Euler’s substitutions for integration


In the integration task

R(x,ax2+bx+c)𝑑x,

where the integrand is a rational functionMathworldPlanetmath of x and ax2+bx+c, the integrand can be changed to a rational function of a new variableMathworldPlanetmath t by using the following substitutions of Euler.

  • The first substitution of Euler.  If  a>0,  we may write

    ax2+bx+c=±xa+t. (1)

    When we take a with the minus sign, then

    ax2+bx+c=ax2-2xta+t2,

    from which we get the expression

    x=t2-cb+2ta;

    thus also dx is expressible rationally via t. We have

    ax2+bx+c=-xa+t=c-t2b+2taa+t.
  • The second substitution of Euler.  If  c>0,  we take

    ax2+bx+c=xt±c. (2)

    With the minus sign we obtain, similarly as above,

    x=2tc+bt2-a.
  • The third substitution of Euler.  If the polynomialMathworldPlanetmathPlanetmath ax2+bx+c has the real zeros α and β, we may chose

    ax2+bx+c=(x-α)t. (3)

    Now

    ax2+bx+c=a(x-α)(x-β)=(x-α)2t2,

    whence  a(x-β)=(x-α)t2. This gives the expression

    x=aβ-αt2a-t2.

    As in the preceding cases, we can express dx and ax2+bx+c rationally via t.

Examples.

1. In the integral dxx2+c we can use the first substitution:  x2+c=-x+t;  then  x2+c=x2-2xt+t2  and thus

x=t2-c2t,dx=t2+c2t2dt,x2+c=-t2-c2t+t=t2+c2t.

Accordingly we obtain

dxx2+c=t2+c2t2dtt2+c2t=dtt=ln|t|+C=ln|x+x2+c|+C.

Especially the cases  c=±1  give the formulas

dxx2+1=arsinhx+C,dxx2-1=arcoshx+C(x>1).

2. The integral c2-x2x𝑑x is needed in deriving the equation of the tractrix. We use for integrating the second substitution  c2-x2=xt-c;  then  c2-x2=x2t2-2cxt+c2, which implies

x=2ctt2+1,dx=2c(1-t2)dt(1+t2)2,c2-x2=2ct2t2+1-c=c(t2-1)t2+1.

We then obtain

c2-x2x𝑑x=-c(1-t2)2t(1+t2)2𝑑t=c(4t(1+t2)2-1t)𝑑t=-2c1+t2-cln|t|+C1.

The equation tying x and t gives  2c1+t2=xt  and  t=c+c2-x2x,  whence

c2-x2x𝑑x=-x2c+c2-x2-clnc+c2-x2x+C1=-c+c2-x2-clnc+c2-x2x+C1,

i.e.

c2-x2x𝑑x=c2-x2-clnc+c2-x2x+C.

3. In the integral dxx2+3x-4, the radicand is (x+4)(x-1). Using the third substitution of Euler, we take  x2+4x-3=(x+4)t. This simplifies to  x-1=(x+4)t2. Then we get

x=1+4t21-t2,dx=10t(1-t2)2dt,x2+3x-4=(1+4tr1-t2+4)t=5t1-t2.

And we obtain

dxx2+3x-4=10t(1-t2)(1-t2)25t𝑑t=21-t2𝑑t=ln|1+t1-t|+C=ln|1+x-1x+41-x-1x+4|+C
=ln|x+4+x-1x+4-x-1|+C.

References

  • 1 N. Piskunov: Diferentsiaal- ja integraalarvutus kõrgematele tehnilistele õppeasutustele. Viies, täiendatud trükk.  Kirjastus “Valgus”, Tallinn (1965).
Title Euler’s substitutions for integration
Canonical name EulersSubstitutionsForIntegration
Date of creation 2013-03-22 17:19:43
Last modified on 2013-03-22 17:19:43
Owner pahio (2872)
Last modified by pahio (2872)
Numerical id 15
Author pahio (2872)
Entry type Topic
Classification msc 26A36
Synonym integration of expressions of square roots of quadratic polynomials
Related topic IntegrationOfRationalFunctionOfSineAndCosine
Related topic Tractrix
Related topic Arsinh
Related topic Arcosh
Defines Euler’s substitutions
Defines substitutions of Euler