The Lie analysis and solutions for a class of second-order nonlinear ordinary differential equations (Q1321902)

The authors investigate the nonlinear ordinary differential equation \(\ddot s+ p\dot s+ (1-\mu)r(t)s= (1-\mu)^ \sigma f(t)s^ \sigma\), where \(s\) is the dependent variable, \(t\) is the independent variable, \(p\), \(r\), \(f\) are given functions of \(t\), and \(\sigma= (n-\mu)/(1-\mu)\), \(n\) and \(\mu\) are constants, to find cases which are exactly solvable (i.e. by quadrature). The main approach of the paper is to apply the theory of Lie symmetries. The authors recast the above equation in various guises (by point transformations) to facilitate this approach. Unfortunately, the paper contains many misprints. For example, in the right side of equation (2.3), \(1-\mu\) should be raised to the power \((n- \mu)/(1-\mu)\), (not \((n+ 1)/(1-\mu)\), similarly in equation (2.6), the power of \(1-\mu\) should be \(\sigma\), the term inside the exponential in equation (2.10) should read \(-\int p(t)dt\), in equation (3.5) the variable \(N\) should be replaced by \(\mu\) (or \(\mu\) in (3.3) replaced by \(N\)), in equation (3.6) only ordinary derivatives of \(\xi\) and \(\eta\) are implied (by the primes), whereas equation (3.4) defines them to be functions of two variables. Furthermore, the quadratures given in equations (4.9) and (4.18) are not valid for the case \(\sigma=-1\). In some cases, the authors are able to find one symmetry but are not able to integrate the equation.