Originally posted by mazza_728
The partial fractions chapter in fitz isnt that bad i thought but these partial fractions are impossible! e.g. Int 2t.dt/((t<sup>2</sup>+t+1)(1+t<sup>2</sup>))
Both t<sup>2</sup> + t + 1 and t<sup>2</sup> + 1 are irreducibile quadratics, so to do the partial fractions we need to find constants A, B, C and D such that:
2t / [(t<sup>2</sup> + t + 1)(t<sup>2</sup> + 1)] = [(At + B) / (t<sup>2</sup> + t + 1)] + [(Ct + D) / (t<sup>2</sup> + 1)]
Eliminating fractions, we have: 2t = (At + B)(t<sup>2</sup> + 1) + (Ct + D)(t<sup>2</sup> + t + 1) _____ (*)
Taking t = 0 shows that B + D = 0 _____ (1)
Equating coefficients of t<sup>3</sup> show that A + C = 0 _____ (2)
Equating coefficients of t<sup>2</sup> shows that 0 = B + D + C _____ (3)
Putting (1) into (3) gives C = 0, which can be put into (2) to give A = 0
So, (*) has become: 2t = B(t<sup>2</sup> + 1) + D(t<sup>2</sup> + t + 1)
Equating coefficients of t shows that D = 2, and using (1) we get B = -2.
So, 2t / [(t<sup>2</sup> + t + 1)(t<sup>2</sup> + 1)] = [-2 / (t<sup>2</sup> + t + 1)] + [2 / (t<sup>2</sup> + 1)]
The other one would be similar.