Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

What is wrong with this method? (partial fractions)

  1. Jan 31, 2012 #1
    1. The problem statement, all variables and given/known data

    So I am trying to integrate the following. I believe I have to use partial fractions.


    2. Relevant equations

    3. The attempt at a solution

    So I am trying to expand the above into something that I can integrate. So I equated the integrand with the following:

    integrand = [itex]\frac{A}{(x-3)}[/itex]+[itex]\frac{Bx+C}{(x^{2}+16+8x) <--just the expanded form of (x+4)^{2}}[/itex]

    So by creating a common denominator you get

    1= A(x[itex]^{2}[/itex]+16+8x)+(Bx+C)(x-3)

    By inputing x=3 first, I solve for A=1/49
    Then by imputing x=0 I get C=-33/49
    And then by imputing x=1 I get B=-45/49

    So plugging that into the original, I get:

    integrand = [itex]\frac{1}{49(x-3)}[/itex]-[itex]\frac{(45/49)x-(33/49)}{(x+4)(x+4)}[/itex]

    When I check my answer by plugging a random value of x into the integrand, and then into the RHS, I find that they are not equal. Where did I mess up?
  2. jcsd
  3. Jan 31, 2012 #2


    User Avatar
    Science Advisor

    I don't know why you multiplied out [itex](x+ 4)^2[/itex]. If you leave it as is, you have
    [tex]\frac{1}{(x+4)^2(x-3)}= \frac{A}{x-3}+ \frac{Bx+ C}{(x+4)^2}[/tex]
    Multiplying on both sides by [itex](x-3)(x+4)^2[/itex] we get
    [tex]1= A(x+4)^2+ (Bx+ C)(x- 3)[/tex]

    Now, you can take x= 3 and get [itex]1= A(7)^2[/itex] so A= 1/49 as you say.

    If you let x= 0 you get [itex]1= 16A+ C(-7)[/itex] and knowing that A= 1/49, you get 7C= 16/49- 1= -33/49 so C= -33/343, not -33/49.

    Finally, if you let x=-4, you get [itex]1= (-4B+ C)(-7)[/itex] and can solve for B.
  4. Jan 31, 2012 #3
    OK Thanks for the help!
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook