This paper deals with the accuracy of time integration methods for linear d
ynamics when applied near the resonance condition. An approach for the anal
ysis is considered which allows spurious resonance conditions to be detecte
d. The analysis of Newmark methods shows the existence of such conditions w
hich can adversely affect the quality of numerical computations, As an alte
rnative, a higher order algorithm, which can be viewed as a generalization
of the trapezoidal rule, is investigated. The analysis reveals that the spu
rious disturbance near the resonance condition is greatly reduced. The repo
rted numerical tests confirm the theoretical predictions and demonstrate th
at high-quality simulations can be obtained by means of higher order algori
thms. (C) 2000 Academic Press.