (b) 65 km/hr65\text{ km/hr}

Full solution

(a) Let v(t)v(t) be the speed of the car at time tt. Then, by the trapezoidal rule with a=0a=0, b=2b=2, Δt=1/3\De t=1/3, the distance traveled is

02v(t)dtΔt[12v(0)+v(1/3)+v(2/3)+v(3/3)+v(4/3)+v(5/3)+12v(2)]=13[1250+70+80+55+60+80+1240]=130 km\begin{align*} \int_0^2 v(t)\,\dee{t} &\approx\De t\Big[\half v(0)+v(1/3)+v(2/3)+v(3/3) +v(4/3)+v(5/3)+\half v(2)\Big]\\ &=\frac{1}{3}\Big[\half 50+70+80+55+60+80+\half 40\Big] =130\text{ km} \end{align*}

(b) The average speed is disttime130 km2 hr=65 km/hr\frac{\text{dist}}{\text{time}} \approx \frac{130\text{ km}}{2\text{ hr}} = 65\text{ km/hr}.

Q20Stage 3

Let s(t)=ets(t) = e^t.

  1. Find the average of s(t)s(t) on the interval [0,1][0,1]. Call this quantity AA.

  2. For any point tt, the difference between s(t)s(t) and AA is s(t)As(t)-A. Find the average value of s(t)As(t)-A on the interval [0,1][0,1].

  3. For any point tt, the absolute difference between s(t)s(t) and AA is s(t)A|s(t)-A|. Find the average value of s(t)A|s(t)-A| on the interval [0,1][0,1].

Hint

To find a definite integral of the absolute value of a function, break up the interval of integration into regions where the function is positive, and intervals where it's negative.

Answer

(a) A=e1A=e-1 (b) 0 (c) 42e+2(e1)log(e1)0.42\displaystyle4-2e+2(e-1)\log(e-1) \approx 0.42

Full solution
  1. Using the definition of an average,

    A=11001etdt=e1A=\frac{1}{1-0}\int_0^1 e^t\,\dee{t} = e-1
  2. Since s(t)A=ete+1s(t)-A = e^t-e+1, its average on [0,1][0,1] is

    11001(ete+1)dt=[etet+t]01=(ee+1)(1)=0\frac{1}{1-0}\int_0^1 \left(e^t-e+1\right)\,\dee{t} = \left[e^t -et+t\right]_0^1 = (e-e+1)-(1)=0

    Remark: what's happening here is that the average difference between s(t)s(t) and AA is zero, because the values of s(t)s(t) that are larger than AA (and give a positive value of s(t)As(t)-A) exactly cancel out the values of s(t)s(t) that are smaller than AA (and give a negative value of s(t)As(t)-A). However, knowing how far the average value is from our calculated average is a reasonable thing to measure. That's where (c) comes in.

  3. Using the definition of an average, the quantity we want is:

    11001ete+1dt\frac{1}{1-0}\int_0^1\left| e^t-e+1\right|\,\dee{t}

    To deal with the absolute value, we consider the integral over two intervals: one where ete+1e^t-e+1 is positive, and one where it's negative. To decide where to break the limits of integration, notice ete+1>0e^t-e+1>0 exactly when et>e1e^t>e-1, so t>log(e1)t>\log(e-1).

    11001ete+1dt=0log(e1)ete+1negativedt+log(e1)1ete+1positivedt=0log(e1)(et+e1)dt+log(e1)1(ete+1)dt=[et+(e1)t]0log(e1)+[et(e1)t]log(e1)1=[(e1)+(e1)log(e1)+1]+[e(e1)(e1)+(e1)log(e1)]=42e+2(e1)log(e1)0.42\begin{align*} &\frac{1}{1-0}\int_0^1\left| e^t-e+1\right|\,\dee{t}=\int_0^{\log(e-1)}|\underbrace{e^t-e+1}_{\text{negative}}|\,\dee{t}+\int_{\log(e-1)}^1|\underbrace{e^t-e+1}_{\text{positive}}|\,\dee{t}\\ &=\int_0^{\log(e-1)}\big(-e^t+e-1\big)\,\dee{t}+\int_{\log(e-1)}^1\big(e^t-e+1\big)\,\dee{t}\\ &=\Big[-e^t+(e-1)t\Big]_0^{\log(e-1)}+ \Big[e^t-(e-1)t\Big]_{\log(e-1)}^1\\ &=\left[-(e-1)+(e-1)\log(e-1)+1\right]+ \left[e-(e-1)-(e-1)+(e-1)\log(e-1)\right] \\&=4-2e+2(e-1)\log(e-1)\\ &\approx 0.42 \end{align*}

    Remark: what we just measured is how far s(t)s(t) is, on average, from AA. We had to neglect whether s(t)s(t) was above or below AA, because (as we saw in (b)) the values above AA “cancel out" the values below AA. That's where the absolute value came in.

    Knowing how well most of your function's values match the average is an important measure, but dealing with absolute values can be a little clumsy. Therefore, the variance of a function squares the differences, rather than taking their absolute value. (In our example, that means looking at (s(t)A)2(s(t)-A)^2, rather than s(t)A|s(t)-A|.) To compensate for the change in magnitude involve in squaring, the standard deviation is the square root of the variance. These are two very commonly used measures of how similar a function is to its average. Compare standard deviation to root-square-mean voltage from Example 2.2.6 in the CLP-2 text and Questions 16 to 18.

Q21Stage 3

Consider the two functions f(x)f(x) and g(x)g(x) below, both of which have average AA on [0,4][0,4].

Figure from prob_s2.2, line 2

Figure from prob_s2.2, line 2

  1. Which function has a larger average on [0,4][0,4]: f(x)Af(x)-A or g(x)Ag(x)-A?

  2. Which function has a larger average on [0,4][0,4]: f(x)A|f(x)-A| or g(x)A|g(x)-A|?

Hint

This is an application of the ideas in Question 20.

Answer

(a) neither–both are zero (b) f(x)A|f(x)-A| has the larger average on [0,4][0,4]

Full solution
  1. Neither: the average of both these functions is zero. We saw this with a particular function in Question 20 (b), but it's actually true in general. It's a quick calculation to prove.

    The average of f(x)Af(x)-A is:

    14004(f(x)A)dx=1404f(x)dxAA=AA=0\begin{align*} \frac{1}{4-0}\int_0^4 \big(f(x)-A\big)\,\dee{x}&=\underbrace{\frac{1}{4}\int_0^4f(x)\,\dee{x}}_{A} - A=A-A=0 \end{align*}

    Similarly, the average of g(x)Ag(x)-A is:

    14004(g(x)A)dx=1404g(x)dxAA=AA=0\begin{align*} \frac{1}{4-0}\int_0^4 \big(g(x)-A\big)\,\dee{x}&=\underbrace{\frac{1}{4}\int_0^4g(x)\,\dee{x}}_{A} - A=A-A=0 \end{align*}
  2. The function f(x)A|f(x)-A| tells us how far f(x)f(x) is from AA, without worrying whether f(x)f(x) is larger or smaller. Looking at our graph, for most values of xx in [0,4][0,4], f(x)f(x) is quite far away from AA, so f(x)A|f(x)-A| is usually a large, positive quantity.

    By contrast, g(x)A|g(x)-A| is a small positive quantity for most values of xx. The function g(x)g(x) is quite close to AA for all values of xx in [0,4][0,4].

    So, since g(x)A|g(x)-A| generally has much smaller values than f(x)A|f(x)-A|, the average of f(x)A|f(x)-A| on [0,4][0,4] will be larger than the average of g(x)A|g(x)-A| on [0,4][0,4].

    As discussed in Question 20(c), the average of f(x)A|f(x)-A| is a measure of how closely f(x)f(x) resembles its average. We see from the graph that f(x)f(x) doesn't resemble the constant function y=Ay=A much at all, while g(x)g(x) seems much more similar to the constant function y=Ay=A.

    This kind of measure–how similar a function is to its average–is also the idea behind the root square mean.

Q22Stage 3

Suppose the root mean square of a function f(x)f(x) on the interval [a,b][a,b] is RR. What is the volume of the solid formed by rotating the portion of f(x)f(x) from aa to bb about the xx-axis?

Figure from prob_s2.2, line 2

Figure from prob_s2.2, line 2

As in Example 2.2.6 of the CLP-2 text, let the root mean square of f(x)f(x) on [a,b][a,b] be 1baabf2(x)dx\displaystyle\sqrt{\frac{1}{b-a}\int_a^b f^2(x)\,\dee{x}}.

Hint

Slice the solid into circular disks of radius f(x)|f(x)| and thickness dx\dee{x}.

Answer

(ba)πR2(b-a)\pi R^2

Full solution

When we rotate f(x)f(x) about the xx-axis, we form a solid whose radius at xx is f(x)|f(x)|. So, its circular cross-sections have area πf(x)2=πf2(x)\pi|f(x)|^2 = \pi f^2(x). If we slice this solid into circular disks of thickness dx\dee{x}, then the disks have volume πf2(x)dx\pi f^2(x)\,\dee{x}. Therefore, the volume of the entire solid is abπf2(x)dx\displaystyle\int_a^b \pi f^2(x)\,\dee{x}. All we need to do now is get this into a form where we can replace the integral with the root mean square, RR.

V=abπf2(x)dx=πbabaabf2(x)dx=π(ba)(1baabf2(x)dx)2=π(ba)R2\begin{align*} V&=\int_a^b \pi f^2(x)\,\dee{x} = \pi\frac{b-a}{b-a}\int_a^b f^2(x)\,\dee{x}\\ &=\pi(b-a)\left(\sqrt{\frac{1}{b-a}\int_a^b f^2(x)\,\dee{x}}\right)^2\\ &=\pi(b-a)R^2 \end{align*}

Remark: the volume of a cylinder with length bab-a and radius rr is π(ba)r2\pi(b-a)r^2. So, the root mean square of f(x)f(x) gave us the radius of a cylinder with the same volume as the solid formed by rotating f(x)f(x). Recall the average of f(x)f(x) gave us the height of a rectangle with the same area as f(x)f(x). Compare this to the geometric interpretations of averages in Questions 1 and 15.

Q23Stage 3

Suppose f(x)=ax2+bx+cf(x)=ax^2+bx+c, and the average value of f(x)f(x) on the interval [0,1][0,1] is the same as the average of f(0)f(0) and f(1)f(1). What is aa?

Hint

The question tells you 11001f(x)dx=f(0)+f(1)2\frac{1}{1-0}\int_0^1 f(x)\,\dee{x} = \frac{f(0)+f(1)}{2}.

Answer

0

Full solution

The question tells you 11001f(x)dx=f(0)+f(1)2\frac{1}{1-0}\int_0^1 f(x)\,\dee{x} = \frac{f(0)+f(1)}{2}. Note f(0)=cf(0)=c, and f(1)=a+b+cf(1) = a+b+c.

11001f(x)dx=f(0)+f(1)2=c+(a+b+c)201(ax2+bx+c)dx=a+b+2c2[a3x3+b2x2+cx]01=a2+b2+ca3+b2+c=a2+b2+ca3=aa=0\begin{align*} \frac{1}{1-0}\int_0^1 f(x)\,\dee{x} &= \frac{f(0)+f(1)}{2} = \frac{c+(a+b+c)}{2}\\ \int_0^1 \big(ax^2+bx+c \big)\,\dee{x}&=\frac{a+b+2c}{2}\\ \left[\frac{a}{3}x^3+\frac{b}{2}x^2+cx \right]_0^1&=\frac{a}{2}+\frac{b}{2}+c \\\frac{a}{3}+\frac{b}{2}+c&=\frac{a}{2}+\frac{b}{2}+c\\ \frac{a}{3}&=a\\ a&=0 \end{align*}

That is, f(x)f(x) is linear.

Q24Stage 3

Suppose f(x)=ax2+bx+cf(x)=ax^2+bx+c, and the average value of f(x)f(x) on the interval [s,t][s,t] is the same as the average of f(s)f(s) and f(t)f(t). Is it possible that a0a \neq 0?

That is– does the result of Question 23 generalize?

Hint

Set up this question just like Question 23, but with variables for your limits of integration.

Note (st)2=s22st+t2(s-t)^2=s^2-2st+t^2.

Answer

Yes, but if a0a\neq0, then s=ts=t.

Full solution

The information given in the question is:

(as2+bs+c)+(at2+bt+c)2=1tsst(ax2+bx+c)dx=1ts[a3x3+b2x2+cx]st=1ts[a3(t3s3)+b2(t2s2)+c(ts)]=a3(t2+st+s2)+b2(t+s)+ca2(s2+t2)+b2(s+t)+c=a3(t2+st+s2)+b2(t+s)+ca2(s2+t2)=a3(t2+st+s2)a[s2+t22t2+st+s23]=0a[s22st+t26]=0a(st)2=0a=0ORs=t\begin{align*} \frac{\big(as^2+bs+c\big)+\big(at^2+bt+c\big)}{2}&=\frac{1}{t-s}\int_s^t \big(ax^2+bx+c\big)\,\dee{x}\\ &=\frac{1}{t-s}\left[\frac{a}{3}x^3+\frac{b}{2}x^2+cx\right]_s^t \\&=\frac{1}{t-s}\left[\frac{a}{3}(t^3-s^3)+\frac{b}{2}(t^2-s^2)+c(t-s)\right]\\ &=\frac{a}{3}(t^2+st+s^2)+\frac{b}{2}(t+s)+c\\ \frac{a}{2}(s^2+t^2) +\textcolor{red}{ \frac{b}{2}(s+t)}+\textcolor{blue}{c}&=\frac{a}{3}(t^2+st+s^2)+\textcolor{red}{\frac{b}{2}(t+s)}+\textcolor{blue}{c}\\ \frac{a}{2}(s^2+t^2)&=\frac{a}{3}(t^2+st+s^2)\\ a\left[\frac{s^2+t^2}{2} - \frac{t^2+st+s^2}{3}\right]&=0\\ a\left[\frac{s^2-2st+t^2}{6}\right]&=0 \\a(s-t)^2&=0\\ a&=0 \quad\text{OR}\quad s=t \end{align*}

So, unless s=ts=t (and we're taking the very boring average of a single point!) then a=0a=0. That is: f(x)f(x) is linear whenever sts \neq t.

Q25Stage 3

Let f(x)f(x) be a function defined for all numbers in the interval [a,b][a,b], with average value AA over that interval. What is the average of f(a+bx)f(a+b-x) over the interval [a,b][a,b]?

Hint

What are the graphs of f(x)f(x) and f(a+bx)f(a+b-x) like?

Answer

AA

Full solution

The function g(x)=f(a+bx)g(x)=f(a+b-x), on the interval [a,b][a,b], is a mirror of the function f(x)f(x), with g(a)=f(b)g(a)=f(b) and g(b)=f(a)g(b)=f(a). So, abf(a+bx)dx=abf(x)dx\int_a^b f(a+b-x)\,\dee{x} = \int_a^bf(x)\,\dee{x}, and hence 1baabf(a+bx)dx=1baabf(x)dx\frac{1}{b-a}\int_a^b f(a+b-x)\,\dee{x} = \frac{1}{b-a}\int_a^bf(x)\,\dee{x}, so the average value of f(a+bx)f(a+b-x) on [a,b][a,b] is AA.

Alternately, we can evaluate 1baabf(a+bx)dx\frac{1}{b-a}\int_a^b f(a+b-x)\,\dee{x} directly, using the substitution u=a+bxu=a+b-x, dx=dx\dee{x}=-\dee{x}:

1baabf(a+bx)dx=1bau(a)u(b)f(u)du=1babaf(u)du=1baabf(u)du=A\begin{align*} \frac{1}{b-a}\int_a^b f(a+b-x)\,\dee{x}&=\frac{-1}{b-a}\int_{u(a)}^{u(b)} f(u)\,\dee{u}\\ &=\frac{-1}{b-a}\int_{b}^{a} f(u)\,\dee{u} \\&=\frac{1}{b-a}\int_{a}^{b} f(u)\,\dee{u} \\&=A \end{align*}
Q26Stage 3

Suppose f(t)f(t) is a continuous function, and A(x)A(x) is the average of f(t)f(t) on the interval from 0 to xx.

  1. What is the average of f(t)f(t) on [a,b][a,b], where a<ba<b? Give your answer in terms of AA.

  2. What is f(t)f(t)? Again, give your answer in terms of AA.

Hint

For (b), express A(x)A(x) as an integral, then differentiate.

Answer

(a) bA(b)aA(a)ba\displaystyle\frac{bA(b) - aA(a)}{b-a} (b) f(t)=A(t)+tA(t)f(t)=A(t)+tA'(t)

Full solution
  1. The function A(x)A(x) only gives us information about an integral when one limit of integration is zero. We can get around this by using properties of definite integrals from Section 1.2 to break our integral into two integrals, each of which has 0 as one limit of integration. So, we find the average of f(t)f(t) on [a,b][a,b] as follows:

    1baabf(t)dt=1ba(a0f(t)dt+0bf(t)dt)=1ba(0af(t)dt+0bf(t)dt)=1ba(a1a0af(t)dtA(a)+b1b0bf(t)dtA(b))=1ba(aA(a)+bA(b))=bA(b)aA(a)ba\begin{align*} \frac{1}{b-a}\int_a^b f(t)\,\dee{t}&= \frac{1}{b-a}\left(\int_a^0 f(t)\,\dee{t} + \int_0^b f(t)\,\dee{t}\right)\\ &=\frac{1}{b-a}\left(-\int_0^a f(t)\,\dee{t} + \int_0^b f(t)\,\dee{t} \right)\\ &=\frac{1}{b-a}\left(-a\cdot\underbrace{\frac{1}{a}\int_0^a f(t)\,\dee{t}}_{A(a)} + b\cdot\underbrace{\frac{1}{b}\int_0^b f(t)\,\dee{t}}_{A(b)} \right)\\ &=\frac{1}{b-a}\left(-aA(a)+bA(b) \right) = \frac{bA(b) - aA(a)}{b-a} \end{align*}
  2. From the definition of A(x)A(x), we know

    A(x)=1x0xf(t)dtA(x) = \frac{1}{x}\int_0^x f(t)\,\dee{t}

    That is,

    xA(x)=0xf(t)dtxA(x) =\int_0^x f(t)\,\dee{t}

    To find f(x)f(x), we differentiate both sides. For the left side, we use the product rule; for the right side, we use the Fundamental Theorem of Calculus part 1.

    A(x)+xA(x)=f(x)A(x)+xA'(x) = f(x)

    So, f(t)=A(t)+tA(t)f(t)=A(t)+tA'(t).

Q27Stage 3
  1. Find a function f(x)f(x) with average 00 over [1,1][-1,1] but f(x)0f(x) \neq 0 for all xx in [1,1][-1,1], or show that no such function exists.

  2. Find a continuous function f(x)f(x) with average 00 over [1,1][-1,1] but f(x)0f(x) \neq 0 for all xx in [1,1][-1,1], or show that no such function exists.

Hint

For (b), consider the cases that f(x)f(x) is always bigger or always smaller than 0. Then, use the intermediate value theorem, Theorem 1.6.12 in the CLP-1 text.

Answer
  1. One of many possible answers: f(x)={1 if x01 if x>0f(x) = \begin{cases} -1&\text{ if } x \leq 0\\ 1&\text{ if } x > 0 \end{cases}.

  2. No such function exists.

    • Suppose f(x)>0f(x)>0 for all xx in [1,1][-1,1]. Then 1211f(x)dx>12110dx=0\frac{1}{2}\int_{-1}^1f(x)\,\dee{x} > \frac{1}{2}\int_{-1}^10\,\dee{x} =0. That is, the average value of f(x)f(x) on the interval [1,1][-1,1] is not zero–it's something greater than zero.

    • Suppose f(x)<0f(x)<0 for all xx in [1,1][-1,1]. Then 1211f(x)dx<12110dx=0\frac{1}{2}\int_{-1}^1f(x)\,\dee{x} < \frac{1}{2}\int_{-1}^10\,\dee{x} =0. That is, the average value of f(x)f(x) on the interval [1,1][-1,1] is not zero–it's something less than zero.

    So, if the average value of f(x)f(x) is zero, then f(x)0f(x)\ge 0 for some xx in [1,1][-1,1], and f(y)0f(y) \le 0 for some y[1,1]y \in [-1,1]. Since ff is a continuous function, and 0 is between f(x)f(x) and f(y)f(y), by the intermediate value theorem (Theorem 1.6.12 in the in the CLP-1 text) there is some value cc between xx and yy such that f(c)=0f(c)=0. Since xx and yy are both in [1,1][-1,1], then cc is as well. Therefore, no function exists as described in the question.

Full solution
  1. One of many possible answers: f(x)={1 if x01 if x>0f(x) = \begin{cases} -1&\text{ if } x \leq 0\\ 1&\text{ if } x > 0 \end{cases}.

  2. No such function exists.

    • Suppose f(x)>0f(x)>0 for all xx in [1,1][-1,1]. Then 1211f(x)dx>12110dx=0\frac{1}{2}\int_{-1}^1f(x)\,\dee{x} > \frac{1}{2}\int_{-1}^10\,\dee{x} =0. That is, the average value of f(x)f(x) on the interval [1,1][-1,1] is not zero–it's something greater than zero.

    • Suppose f(x)<0f(x)<0 for all xx in [1,1][-1,1]. Then 1211f(x)dx<12110dx=0\frac{1}{2}\int_{-1}^1f(x)\,\dee{x} < \frac{1}{2}\int_{-1}^10\,\dee{x} =0. That is, the average value of f(x)f(x) on the interval [1,1][-1,1] is not zero–it's something less than zero.

    So, if the average value of f(x)f(x) is zero, then f(x)0f(x)\ge 0 for some xx in [1,1][-1,1], and f(y)0f(y) \le 0 for some y[1,1]y \in [-1,1]. Since ff is a continuous function, and 0 is between f(x)f(x) and f(y)f(y), by the intermediate value theorem (Theorem 1.6.12 in the CLP-1 text) there is some value cc between xx and yy such that f(c)=0f(c)=0. Since xx and yy are both in [1,1][-1,1], then cc is as well. Therefore, no function exists as described in the question.

Q28Stage 3

Suppose f(x)f(x) is a positive, continuous function with limxf(x)=0\lim\limits_{x \to \infty} f(x)=0, and let A(x)A(x) be the average of f(x)f(x) on [0,x][0,x].

True or false: limxA(x)=0\lim\limits_{x \to \infty} A(x) = 0.

Hint

Try l'H^opital's rule.

Answer

true

Full solution

This seems like it might be true: if ff is getting closer and closer to zero, as xx grows towards infinity, then over time the later values will become a larger and larger portion of the total interval we're looking at, and so the average should look more and more like f(x)f(x) when xx is large–that is, like 0. That's some intuition to start us out, but it isn't a rigorous argument. To be sure we haven't overlooked something, let's use the definition of an average to express A(x)A(x).

A(x)=1x0xf(t)dtlimxA(x)=limx1x0xf(t)dt\begin{align*}A(x)&=\frac{1}{x}\int_0^x f(t)\,\dee{t}\\ \lim_{x \to \infty}A(x)&=\lim_{x \to \infty}\frac{1}{x}\int_0^x f(t)\,\dee{t}\end{align*}

If 0f(t)dt\int_0^\infty f(t) \,\dee{t} converges, then this limit is 0, and the statement is true. So, suppose it does not converge. Since f(x)f(x) is positive, that means limx0xf(t)d(t)=\lim\limits_{x \to \infty} \int_0^x f(t)\,\dee(t) = \infty, so we can use l'H^opital's rule. To differentiate the numerator, we use the Fundamental Theorem of Calculus part 1.

limxA(x)=limx0xf(t)dtx=limxf(t)1=0\begin{align*}\lim_{x \to \infty}A(x)&=\lim_{x \to \infty}\frac{\int_0^x f(t)\,\dee{t}}{x} = \lim_{x \to \infty}\frac{f(t)}{1}=0\end{align*}

So, the statement is true whether 0xf(t)dt\int_0^xf(t)\,\dee{t} converges or not.

Q29Stage 3

Let A(x)A(x) be the average of the function f(t)=et2f(t)=e^{-t^2} on the interval [0,x][0,x]. What is limxA(x)\displaystyle\lim_{x \to \infty} A(x)?

Hint

Use the result of Question 28.

Answer

0

Full solution

Note f(t)f(t) is a continuous function that takes only positive values, and limtf(t)=0\lim\limits_{t \to \infty} f(t)=0. By the result of Question 28, limxA(x)=0\displaystyle\lim_{x \to \infty} A(x)=0.

My list

nothing marked yet

Loading…

Open the whole list →

Your tutor can open this list with you. It follows your account, so it is there on whichever device you study on.

From the UBC Math 101 open textbook project, © Joel Feldman, Andrew Rechnitzer, Elyse Yeager and others. Licensed CC BY-NC-SA 4.0; this HTML adaptation is released under the same licence. Full attribution.