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!

Homework Help: Help with an integral

  1. Aug 11, 2015 #1
    • Member warned about posting with no effort shown
    1. The problem statement, all variables and given/known data
    Here is the integral:


    3. The attempt at a solution

    I tried doing a partial fraction decomposition, but I'm not sure if that is permitted since the numerator is not a regular polynomial. Any help would be greatly appreciated! Thanks.
  2. jcsd
  3. Aug 11, 2015 #2
    There's a tricky u-substitution you can use that's not the most obvious and it has something to do with that numerator. Can you see what it is?
  4. Aug 11, 2015 #3
    How about another hint?
  5. Aug 11, 2015 #4
    For future reference, you have to type out your try at a solution. I'll be nice this time, and hopefully I do not get a warning for helping you.

    You can combine multiple techniques. You may have to preform a substitution or a algebraic manipulation. Since you tried partial fractions, yes you cannot proceed because of that e^(1/x).

    Try a u-sub and tell me what you get. I actually solved this problem. It is very long. Maybe there is a short-cut but i could not see it.
  6. Aug 11, 2015 #5
    Physicsnorum Physics Forums does not operate the same way google answer does. You have to actually make attempts and try. This forum is not a solutions manual.
    Last edited by a moderator: Aug 11, 2015
  7. Aug 11, 2015 #6
    Thanks for your reply.

    I guess I'm getting stuck pretty early in the problem. I tried letting u=1/x, then the integral turns into -e^(u)x/[(x+1)^2] du, but that is a mess. I don't really know what else to try. Maybe you can point to a step I can make to kelp me get my foot in the door of a solution?
  8. Aug 11, 2015 #7


    Staff: Mentor

    I haven't worked it through like MidgetDwarf has, so I don't know what works. It might be that u = 1/x is a good substitution, but once you've done the substitution, your integrand should be entirely in terms of u and du -- no x terms or dx should still remain.

    One thing you might try is to use partial fractions on the ##\frac 1 {x(x + 1)^2}## part. That way you could break up the integral into three integrals of the form
    $$A\int \frac{e^{1/x} dx}{x} + B\int \frac{e^{1/x} dx}{x + 1} + C\int \frac{e^{1/x} dx}{(x + 1)^2}$$
    I don't know if this hint is helpful. My aim is splitting up one harder integral into three that are easier.
  9. Aug 11, 2015 #8
    How can you rewrite the x terms as u's using that substitution? In other words what does x equal in terms of u?
  10. Aug 11, 2015 #9
    I think I worked through the substitution as you suggested and now have: [-u*e^u]/[(u+1)^2] du. I got this using x=1/u. I think I must be missing something because this looks just as tough to me as the original one. Is there some way I should be leveraging integration by parts at this point?
  11. Aug 11, 2015 #10
    No, you are on the right track. What other integration techniques do you have? We have trig, partial fractions, u-sub, by parts? Which one of these will work?

    and no it is not as tough as the original. You got rid of the 1/x exponent on the e.

    The point is, no matter how scary the integral problems look. 90 percent of the problems in your book can be worked out without resorting to more advance methods.

    It just simply comes down to noticing what techniques may work and how to algebraically manipulate the function to fit the integral formulas and techniques we already have.
  12. Aug 11, 2015 #11
  13. Aug 11, 2015 #12

    Ray Vickson

    User Avatar
    Science Advisor
    Homework Helper

    If u = 1/x, isn't it really easy to get x in terms of u?
  14. Aug 11, 2015 #13
    Alright, I have taken your hints, for which I am of course very thankful, and have now "decomposed" the u-substitution into two separate terms. That is, from [-u*e^u]/[(u+1)^2]du, I got: e^u/[(u+1)^2] - e^u/(u+1). As far as I am concerned, I have just gone from having one integral that I am incapable of solving, to having two of them ...:nb). Was it correct to do that decomposition?
  15. Aug 11, 2015 #14


    User Avatar
    Staff Emeritus
    Science Advisor
    Homework Helper
    Gold Member

    Now, maybe let t = u+1, i.e: u = t-1 .

    It looks like after some simplifying, you may need integration by parts.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted