Our primary objective with MOSES is to provide engineers with the tools necessary to realistically design and analyze marine structures and operations. The increasing sophistication of the offshore industry, coupled with the rapid evolution of the computational power available, has made it obvious that to achieve this goal, a major departure from the traditional approaches would be necessary.

In the past, a problem was analyzed in several distinct parts - each requiring a different view of reality and subsequent model. While this approach is suited to existing organizational structure, it is highly inefficient and error prone. Different models are ipso facto different. A substantial quality assurance effort has been required to reconcile these differences, and more importantly, one could not hope to obtain a proper analysis of the complete picture by simply viewing selected parts of it. Obviously, what was really necessary was something that integrated all aspects of the problem.

To bridge this gap, we have created MOSES, a new language for modeling, simulating, and analyzing the stresses which arise in marine situations. This new language offers the necessary flexibility along with the rigor of a programming language. Now, one can easily create new models, document them, and assess their validity - all with a single program.

In addition to specialized capabilities, the MOSES language is rich in general utilities to make one's life easier. Most results of a MOSES simulation are available for interactive reporting, graphing, viewing in three dimensions, and statistical interpretation. Instead of manually repeating blocks of data, MOSES provides for loops. Instead of having different sets of data for slightly different situations, MOSES provides for conditional execution. Instead of having the same data defined in different places, MOSES allows one to define variables and use them later. Instead of repeating commands with minor alterations, MOSES allows the user to create his own commands called macros.

The MOSES language is built upon a proprietary database manager specifically designed for its purpose - the storage and retrieval of scientific models and the results of their simulations. By storing all data in a database, MOSES is totally restartable. One can perform some tasks interactively, stop, then seamlessly restart the program to perform other tasks in the background. The database even allows different types of simulation with the same model and a stress analysis to be performed for all types concurrently.

Before MOSES, most marine problems were considered in two steps: a simulation followed by a stress analysis. Two different programs were required. Since MOSES performs both of these analyses, one needs only a single program to investigate all aspects of the problem. Also, with MOSES, one is spared the agony of transferring files and of learning the idiosyncrasies of several programs.

Since it must cope with the demands of both simulation and stress analysis, the MOSES modeling language is richer than the norm. From a stress analysis point of view, a MOSES model consists of a set of beams, generalized plates, and connectors. Here, however, these structural elements can also model load generating attributes. To allow for other types of loads, one can define areas and masses, along with constructs called "hulls". This gives MOSES the ability to compute hydrodynamic forces on a system via three hydrodynamic theories: Morison's Equation, Two Dimensional Diffraction theory, or Three Dimensional Diffraction theory.

With MOSES, connectors are not simply "restraints", but the way one connects different bodies. One can select from catenary mooring lines, tension-only and compression-only nonlinear springs, rigid connectors such as pins and launchways, and even true nonlinear rod elements. These connectors are automatically applied during a stress analysis so that one can correctly perform a stress analysis of several connected bodies.

The MOSES modeling language is rich enough so that models suitable for other programs can be converted to MOSES models with minimal effort. In fact, interfaces are available for several programs, and others can be quickly developed.

Not being content with simply analyzing a given situation, MOSES provides a menu which aids in the design of mooring lines and lifting slings. Commands are also available which will automatically alter connectors so that different scenarios can be assessed with minimal effort. With a rod connector, the effect of the inertia and damping of the connections may be assessed.

As with connectors, MOSES allows for the basic computations traditionally performed by a naval architect. One can compute the curves of form, the intact or damaged stability, and the longitudinal strength of a vessel. MOSES, however, does not stop here. One can specify interactively, the ballast in any or all of the vessel's tanks and immediately find the resulting condition. If one wishes, he can ask MOSES to compute a ballast plan which will achieve a given condition and then alter it. Finally, if desired, one can ask MOSES to perform a detailed stress analysis of the condition. The program will take care of all of the details of computing the correct inertia, loads, and restraints.

Once a suitable condition has been found, a traditional seakeeping study can be performed with MOSES by issuing a single command. MOSES will then use the hydrodynamic theory selected from the three available to compute the response operators of both the motions of each body and the connector forces. An entire menu of commands is available to post-process these response operators. One can easily find the statistical results for specified sea conditions and create time domain samples of the results to assess phasing. All results can be graphed or reported. Only four additional commands are necessary to produce a detailed stress analysis of the system in the frequency domain.

At any point, one may perform a time domain simulation of the current system. This is accomplished by issuing a command to define the environment, and a second to initiate the time domain simulation. MOSES then takes the hydrodynamic forces computed via the proper hydrodynamic theory, combines them with the other forces which act on the system, and integrates the nonlinear equations of motion in the time domain. At the conclusion, again a menu of post-processing commands are available to assist the analyst in deciphering the results - trajectories of points, forces on elements, connector forces, etc. As before, a stress analysis at events during the simulation requires only a few additional commands.

To simulate the process of lifting a structure off of a barge, lowering it into the water, and bringing it upright, MOSES offers a menu of alternatives. One can interactively ballast compartments and move the hook up or down to assess the results of any field action. These results are stored by event so that they can be reviewed and the action changed, until the desired outcome is attained. As with other simulations, at the conclusion, the results can be post-processed and used for a stress analysis.

A specialized type of time domain simulation is a jacket launch. Here, a single body is moved until it comes free of other bodies upon which it was towed to location. Traditionally, a jacket was launched from a single barge. In anticipation of such an operation, MOSES can simulate a launch from several barges which may be connected.

By combining a nonlinear rod element with other connectors, one can simulate the laying of pipe either from a stinger or from davits. With MOSES, all aspects of the problem can be modeled. The lay vessel and the stinger can be modeled as separate bodies connected via the pipe, hinges, tensioners, and rollers. Once the system is assembled, one can perform static, time, or frequency domain simulations of the laying process.

MOSES can perform a detailed stress analysis for events during a time domain simulation, a static process, or a frequency domain process. There are no essential limits on either the model size, the number of bodies which can be analyzed, or the number of load cases. The solution algorithms are state of the art and the structural post-processing is superior. MOSES can consider not only linear but also spectral combinations of the basic load cases. Thus, if one performs a stress analysis in the frequency domain, he can then consider member and joint checks spectrally. In addition, spectral fatigue can be considered in beams, generalized plates, and tubular joints.