Integrating sec(x): Understanding the Answer

  • Context: High School 
  • Thread starter Thread starter ForceBoy
  • Start date Start date
  • Tags Tags
    Integral
Click For Summary

Discussion Overview

The discussion revolves around the integration of the function sec(x) and the discrepancies between different solutions obtained by participants. It includes explorations of complex analysis and the validity of certain integral forms, as well as verification through differentiation.

Discussion Character

  • Exploratory
  • Technical explanation
  • Debate/contested
  • Mathematical reasoning

Main Points Raised

  • One participant presents their integration process for sec(x) using complex analysis, arriving at a solution that differs from the standard integral found in tables.
  • Another participant points out that the integral of 1/(1+u^2) = arctan(u) is valid only for real u, questioning the application of this result to the complex variable u = e^(ix).
  • A subsequent participant suggests evaluating the integral with a complex variable z instead, indicating a shift in approach.
  • Further discussion arises about the limitations of arctan(u) when u is complex, with one participant noting that the logarithmic form is more useful in this context.
  • One participant discovers a way to express their result in terms of ln|sec(x) + tan(x)|, suggesting a connection to the standard solution.
  • Another participant expresses satisfaction upon reaching the standard solution, indicating a resolution to their initial confusion.

Areas of Agreement / Disagreement

Participants do not reach a consensus on the validity of using arctan(u) for complex u, and there are competing views on the best approach to integrate sec(x). However, there is agreement on the usefulness of the logarithmic form in expressing the integral.

Contextual Notes

The discussion highlights the complexities involved in integrating functions with complex variables and the assumptions underlying certain integral forms. The transition from arctan to logarithmic expressions is noted but remains context-dependent.

Who May Find This Useful

This discussion may be of interest to those studying complex analysis, integration techniques, and the nuances of mathematical expressions in different contexts.

ForceBoy
Messages
47
Reaction score
6
Hello.

I integrated ##\sec(x) ## and got an answer. I differentiated it to verify it and it came out well. Later when I was looking in a table of integrals I saw the solution ## \ln| \sec(x) +\tan(x) | + c##. This was completely different than my solution. I do not think I made a mistake (unless I'm wrongly convinced of something) and I can't see how my answer is related to the given one.

My work:

##I = \int\sec(x) dx ##

## \sec x = \frac{1}{\cos x}## and ##\cos(x) = \frac{e^{ix}+e^{-ix}}{2}## so

##I = \int \frac{2dx}{e^{ix}+e^{-ix}}##

## I = 2 \int \frac{dx}{e^{ix}+e^{-ix}} ##

Multiplication by ##\frac{e^{ix}}{e^{ix}}## yields

## I = 2 \int \frac{e^{ix}dx}{e^{2ix}+1} ##

Let ##u = e^{ix}##
then ##du = i u dx##
so ##\frac{du}{iu} = dx ##

The integral now:
##I = 2\int \frac{u}{u^{2}+1} \frac{du}{iu}##

##I = -2i\int \frac{1}{u^{2}+1}##

We know ##\int \frac{1}{u^{2}+1} = arctan(x)+c##

So it follows that
##I = -2i \arctan(u) + c##

and finally that

##I = -2i \arctan(e^{ix}) + c##

Differentiation:

## \frac{d}{dx}(-2i \arctan(e^{ix})) ##

## -2i \frac{d}{dx}(\arctan(e^{ix})) ##

##t = ix##
##u = e^{t} ##
##v = \arctan(u)##

so

## -2i ( \frac{dv}{du} * \frac{du}{dt} *\frac{dt}{dx}) ##

##-2i(\frac{1}{u^{2}+1}*u*i) ##

##-2i( \frac{i e^{ix}}{e^{2ix}+1}) ##

## -2i(\frac{i e^{ix}}{e^{ix} (e^{ix} +e^{-ix})})##

##-2i(\frac{i}{e^{ix} +e^{-ix}})##

##\frac{2}{e^{ix} +e^{-ix}}##

##\sec(x)##

This is then clearly different than the mentioned solution. I don't see where I could've made a mistake. If I'm right, my solution would require more simplification. I don't see how this could be done.

Any thoughts would be appreciated. Thanks.
 
Physics news on Phys.org
##\int \frac{1}{1+u^2}du=arctan(u)+c## holds only for real u. Here ##u=e^{ix}## is complex.
 
Last edited:
I understand. Would my next step then be to evaluate ##\int \frac{1}{z^{2} +1} dz ## where z is a complex number?
 
Also a question.
mathman said:
##\int \frac{1}{1+u^2}du=arctan(u)+c## holds only for real u. Here ##u=e^{ix}## is complex.

Why is this? In the differentiation of arctan(x), where is it that it is assumed that x is real?
 
Technically you are correct. However arctan(u) when u is complex is not very useful. It is more useful with the form given (ln|...|)..
 
I now saw how to evaluate ## \int \frac{du}{u^2+1} ## for complex u. It's ## \frac{i}{2}\ln|\frac{u+i}{u-i}| ## if I'm not wrong. If I use this instead of ##arctan(u)## I get:

## I = \ln|\frac{e^{ix}+i}{e^{ix}-i}| ##

When looking at this, something tells me it can quickly be expressed as ## \ln|\sec(x)+\tan(x)| ##. Is this right?
 
Last edited:
I just found out how to make it ## \ln|\sec(x)+\tan(x)| ##. I now understand this better and know how to integrate sec(x). Thank you for your help and time. :smile:
 

Similar threads

  • · Replies 1 ·
Replies
1
Views
3K
  • · Replies 5 ·
Replies
5
Views
3K
  • · Replies 4 ·
Replies
4
Views
3K
  • · Replies 1 ·
Replies
1
Views
4K
  • · Replies 27 ·
Replies
27
Views
3K
  • · Replies 2 ·
Replies
2
Views
3K
  • · Replies 11 ·
Replies
11
Views
7K
  • · Replies 8 ·
Replies
8
Views
1K
  • · Replies 5 ·
Replies
5
Views
1K
  • · Replies 9 ·
Replies
9
Views
2K