1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Basic Integration question

  1. Dec 22, 2015 #1
    1. The problem statement, all variables and given/known data
    ∫ sec2 (7-3θ) dθ

    2. Relevant equations
    ∫ cos(kx+b)dx= ((sin(kx+b))/k)+C
    ∫ sin (kx+b)dx=-cos "
    (k and b are constants)
    3. The attempt at a solution
    The answer in the book is: ∫ -1/3(tan(7-3x))+C. I assume by x they meant θ.
    The book only gave the two formulas above for solving this, but my question is can these by simplified and expanded by saying ∫ f(g(x)) dx = F(g(x))/g'(x)? Because I looked online for a rule like this and found nothing. If this involves U-substitution or something of the sort, I couldn't find that in the textbook. Also, if you have any really good textbooks or sources for learning integration I don't like my textbook.
  2. jcsd
  3. Dec 22, 2015 #2


    User Avatar
    Staff Emeritus
    Science Advisor
    Homework Helper
    Gold Member

    What is the derivative of ##\ \tan(7-3\theta) \ ## ?
  4. Dec 22, 2015 #3


    User Avatar
    Staff Emeritus
    Science Advisor
    Homework Helper

    Instead of going off on wild goose chases and inventing spurious integration rules, refresh your knowledge of the derivatives of trig functions:

  5. Dec 22, 2015 #4
    Two ways to this problem.

    You can prove the integral of the basic sec^2(x) or you can use the following fact. the derivative of tan(x) is sec^2(x). This should be in your book, right after they prove it.
  6. Dec 22, 2015 #5
    ∫ f(g(x)) dx = F(g(x))/g'(x)
    In this case f(x)=sec^2 (x) and g(x)= 7-3x (or θ, doesn't matter)
    So F(g(x))=tan(7-3x) because ∫sec^2 (x) dx = tan x
    That would be divided by only g'. f is not in the denominator, it is just d/dx (7-3x) = -3. That is on the bottom. I don't use derivatives of trig functions

    I'm just saying that this: ∫ f(g(x)) dx = F(g(x))/g'(x), gave me the right answer here, but I don't want to use it unless I know that it will always work when integrating composite functions.
  7. Dec 22, 2015 #6
    I'm not trying to prove trig identities. I am taking ∫ sec^2(x)=tan x to be a given. The thing is that there is a function inside of the sec^2(x) function, which I have named g. Putting this function inside of the trig function creates a composite function which can't be integrated by only integrating the outer function (the trig identity).
  8. Dec 22, 2015 #7


    User Avatar
    Staff Emeritus
    Science Advisor
    Homework Helper

    You're making this too complicated.

    The basic integral is ∫ sec2(u) du = tan (u) + C

    Now, if your particular integral does not have a simple u as the argument of the secant function, you can use u-substitution to simplify things.

    Your particular argument is u = (7 - 3x). You should be able to calculate du very easily, do a little algebra, and obtain the correct result. :wink:
  9. Dec 22, 2015 #8
    thank you, and I found exactly what I was looking for on wikipedia under "leibniz's notation":
    Would have made things so much simpler to put this in the text book instead of arbitrary sin and cos integrals.
  10. Dec 22, 2015 #9

    Ray Vickson

    User Avatar
    Science Advisor
    Homework Helper

    [tex] \int f(g(x) \, dx = \frac{F(g(x))}{g'(x)} \Longleftarrow \text{FALSE!} [/tex]
    at least if you mean that ##F## is the antiderivative of ##f##. For one thing, differentiation of ##F(g(x))/g'(x)## does not give back ##f(g(x))##. For another thing it is easy to give counterexamples. For example, take ##f(y) = \sqrt{y}## and ##g(x) = x^2 + 1##., so that ##f(g(x)) = \sqrt{x^2+1}##. Then, we have ##F(y) = (2/3) y^{3/2}##, so
    [tex] F(g(x))/g'(x) = \frac{1}{3x} (x^2+1)^{3/2}, [/tex]
    [tex] \int \sqrt{x^2+1} \, dx = \frac{1}{2} x \sqrt{x^2+1} + \frac{1}{2} \, \text{arcsinh}(x) + \text{const.} [/tex]

    You have already been advised to not try making up you own, personal, integration rules, because if you are not an expert you are almost guaranteed to get it wrong.

    You say you do not like your textbook. Well, buy a cheap, used copy of another, or go to the library and check out an alternative. Or, do the modern equivalent of going to the library, and download one of the many free calculus textbooks available on-line.
  11. Dec 26, 2015 #10
    you are kind of right
    ∫ f(g(x))dx = ∫ { f(u)du/g'(x) } taking g(x)=u; [note:g'(x)is still inside the integral to be evaluated]
    now this integral is equal to F(g(x))/g'(x) (assuming ∫f(x)dx=F(x) ) only if g(x) is a linear function ie g(x)=ax+b a≠0;
    now g'(x)=a (a constant which can be taken out of the integral woohoo!!!)
    this gives us ∫ { f(u)du/g'(x) } =(1/a)*∫f(u)du= F(g(x))/g'(x) [i now remind you again that g must be a linear function only]
    now try out this integral just for fun
    ∫x sin(3x^2)dx (hint take 3x^2 as some other variable t)

  12. Dec 26, 2015 #11
    You could also use u-substitution to better simplify the integral for yourself before arriving at the answer.
  13. Dec 28, 2015 #12

    Thank you so much. This really helps.I took ∫x sin(3x^2)dx and got -cos(3x^2)/6+C. I checked this on an online integral calculator. Thanks again!!
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted

Similar Discussions: Basic Integration question