3. Explicit Dynamic Time Step Control

This chapter discusses time control for explicit dynamic Sierra.SM analyses. We begin with a broad overview of explicit dynamic time control and then describe the options that are available to users for time control. The commands and features documented in this chapter are applicable only for explicit dynamic analyses. For information on time stepping commands for implicit analyses, see Section 4.

The user initiates time control by setting a start time and a termination time for an analysis. The analysis is typically carried out with a large number of time steps, with each time step being much smaller than the analysis time. In explicit transient dynamics analyses, the time step must be less than a critical value. Sierra/SM has a number of methods for computing an estimate for the critical time step. These methods are discussed in detail in this chapter.

The primary time control uses a TIME CONTROL command block that appears in the procedure scope. Use of the TIME CONTROL command block gives the user, by default, access to an element-based method for estimating the critical time step. The user can access three other methods for estimating the critical time step by using specific command blocks that are placed in the region scope. These other methods tend to give better (larger) estimates for the critical time step. In addition, there is a technique for adjusting the time step known as mass scaling.

Section 3.1 describes the TIME CONTROL command block. In Section 3.2 we discuss the other methods for estimating the critical time step. One approach for improving this estimate is to compute the maximum eigenvalue for a problem. Two methods for computing the maximum eigenvalue are available: the Lanczos method and the power method. Section 3.2.1 discusses the Lanczos method; Section 3.2.1.5 describes the command block required to implement the Lanczos method. Section 3.2.2 discusses the power method; Section 3.2.2.5 describes the command block required to implement the power method. Another approach for improving the time step estimate relies on a node-based estimate. Section 3.2.3 discusses the node-based method; Section 3.2.3.1 describes the command block required to implement the node-based method. You should read the introductory material for the maximum eigenvalue calculation methods and the node-based method and understand this material thoroughly before you attempt to use these methods. Although these other methods give larger time step estimates than the element-based method, they may not result in a net reduction of central processor unit (CPU) time for an analysis unless they are used properly. In those sections dealing with these other methods, we discuss how to use these methods in a cost-effective manner. In Section 3.3 we describe the technique of mass scaling, which is a technique that artificially scales up nodal masses to allow larger time steps to be taken, and must be used with the node-based critical time step estimation method.