How to Integrate dz/ (y^2 + (x-z)^2))^1/2 Using Inverse Hyperbolic Functions?

  • Thread starter Thread starter Arthy
  • Start date Start date
  • Tags Tags
    Integrate
Join the discussion
Registration is free. Start your own thread to ask a follow-up.
5 replies · 2K views
Arthy
Messages
2
Reaction score
0
Integral of dz/ (y^2 + (x-z)^2))^1/2
 
Physics news on Phys.org
I would do it as a sequence of substitutions (planck42's use of a single substitution is correct and faster but harder to see).

First, because that "[itex](x- z)^2[/itex]" is a nuisance, let u= x- z. Then du= -dz so the integral becomes
[tex]-\int \frac{du}{y^2+ u^2}[/tex].

Now, remembering that [itex]sin^2(\theta)+ cos^2(\theta)= 1[/itex] so that [itex]tan^1(\theta)+ 1= sec^2(\theta)[/itex] and that [itex]y^2+ u^2= y^2(1+ (u/y)^2[/itex], let [itex]u/y= tan(\theta)[/itex] so that [itex]du= y^2 sec^2(\theta)d\theta[/itex] and [itex]y^2+ u^2= y^2(1+ (u/y)^2)= y^1(sec^2(\theta))[/itex] (Again, Planck42's hyperbolic substitution works fine but I learned trig substitutions before hyperbolic substutions so I tend to think of them first!). That makes the integral
[tex]\int\frac{y^2 sec^2(\theta)d\theta}{y^2 sec^2(\theta)}[/tex]
 
Thank you for the reply, but there is a square root at the denominator (y^2+ (x-z)^2)^1/2.

I know the answer of this derivation, with respect to limits 0 to L,

Integral of
dz/ Square root of (y^2+(x-z)^2) =
ln (x+ Square root of (x^2+y^2)/ x-L + square root of ((x-L)^2+y^2))

Please explain me the derivation part.
 
Going off of Halls' steps,
u/y = tanθ → u = y·tanθ
du = y·sec2θ dθ
[tex]\sqrt{(y^2+ u^2)} = \sqrt{(y^2(1+ (u/y)^2))} = \sqrt{(y^2(\sec^2\theta)} = y\sec \theta[/tex]

Then you get
[tex]\int \frac{y \sec^2 \theta ~d\theta}{y\sec \theta} = \int \sec\theta ~d\theta[/tex]
 
Let x-z be y sinh t
dz=-ycosh t dt

Using the identity [tex]cosh^{2}t=1+sinh^{2}t[/tex], we get

[tex]{\int}-dt[/tex]

Which is just -t. Now to put back the x's, y's, and z's.

[tex]x-z=ysinht[/tex] so
[tex]t=sinh^{-1}\frac{x-z}{y}[/tex]

The rest is just a matter of inserting the bounds L and 0(and looking up the inverse hyperbolic functions; they're why your answer is a nasty ln function)