The simulation model supported by a Petri net description of the degradation process has been applied to the case study presented above, with the same parameter settings as in. It is noted that in their formulation, the transition rates, and are time-dependent. The detailed information about the definitions of the transition rates can be found in.
The Petri net representing the multi-state physics model of the crack growth process, is given in Fig. The circumferential crack grows relatively evenly around the circumference, potentially leading to a rupture. The radial crack mainly grows outward from the initiation site towards the outer diameter the process can lead to a leak and potentially to rupture. The crack growth has two steps (1) crack initiation, (2) crack propagation. The latter two types can lead to the rupture of the component.
Cracks can grow from the inner to the outer diameter of the dissimilar metal welds in one of the three major morphologies: axial, radial, and circumferential. The case study refers to the cracking process in an Alloy 82/182 dissimilar metal weld in a primary coolant system of a nuclear power plant. It is noted that the derived distributions and, may have complicated mathematical expressions under these circumstances, the Markov Chain Monte Carlo technique can be used to sample random values. to the execution of the simulation algorithm, an estimate of the state probability vector is computed by dividing the total number of visits to each state by the total number of simulations :, where is the total number of visits to state i at time t. But Simio has another trick up its sleeve, using the Input Parameters tab and defining the distributions with the sample size, provides a neat way to analyse the impact the distribution has on specific result parameters. Of course with distributions come replications, which using the Experiments function allow the analyst to create and execute scenarios using different user defined input parameters and compare results based on user defined KPIs. Note to developers, it would be great if intellitype is available in all fields. Simio has the standard list of distributions which using a form of intellitype is easy to define. The power of discrete event simulation is its ability to handle random events. Gaining model acceptance is generally easier with collaboration, so showing the team their input has been incorporated in a way that resembles the conceptual model goes a long way. Of course the model is an object that can be imported into another model. The object orientated approach means that these basic objects can be modified by adding more detailed lower level processes or modifying original logic. Simio supports this basic concept in its standard library of objects. This boils down to possible resources and a general set of inputs that trigger delays and produce outputs. The start of any simulation project should start with the agreed conceptual model.
This simulation engine seems to have been developed with not only the ability to quickly model with little to no code writing experience, but also empowers one to devise and quickly run scenarios that can answer the real questions and find those sensitive areas/ optimal solutions. Developers please add STP or IGES file format. Add Google Warehouse and importing 3DSMax files and decent visuals can be shown. Simio like most major simulation engines allows 3D visualisation and what is nice is a simple press of the '2' and '3' keys smoothly transitions you from 2D to 3D respectively. Actually, there is another way, demonstrate that all scenarios have been considered and with an inflated chest lay down the challenge, 'tell me where I have gone wrong and I'll tell you what data I need to determine whether its fact or fiction'.
After all, we need to convince the stakeholders that the solution we propose will deliver what it says on the tin and what better way than to mesmerise them with animation. Overall: It is very easy when building simulation models to focus on the visuals and forget the analysis.