What is #int_(-pi)^pi ((x^2+1)/(x^3cosx)) dx# ?
2 Answers
Actually, it does not converge. It is of indeterminate form
Explanation:
Note that:
-
The integration range is symmetrical about
#0# . -
The function
#x^2+1# is even. -
The function
#x^3# is odd. -
The function
#cosx# is even.
Hence the function
So the positive and negative contributions to the integral will cancel out.
Hence:
#int_(-pi)^pi ((x^2+1)/(x^3 cos x)) dx = 0#
graph{(x^2+1)/(x^3cosx)(sqrt(pi^2-x^2)/sqrt(pi^2-x^2)) [-3.2, 3.2, -10, 10]}
The integral does not converge
The problem is that neither of the integrals:
#int_(-pi)^0 (x^2+1)/(x^3cos x) dx" "# #" "int_0^pi (x^2+1)/(x^3cos x) dx#
converges.
As
So if
#int_0^a (x^2+1)/(x^3 cos x) dx ~~ int_0^a 1/x^3 dx#
#color(white)(int_0^a (x^2+1)/(x^3 cos x) dx) = [-1/(2x^2)]_0^a#
#color(white)(int_0^a (x^2+1)/(x^3 cos x) dx) = -1/(2a^2)-(-1/(2(0^+)^2))#
#color(white)(int_0^a (x^2+1)/(x^3 cos x) dx) = +oo#
Hence the integral over the whole
Explanation:
The function:
does not have a primitive in terms of elementary functions. However you do not need to actually solve the integral.
Just note that
In general, the definite integral of an odd function over an interval that is symmetric with respect to the origin must be null, since:
Substitute in the first integral:
Inverting the limits of integration of the first integral changes its sign:
and as
Now in this particular case the integral:
and see that
which means that the integral is null in the sense of Cauchy's principal value.