Math 1B Calculus II
Chapter 8 test (1)
fall ’06 Name__________________________
Show your work for credit. Write all responses on separate paper. Don’t abuse a calculator.
1.
Show that can be evaluated using a geometric series of
the form,
.
What is the ratio, r and the leading coefficient, a?
What is the convergent value?
Hint: you’ll need basic properties of
exponents like and
.
2.
Use the integral test to determine whether is convergent or divergent.
Start by stating what continuous function is and how it meets the conditions of
the integral test.
3.
Use the integral test to determine whether the series
converges or diverges: .
Start by stating what continuous function is and how it meets the conditions of
the integral test.
4.
Evaluate . Hint:
factor the denominator and find its partial fractions representation…this leads
to a telescoping sum. Write out some
terms of the partial fraction representations to see the pattern of
cancellation.
5.
Determine whether the following series converges or
diverges: .
As usual, describe your method in detail.
6.
Use the limit comparison test to determine whether the
series converges or diverges: .
Hint: compare with a geometric
series.
7.
Show that is convergent. What value of N is needed to be sure the approximation
is correct to 20 significant digits?
8.
Write the function as a power series. Hint:
. Use the ratio test to determine the radius of
convergence of the series. What is the
interval of convergence?
Math 1B Calculus II
Chapter 8.1
8.6 test Solutions
fall ’06
1.
r = 4/9 and the leading coefficient, a = 324.
2.
Since is a positive valued and decreasing function,
we can use the integral test to determine whether
is convergent or divergent. Substituing
we see that, since the integral is divergent:
,
the series must also be divergent.
3.
Since is a positive valued function and, since
on
decreasing function, we can use the integral
test to determine whether
converges or diverges. Substitute
,
to see the integral is convergent:
. So the series is also convergent.
4.
The partial fractions expansion is a simple one: . This leads to a telescoping sum:
.
5.
The series converges since
and since
,
the series must converge.
Alternatively, use the limit comparison test with the convergent p-series, :
so, like the p-series with p = 2, the series converges.
6. Using
the limit comparison test with the geometric series ,
we have
must converge since, after a few applications
of L’Hopital’s rule,
.
7. is convergent since it’s alternating and
. To be sure the approximation
is correct to 20 significant digits, we need
only require that the magnitude of the first neglected term is
.
Somehow, I don’t think that 1 subtracted from a huge number is going to make a
difference, but how can I be more sure? Well,
here’s some calculations in Mathematica that make it seem like the convergence
is much faster than that. Can you
confirm the following somehow?
Entering, the command to numerically approximate
N[Sum[Cos[n*Pi]/Sqrt[n],{n,2,4*^40}],50]
0.39510135657836962975523408576404450024023745486975
whereas
N[Sum[(-1)^n/Sqrt[n],{n,2,4*^80}],50]
0.39510135657836962975273408576404450024026245486975
So that the difference is
N[Sum[Cos[n*Pi]/Sqrt[n],{n,2,4*^40}],50]-
N[Sum[Cos[n*Pi]/Sqrt[n],{n,2,4*^80}],50]
![]()
Which is twice
as good as was required. This makes
sense if you consider the nature of the path of this alternating series. Like a
pendulum swinging steadily with decreasing arcs of passage, each time it
swings, half the length of the swing is a better estimate of the distance from
one extreme of the string to the point at equilibrium (about halfway through the
swing.) The distance to the next
opposite peak will still cause a rounding error:
N[Sum[Cos[n*Pi]/Sqrt[n],{n,2,1*^40}],50]- |
N[Sum[Cos[n*Pi]/Sqrt[n],{n,2,1*^80+1}],50]
![]()
8.
Writing as a power series, The ratio test
leaves the endpoints to check. Since the function is not even defined at x = -1 and at x = 1 the terms are alternating between 1 and -1, the series does
not converge at the endpoints of this interval and the interval of convergence
is (-1,1).