When simulating the behavior of a mechanical system, the time evolution of the generalized coordinates used to represent the configuration of the model is computed as the solution of a combined set of ordinary differential and algebraic equations (DAEs). There are several ways in which the numerical solution of the resulting index 3 DAE problem can be approached. The most well-known and time-honored algorithms are the direct discretization approach, and the state-space reduction approach, respectively. In the latter, the problem is reduced to a minimal set of potentially new generalized coordinates in which the problem assumes the form of a pure second order set of Ordinary Differential Equations (ODE). This approach is very accurate, but computationally intensive, especially when dealing with large mechanical systems that contain flexible parts, stiff components, and contact/impact. The direct discretization approach is less but nevertheless sufficiently accurate yet significantly faster, and it is the approach that is considered in this paper. In the context of direct discretization methods, approaches based on the Backward Differentiation Formulas (BDF) have been the traditional choice for more than 20 years. This paper proposes a new approach in which BDF methods are replaced by the Newmark formulas. Local convergence analysis is carried out for the proposed method, and step-size control, error estimation, and nonlinear system solution related issues are discussed in detail. A series of two simple models are used to validate the method. The global convergence analysis and a computational-efficiency comparison with the most widely used numerical integrator available in the MSC.ADAMS commercial simulation package are forthcoming. The new method has been implemented successfully for industrial strength Dynamic Analysis simulations in the 2005 version of the MSC.ADAMS software and used very effectively for the simulation of systems with more than 15,000 differential-algebraic equations.