The Dynamic Flowgraph Methodology (DFM)
Introduction to DFM
The DFM models and analytical techniques have been developed and perfected by ASCA over a period of several years. They have been successfully applied in projects ranging across the fields of nuclear and space systems risk and reliability analysis, software risk assessment, and dynamic control system V&V. DFM models are multi-valued / discrete-time logic-dynamic models that describe a system and its behavior via discretized representation of system component cause-effect and time interactions. They are laid out and visualized in directed graph (“digraph”) form, and can be analyzed deductively or inductively by automated algorithms. These analyses provide Fault Tree Analysis (FTA)-like output when run in deductive mode, and Failure Modes and Effects Analysis (FMEA)-like output when run in inductive mode. DFM is substantially more advanced and capable than traditional FTA and FMEA in the following aspects:
– It describes in one model both system nominal behavior and any faulted behavior resulting from identifiable component failure-modes, not just system failure events, each represented in a separate model, or model page, as FTA or FMEA; this results in a more efficient and compact format of system visualization and data storage.
– It is function and physical parameter oriented, and, thanks to this and the use of multi-valued time-dependent logic instead of binary static logic, provides greater granularity and fidelity of system representation.