Although it is a lot simpler and less costly to use "go or no-go" criteria for inspection, vital information about the manufacturing process is lost. There are some items that lend themselves to this type of inspection, but these are normally attribute type features where only two choices are available. For example, has the part been part marked correctly? This is a standard binary question for which a “yes/no” is perfectly adequate. This is not the case with a feature that has a specific tolerance i.e.: a nominal and an upper and lower spec limit (1.0 +/- .005). It might be tempting to put a quality stamp on a feature if it is between these ranges without recording the measurement, but if this is done, the company loses vital manufacturing and quality information. This information accurately collected is the only basis of implementing quality improvement. The more granular the information, the more effective it can become.
If the only options are in-tolerance and out-of-tolerance, quality can degrade gradually. There is no ability to determine deteriorating quality until it is too late. This situation was dramatically highlighted by the Alaska Airlines Flight 261 accident in early 2000. Poor maintenance (lack of appropriate lubrication) had resulted in excessive wear on a jack screw that was the single point of failure for the elevator system. When the part was being reviewed by maintenance, it was initially determined it was out-of-tolerance and should not be put back on the aircraft. Because of cost considerations, the part was re-measured by a supervisor who determined it was within tolerance by a few tenths of an inch. A classic go or no-go measurement process. This made the reinstallation of this part perfectly legal but not wise. Common sense would have shown the wear would continue, which it did, until it finally completely failed mid-flight resulting in a complete loss of control of the aircraft and the loss of lives for all the passengers.
Accurate and consistent measurement techniques may sound like an obvious requirement but these all too frequently compete with the need to ship and in many cases, take second place to a quick go or no-go inspection. The value in collecting accurate and consistent measurements and analyzing them in real time is an invaluable tool for not only process improvement, but also maintaining the improved process. Frequently, when a six sigma process improvement is initiated, improvements are quickly found and the corrective action is put in place. However, many process improvements are elastic and, unless rigorously monitored, drift back to their original state. Determining process capability requires measurements to be accurately and consistently measured. The more automated the process is, the less likely there is to be variation caused by the inspection process, which can hide or distort what is actually happening to the manufacturing process.
With go or no-go, incredibly valuable manufacturing information is lost forever. Drift in process is completely hidden until the part is finally out of tolerance and scrapped. Historical trends and process variance is also lost, this is particularly important in aerospace because the batch quantities are frequently small, but the duration of making the part can be as long as 20 years.
What are the steps a company should follow to take advantage of the wealth of data acquired during the inspection process?
- Review all inspection plans to make sure that go or no-go inspections are reserved for true attribute features for which there is either no measurement is available. For example, has varnish been applied to the part mark to protect it? The answer to this is either yes or no.
- Review all places where go or no-go pins and gauges are used. Not only is this process somewhat subjective but as discussed above the actual measurement data is either nor retained or its granularity is lost due to the limited nature of the options.
- Check that the same measuring instruments are used for similar features on all parts. This will allow an analysis of process as well as a specific feature.
- Review all measuring devices to see whether there is a wireless attachment or a wireless system built into the tool. This way, measurements can be recorded accurately without having to be keyed in.
- Look at using CMMs or a comparator device like the Renishaw Equator to capture data without any chance of someone putting their foot on the scale to achieve favorable results.
- Have the machinists monitor any significant variance in the process in order to eliminate drift in process. Remember, if the operators just put in machine offsets you have changed the process. This might hide a serious defect in the process and impact obtaining information to accurately determine process stability and tool life management.
By implementing these steps to gain accurate data over a long period, you can gain a wealth of analysis about the capability of tools, equipment and processes that would have not been available if the measurement process was not standardized and the measurements were not stored for future analysis. Net-Inspect provides a full quality improvement system that captures and stores all information associated with a single measurement result: specification, machine, operator, measurement tool, and more. Net-Inspect also provides advanced analytics to analyze the sometimes millions of data.
For more information, contact our sales team.