Vibration Diagnostics System
VDS by Crystal Instruments
The Vibration Diagnostic System (VDS) is a vibration data management system designed specifically for the Machinery Predictive Maintenance (PdM) community. It harnesses the graphic display capabilities of Crystal Instruments EDM Software for the work of machinery vibration analysts. It allows the user to quickly get to the data for a machine of interest and display that data in the familiar Tri-axial or Single Axis view. It lets users quickly compare to other data from the same machine, quickly navigate back into the historical data of the machine, and quickly compare the data to that of other machines in the database. The software supports the construction of Average (a.k.a. baseline) data for a class of machines and allows easy comparison to that data as well. When users interact with the data they will have access to a full suite of cursors designed specifically for PdM analysis. Model your machines with our machine components. We have over 60 common components that you can use to model your machines.
In addition to great graphics, Crystal Instruments has developed an extensible machine modeling system specifically for vibration analysis. It allows users to model machines based on the elements of the machine that can contribute to the vibration energy of the whole. Model bearings, rotors such as motor bars, couplings such as flexible and fluid couplings, account for slip in things like fluid couplings, model gears and pulley systems, and model turbines by accounting for each stage. It’s wide open, users can create new elements to use in their system.
Not all users may want to do all that, and for those users we’re including a comprehensive library of machine components such as AC and Induction Motors, Couplings, Gears, Pulleys, Pumps, and other components. With our single and double gear shafts, users will be able to model many kinds of gear boxes with an unlimited number of stages. This ability alone lets you model your machines to view and use to keep track of vibration pickup locations, record the forcing frequencies of each part of the system, and attach attributes such as manufacturer and other part information.
Machinery Modeling System
The Machinery Modeling system provides the structure for our narrow band automated diagnostic system. VDS provides the following features in support of an automated diagnosis system:
The Machine Modeling System, which allows us to model machine components and to know the relative rotational speeds of all the shafts of the system.
A method to organize Machine Class Average (baseline) data for each designated pickup location.
Associate a physical machine with a Machine Class.
A diagnostic rule processing system based on a forward chaining, probabilistic, inference engine.
A method to define machinery faults.
A method to define recommended actions based on recognized faults.
Provide functions to support basic vibration analysis that takes care of all the mundane details such as extracting dominant peaks from each data set, matching the extracted peaks to forcing frequencies defined by the Machine model, comparing individual datasets with the appropriate Average data and returning the difference in amplitude between the two.
In essence, VDS takes care of all the math and accounting to let the rule writer focus on applying their knowledge of machine condition analysis.
Powered by Artificial Intelligence
We know that most analysts don’t have the time or desire to write their own rules for identifying machinery faults so we hired an expert to write a comprehensive set of rules for all of the machine components in our component library. When you model your machines with our component library you will get the benefit of Artificial Intelligence (AI) that will find many common machine problems.
Because our inference engine is based on a well-known open source scripting language called Lua, those that do want to build their own rule base or add to our rules will be able to do so. Using our low level, documented API, and open source tools, anyone could attempt to build or add to a knowledge base. But so as not to disappoint, keep in mind that this is not for the faint of heart, it can be a challenging task. Entities that may want to do this include:
Predicative Maintenance organizations that specialize in a certain kind of machine and have years of detailed knowledge about these machines, could encode their knowledge with rules, thus extending our system to meet their needs.
University Engineering departments may want to work with our system to teach and experiment with machine vibration analysis.