Tag: Parameter Integrals

  • A Surprising Improper Integral: Why the Integral of sin(ax)/x Is Always π/2

    A Surprising Improper Integral

    Consider the improper integral

    ∫ 0 ∞ sin ( ax ) x dx.

    At first glance, this integral looks difficult. The factor 1x suggests a singularity at the origin, while the sine function continues to oscillate forever as x→∞. There is no elementary antiderivative that immediately resolves the problem.

    Nevertheless, for every positive number a, the answer is remarkably simple:

    ∫ 0 ∞ sin ( ax ) x dx = π2.

    Even more surprisingly, the answer does not depend on the positive value of a. Let us see why.

    The Main Idea: Add a Damping Factor

    Instead of attacking the original integral directly, introduce a positive parameter t and define

    F ( t ) = ∫ 0 ∞ e −tx sin ( ax ) x dx , t>0.

    The factor e −tx suppresses the oscillations for large values of x. This makes the parameter-dependent integral easier to work with.

    The key step is to differentiate with respect to the parameter t. We obtain

    F′ ( t ) = − ∫ 0 ∞ e −tx sin ( ax ) dx.

    Notice what happened: the troublesome factor 1x has disappeared. We are left with a standard Laplace-type integral.

    Evaluating the Easier Integral

    For t>0, we have

    ∫ 0 ∞ e −tx sin ( ax ) dx = a t2 + a2 .

    Therefore,

    F′ ( t ) = − a t2 + a2 .

    Now the problem has been reduced to an elementary integral.

    Recovering F(t)

    As t→∞, the exponential damping becomes stronger and

    F ( t ) → 0.

    Thus, for a>0,

    F ( t ) = ∫ t ∞ a u2 + a2 du.

    Evaluating this integral gives

    F ( t ) = π2 − arctan ( ta ).

    Equivalently, using the elementary arctangent identity,

    F ( t ) = arctan ( at ).

    So we have actually obtained the more general and useful formula

    ∫ 0 ∞ e −tx sin ( ax ) x dx = arctan ( at ), a>0, t>0.

    Removing the Damping

    We introduced the exponential factor only to make the integral easier to evaluate. Now let t→0+ . Then

    arctan ( at ) → π2.

    and the damping factor approaches 1. This leads to the celebrated Dirichlet integral

    ∫ 0 ∞ sin ( ax ) x dx = π2, a>0.

    Why Does the Answer Not Depend on a?

    There is also a simple scaling argument that explains why the answer must be the same for every positive a. Set

    u=ax.

    Then

    x = ua, dx = du a .

    Therefore,

    ∫ 0 ∞ sin ( ax ) x dx = ∫ 0 ∞ sin ( u ) u du.

    The parameter a has completely disappeared. Changing a changes how rapidly the sine function oscillates, but the total value of the improper integral remains unchanged.

    What If a Is Zero or Negative?

    If a=0, the integrand is identically zero, so the integral is zero.

    If a<0, use the oddness of the sine function:

    sin ( ax ) = − sin ( −ax ).

    Hence the complete result is

    ∫ 0 ∞ sin ( ax ) x dx = π2 if a>0, 0 if a=0, −π2 if a<0.

    One Important Detail: The Integral Is Not Absolutely Convergent

    The convergence of this integral is subtle. Although

    ∫ 0 ∞ sin ( ax ) x dx

    converges for a≠0, the corresponding absolute-value integral

    ∫ 0 ∞ | sin ( ax ) | x dx

    diverges. Thus the positive and negative oscillations of the sine function are essential. They cancel one another just enough for the original improper integral to converge.

    A Useful Lesson

    The most interesting part of this calculation is not simply the final answer π2. It is the method.

    When an integral is difficult to evaluate directly, it can sometimes be embedded into a family of integrals depending on a parameter. Differentiating with respect to that parameter may transform the original problem into a much easier one. After solving the parameterized problem, we return to the original integral by taking a limit.

    In this example, the chain of ideas is

    sin ( ax ) x → e −tx sin ( ax ) x → F′ ( t ) → F ( t ) → π2.

    A difficult oscillatory improper integral has been reduced to an elementary rational integral. That is what makes the Dirichlet integral such a beautiful example of the power of introducing a parameter.


    Another Surprise from an Improper Integral

    The integral involving sin ⁡ ( a x ) x shows that an oscillating function extending over an infinite interval can nevertheless produce a beautifully simple finite value.

    There is another famous improper-integral paradox in which infinity appears in a completely different way: a surface extending forever can enclose a finite volume while having infinite surface area.

    Continue exploring: Gabriel’s Horn: When Can an Infinite Horn Be Painted?


    From the Dirichlet Integral to the Gaussian Integral

    There is another beautiful connection behind the Dirichlet integral. The Gaussian function e − x 2 leads to one of the most famous improper integrals in mathematics. Its evaluation introduces a remarkably powerful idea: turn a one-dimensional integral into a two-dimensional one and then use geometry.

    That same Gaussian structure appears in many unexpected places and provides another route into the world of remarkable improper integrals.

    Continue exploring: The Gaussian Integral and Beyond: From e^(-x²) to a Family of Integrals