Different Substitution for Solving an Integral

Click For Summary

Discussion Overview

The discussion revolves around solving the integral ##\int \frac{1}{x\sqrt{3x^2+2x-1}}dx## using different substitution methods. Participants explore various approaches, compare results, and analyze the implications of their findings, focusing on the mathematical reasoning behind their solutions.

Discussion Character

  • Exploratory
  • Mathematical reasoning
  • Debate/contested

Main Points Raised

  • One participant describes their method of substitution ##\frac{1}{u}## and completing the square, leading to a trigonometric substitution.
  • Another participant points out that both integrals derived from different methods are equivalent but differ by a constant, suggesting that the discrepancy in results may stem from hidden constants associated with the integrals.
  • Some participants provide specific forms of the integral's solutions, indicating that the constants can vary depending on the domain of the integrand.
  • There is mention of the domain being disconnected, which could affect the equality of constants across different components.
  • One participant expresses confusion regarding the presence of hidden constants, noting that they can arise in various forms, not just logarithmic functions.
  • Further analysis involves checking the relationship between the two expressions derived from different substitution methods, using trigonometric identities to validate their equivalence.

Areas of Agreement / Disagreement

Participants generally agree that the two different expressions for the integral are equivalent, differing only by constants. However, there is no consensus on the specific values of these constants or the implications of the disconnected domain on their equality.

Contextual Notes

The discussion highlights the complexity of integrating functions with multiple substitution methods and the potential for hidden constants that can arise in different domains. The participants note that the integrand's domain is disconnected, which may influence the behavior of the constants in the solutions.

Who May Find This Useful

This discussion may be useful for students and practitioners in mathematics and physics who are exploring integration techniques, particularly those involving substitutions and the implications of constants in indefinite integrals.

12Element
Messages
11
Reaction score
1
TL;DR
Asking about multiple approaches for integration by substitution with an integral involving the reciprocal of a square root of a quadratic equation.
I have a question about how to solve the following integral: ##\int \frac 1 {x\sqrt{3x^2+2x-1}}dx##, I know how to get the right answer (which agrees with WolframAlpha and the book I got the question from) through substituting ##\frac 1u## for ##x##, completing the square and finally doing a trig substitution. However, I'm not sure why I'm not getting the right answer with the below approach, I would appreciate it if someone looks at my solution and tell me where and why I'm wrong.
$$ \int \frac 1 {x\sqrt{3x^2+2x-1}}dx$$
Using the below substitution:
$$3x^2+2x-1=3\left( t-x \right)^2$$
$$3x^2+2x-1=3t^2-6tx+3x^2$$
$$6tx+2x=3t^2+1$$
$$x=\frac {3t^2+1}{6t+2}$$
The derivative of ##x## with respect to ##t## is:
$$\frac {dx}{dt} = \frac {6t \left( 6t+2 \right) - 6 \left( 3t^2+1 \right)} {\left( 6t+2 \right)^2} = \frac {36t^2+12t-18t^2-6} {\left( 6t+2 \right)^2}$$
$$dx = \frac {6 \left( 3t^2+2t-1 \right)} {\left( 6t+2 \right)^2}dt$$
Also, to find the value of ##t##:
$$3x^2+2x-1=3\left( t-x \right)^2$$
$$\sqrt {3x^2+2x-1}=\sqrt{3}\left( t-x \right)$$
$$t= \frac {\sqrt {3x^2+2x-1}}{\sqrt{3}} + x$$
To find the value of ##\sqrt {3x^2+2x-1}## in terms of ##t##:
$$\sqrt {3x^2+2x-1}=\sqrt{3}\left( t-x \right)$$
substituting for ##x## with the value derived above:
$$\sqrt {3x^2+2x-1}=\sqrt{3}\left( t-\frac {3t^2+1}{6t+2} \right)=\sqrt{3}\left( \frac {6t^2+2t-3t^2-1}{6t+2} \right)$$
$$\sqrt {3x^2+2x-1}=\sqrt{3}~ \frac {3t^2+2t-1}{6t+2} $$
Substituting back into the original integral:
$$ \int \frac 1 {x\sqrt{3x^2+2x-1}}dx = \int \frac {\frac {6 \left( 3t^2+2t-1 \right)} {\left( 6t+2 \right)^2}}{\frac {3t^2+1}{6t+2} \sqrt{3}~ \frac {3t^2+2t-1}{6t+2}}dt= \frac6{\sqrt{3}}\int\frac1{3t^2+1}dt $$
$$= \frac2{\sqrt{3}}\int\frac1{t^2+ \frac13}dt = \frac2{\sqrt{3}}\sqrt{3} \tan^{-1} \left(\sqrt{3} t\right) + C$$
Substituting back for ##t## with ##x##:
$$= 2\tan^{-1} \left[ \sqrt{3}\left( \frac {\sqrt {3x^2+2x-1}}{\sqrt{3}} + x\right) \right] + C$$
$$= 2\tan^{-1} \left( \sqrt{3x^2+2x-1} + \sqrt{3}x \right) + C$$
 
Last edited:
Physics news on Phys.org
For comparison, what is the right answer you got first ?
 
12Element said:
I would appreciate it if someone looks at my solution and tell me where and why I'm wrong.
You're not wrong: both integrals are the same, they just differ by a constant!
Using Mathematica, your original indefinite integral is given by:$$I\equiv\int\frac{1}{x\sqrt{3x^{2}+2x-1}}dx=-2\tan^{-1}\left(\frac{\sqrt{3x^{2}+2x-1}}{3x-1}\right)+C\tag{1}$$whereas using your substitution it evaluates to:$$I_{S}\equiv2\tan^{-1}\left(\sqrt{3x^{2}+2x-1}+\sqrt{3}x\right)+C_{S}\tag{2}$$(Incidentally, Mathematica's change-of-variable operation agrees with your substitution and the form of the resulting integral.) Taking the difference of (1) and (2) and performing some algebra (or what's easier, just graphing the difference vs. ##x##) shows that:$$
I-I_{S}=C-C_{S}=\begin{cases}
\frac{2\pi}{3} & \left(x<\frac{1}{3}\right)\\
-\frac{4\pi}{3} & \left(x>\frac{1}{3}\right)
\end{cases}\tag{3}
$$
 
  • Like
Likes   Reactions: dextercioby and 12Element
anuttarasammyak said:
For comparison, what is the right answer you got first ?
The answer can be in either form ##\ -sin^{-1} \left(\frac{1-x}{2x} \right)+C= \ -2tan^{-1} \left( \frac {\sqrt{3x^2+2x-1}}{3x-1} \right)+C ##.

renormalize said:
Taking the difference of (1) and (2) and performing some algebra (or what's easier, just graphing the difference vs. ##x##) shows that:$$
I-I_{S}=C-C_{S}=\begin{cases}
\frac{2\pi}{3} & \left(x<\frac{1}{3}\right)\\
-\frac{4\pi}{3} & \left(x>\frac{1}{3}\right)
\end{cases}\tag{3}
$$
Thanks a lot, this is really interesting, I only have a Casio calculator and whenever substituting random values in both solution does not give the same number I assume my solution is wrong. I only thought that hidden constants are associated with logarithmic functions, now I know they can be more devious than what I thought. Not to mention improper integrals (like in this case) where multiple constants can be present along the domain. Thanks again.
 
Thanks. I let wolfram draw the graphs.
https://www.wolframalpha.com/input?i=plot+y=-+arcsin+(+(1-x)/(2x))
1728687812256.png



The right answer for x<0 needs -1 to be multiplied.
https://www.wolframalpha.com/input?i=plot+y=-+sgn(x)+arcsin+(+(1-x)/(2x))
1728697512563.png

So for x<-1
\int_x^{-1} \frac{dt}{t\sqrt{3t^2+2t-1}}=-\sin^{-1}\frac{1-x}{2x}-\frac{\pi}{2}
for 1/3 <x
\int_{1/3}^x \frac{dt}{t\sqrt{3t^2+2t-1}}=-\sin^{-1}\frac{1-x}{2x}+\frac{\pi}{2}

https://www.wolframalpha.com/input?i=plot+y=-2+arctan+(+\sqrt(3x^2+2x-1)/(3x-1))

1728685909493.png

So for x<-1
\int_x^{-1} \frac{dt}{t\sqrt{3t^2+2t-1}}=2\tan^{-1}\frac{\sqrt{3x^2+2x-1}}{3x-1}
for 1/3 <x
\int_{1/3}^x \frac{dt}{t\sqrt{3t^2+2t-1}}=-2\tan^{-1}\frac{\sqrt{3x^2+2x-1}}{3x-1}+\pi


https://www.wolframalpha.com/input?i=plot+y=2+arctan+(+\sqrt(3x^2+2x-1)+\sqrt(3)x+)
1728686188858.png


So for x<-1
\int_x^{-1} \frac{dt}{t\sqrt{3t^2+2t-1}}=-2\tan^{-1}(\sqrt{3x^2+2x-1}+\sqrt{3}x)-\frac{2\pi}{3}
for 1/3 <x
\int_{1/3}^x \frac{dt}{t\sqrt{3t^2+2t-1}}=2\tan^{-1}(\sqrt{3x^2+2x-1}+\sqrt{3}x)-\frac{\pi}{3}
 
Last edited:
12Element said:
I only thought that hidden constants are associated with logarithmic functions, now I know they can be more devious than what I thought.

An antiderivative of zero on a disconnected domain is constant on each connected component of the domain. but the constants on different connected components don't have to be equal. Here, the domain of the integrand as a real function is (-\infty, -1) \cup (\frac13,\infty) which is disconnected.
 
  • Like
Likes   Reactions: SammyS
As noted in post #5
For the two different expressions are same we expect the relation
\tan^{-1} (\sqrt{(3x-1)(x+1)}+\sqrt{3}x) + \tan^{-1} \frac{\sqrt{(3x-1)(x+1)}}{3x-1}=-\frac{\pi}{3}
\tan(\tan^{-1} (\sqrt{(3x-1)(x+1)}+\sqrt{3}x) + \tan^{-1} \frac{\sqrt{(3x-1)(x+1)}}{3x-1})=-\sqrt{3}
for x<-1
\tan^{-1} (\sqrt{(3x-1)(x+1)}+\sqrt{3}x) + \tan^{-1} \frac{\sqrt{(3x-1)(x+1)}}{3x-1})=\frac{2\pi}{3}
\tan(\tan^{-1} (\sqrt{(3x-1)(x+1)}+\sqrt{3}x) + \tan^{-1} \frac{\sqrt{(3x-1)(x+1)}}{3x-1})=-\sqrt{3}
for x>1/3
They are same. By tan addition law I have checked that it is valid.
 
Last edited:

Similar threads

  • · Replies 4 ·
Replies
4
Views
3K
  • · Replies 6 ·
Replies
6
Views
3K
  • · Replies 12 ·
Replies
12
Views
3K
  • · Replies 4 ·
Replies
4
Views
4K
  • · Replies 3 ·
Replies
3
Views
3K
  • · Replies 27 ·
Replies
27
Views
2K
  • · Replies 19 ·
Replies
19
Views
4K
  • · Replies 5 ·
Replies
5
Views
2K
  • · Replies 8 ·
Replies
8
Views
2K
  • · Replies 2 ·
Replies
2
Views
2K