MIDAS Thermal Analysis (MTA) and Predictive Maintenance
We hear the term Predictive Maintenance talked about all the time from many different manufacturers, but what does it really mean?
Does it mean that they can tell us when something is going to break? Does it mean that they’ll watch the calendar and tell us when it’s time to do some procedure? Or maybe it means that they have a problem on other equipment and you’ll probably get it too.
Since it release, the MTA program has been having some success doing predictive maintenance with Thermal Trending. It has been possible to locate incorrect and defective components by plotting individual machines against their past behavior and even better against fleets of the same machine.
The built in protections on machines will tell you when it’s already a problem (i.e. > component specification), however, to know when a component is running at the correct temperature is a more complicated situation. It often depends on the ambient and the exact machine operation for that time period.
The MIDAS approach is to allow visual and statistical comparisons between the machine on different days and also comparisons against similar machines, short term, log term, and/or statistical.
As with any new application, users have reported certain bugs and features that are needed in the program. Here is the list, all slated to be fixed in the next release:
- When the database gets large, it takes too long to run “Long Term Thermal Trending”
- When the database gets large, the report function sometimes bombs the program.
- The working copies sometimes don’t delete, seems to have something to do with how I exit the program.
If you have others, please report them to MUG so they can be included in the next release.