Intermittent or momentary faults can be very hard to find. On some machines, where possible, I added a latch to the alarm. Other times, you just had to watch for the error. One of the machines that I worked on had an air blow for spindle cleaning during tool change. If the air pressure dropped below the switch setting the machine Emergency-Stopped on low air. As soon as it stopped and shut off the air blow, the pressure built back up above the switch setting, and the error automatically reset . Machine operators were notorious for not paying attention, so the complaint was usually "the machine just stopped". Often the fix was just replacing the air supply hose with a shorter, larger diameter hose.
GlennA, I was working for a company that made simulators at one time. I was involved in a R&D project to track errors and predict faults. It was very interesting and an early application of AI in a real-time system. We acutally logged every fault, even intermittent ones. I learned about this working on satellites, actually. Having an indicator that just blinked is not really useful, you need to keep that information.
TJ McDermott; Yes, too many alarm messages can also be a problem. The ABB IRB 6400 S4c controller had several message lists, and a general list. And some alarms would generate a secondary message, or even a tertiary message; e.g. on a servo fault, each of the 6 axes would reply with a shutdown message. It took some time to sort through the alarms to get to the primary error.
You have to remember that this was around 1968 and control system design was moving from traditional relay technology to use of early (and expensive!) integrated circuits. It just wasn't possible to latch every one of the hundreds of possible fault conditions and with hindsight, use of hard-wired printed circuit logic with IC's (mostly 7400 family in flat-packs) was smaller but no better than the relay version. Only with the arrival of the microprocessor and true Programmable Logic Controller or PLC would all this be resolved sensibly.
I'm surprised this wasn't discovered during testing of the lubrication system. It's not intermittent, and there wasn't anything odd about 11:30, it's just that the lubrication system caused the machine to fail. It makes me think that the lubrication system was tested outside of the machine, or when it was at idle.
Rod, we had a similar experience while working with the rocket launch division. During the testing phase, when time counter is incremented from 23.59 to 00.00, the vernier engines stops fueling the booster rockets and hence there is change in thrust pressure. We had done all sort of testing to find out why this fuel injection missing happens at that particular time and finally found that a small bug in the Micro controller program. According to the program 00.00 is the resetting cycle, so whenever the clock reaches 00.00, it reads as a reset instruction.
Tekochip, it seems that integration testing had not happened. By component wise all the devices or parts may work fine and by integration testing we can find how the system is behaving in a complex atmosphere or in collaborative mode.
Now you have put your finger on the root of the problem! The gearboxes and lubrication systems had been supplied and installed by a sub-contractor and the only "official" interface to our control system was the oil pressure signal. The timeswitch didn't appear on any of our system diagrams and was fed from a spur from the antenna domestic mains, installed on an ad hoc basis by a sub-contract "sparks" at the request of the gearbox supplier. That's how it got as far as it did - we were as surprised as anyone else when we found it. Later earth-stations projects paid much more attention to such integration issues and consequently the commissioning threw up fewer and fewer surprises. We still had some nasty experiences though, one in particular from an air-conditioning system. But that's another story...
Using wireless chips and accessories, engineers can now extract data from the unlikeliest of places -- pumps, motors, bridges, conveyors, refineries, cooling towers, parking garages, down-hole drills and just about anything else that can benefit from monitoring.
With strong marketplace demand for qualified engineers across the board that currently outstrips the available supply, there may never be a better time for engineers and project managers to advance their careers and salaries. Whether those moves are successful in the short-term and long-term is likely to depend on how the transition from one job to the next is handled.
Focus on Fundamentals consists of 45-minute on-line classes that cover a host of technologies. You learn without leaving the comfort of your desk. All classes are taught by subject-matter experts and all are archived. So if you can't attend live, attend at your convenience.