April 24, 2001
A common thread runs through all effective troubleshooting approaches: the skill of observation. Learn to use it to your advantage.
A manufacturing process can be defined as altering the configuration or shape of raw material and/or previously manufactured components with a combination of equipment, tools, and operators to create a new product. Over time, any manufacturing process becomes vulnerable to alteration through wear and misuse of the equipment or changes in the raw materials or components used, which ultimately can lead to process failure.
Troubleshooting can be used to determine why a manufacturing process fails to produce the desired result.
The first step in troubleshooting is to become fully familiar with the manufacturing process and the parts it produces. The troubleshooter must know what takes place when the process is functioning correctly. If the process produces an assembly, then all the individual parts that comprise that assembly must be understood clearly.
It is not enough, however, just to know what the end result of a process should be. Each separate operation performed in the process also must be understood thoroughly.
Also critical to the full understanding of a process is the sequence in which each operation is performed. If the operation sequence is not entirely understood, a lot of time will be wasted working on a symptom rather than a cause, yielding unsatisfactory, if not catastrophic, results.
Finally, it is essential that the process variables be understood. All manufacturing processes have variables; some are controlled within the process, and some are not. These variables must be identified for the process being studied:
1. Controlled variables include the capabilities of the tools and equipment used in the process. When these are known, and the process result indicates that the deviations exceed the allowable limits for the tools or equipment, then a possible cause has been found.
2. Uncontrolled variables include raw material and component variations, as well as operator error. Although these variables most often are responsible for creating problems, they are frequently overlooked. Consequently, causes of problems often are diagnosed incorrectly.
The obvious way to become familiar with a process, of course, is to study its design through instructions or design drawings. The other way is to study physically the elements of the process to determine how they all work together.
A common thread runs through all effective troubleshooting approaches: the skill of observation. Some suggest that such a skill is more innate than learned. However, it can be developed. The disciplined observer will learn, through careful study of cause and effect, the relative importance of all features of the process.
For instance, if a police officer and a businessman are standing on a sidewalk and witness a car go through a red light, the police officer likely will be able to identify the make, model, year, color, and state of registration of the car. The businessman, on the other hand, might recall only the color. The police officer's ability to see detail is not innate. The police officer simply is disciplined to recognize the distinctive features of automobiles to the extent that identification can be made at a glance, while the businessman does not have this discipline.
This kind of discipline must be developed for successful troubleshooting, regardless of the methodology employed.
The observation stage of the troubleshooting process should follow a certain sequence:
1. Study the symptoms of the problem. First, the troubleshooter must determine what has changed from before the problem was observed. It is always a temptation, in the interest of speed, to move too quickly through this process. Subtle changes can be just as important as obvious ones to point to the cause of a problem.
2. Determine where a discrepant condition or malfunction has occurred in the process. The problem-causing malfunction often does not happen at the point in the process where the problem first becomes apparent. People can expend great effort making changes to a process where they first see a problem, only to find out that the cause of the problem was earlier in the process.
3. Ensure that all process variables have been taken into consideration. Significant time can be saved when the process variables are considered early in the troubleshooting procedure. Sometimes the problem does not lie with the process, but with minor material discrepancies or components that do not meet the specifications
After becoming familiar with the manufacturing process, observing existing conditions, and noting changes, the troubleshooter can determine the cause of the problem by interpreting the observations that were made. At this point, the troubleshooter can rely on his own experience for determining the necessary remedial action or consult an expert.
Heeding the following advice, and thus avoiding the temptation to take shortcuts that could result in errors, can help a troubleshooter be successful:
Troubleshooting a stamping process should not be confused with the simple process of removing a damaged or broken die from the press and repairing it. Often, stampers repair dies to compensate for the problem, rather than determining and correcting the underlying cause of the breakage. A stamping shop that does not follow the principles of troubleshooting in its die maintenance program will find that die life, productivity, and part quality are reduced, while die maintenance costs are increased.
To showcase the principles outlined in this article, following is an example of an actual progressive die problem that occurred in the author's shop.
1. Problem identification. After the shop ran about 100,000 in-tolerance parts as described in the following paragraph, the flatness error began to exceed the tolerance, and the out-of-flat condition caused a problem with part ejection.
2. Process and part understanding. Following is the type of information that must be gathered to become adequately knowledgeable about a process and part.
The part was a faceplate of an electronic data networking device. It measured about 2 by 12 inches and was made from 0.030-inch-thick cold-rolled steel. Several dimensions had a tolerance of ±0.005 inch. The part had several forms, two of which were very narrow tabs bent at 90-degree angles to the main surface of the part. Flatness in the 12-inch direction had to be within 0.015 inch.
The die had 12 stations and produced five different parts, all with common major features. Those features that were unique to each part could be gagged in or out so that only one set of unique features were active at a time. The die had a feed length sensor and a precision-guided spring stripper and performed several operations: cutting, bending, embossing, offset forming, and coining.
As the material progressed through the various stations, excess material was cut away first, except for the material needed to hold the part in the strip. Toward the end of the die, the forming operations were performed, requiring minimal clearing of die blocks for the formed features. The final station cut material from each end of the part, separating the part from the strip.
3. Process observation. The out-of-flat condition was found during the customary periodic part inspection. Preliminary inspection of the die and strip revealed no obvious problem. A comparison of an earlier strip to the problem strip also did not show any difference. Further inspection of the die found nothing worn or broken.
A more careful inspection of the strip, however, revealed that the bow condition began to be apparent in the eighth station and got slightly worse in the subsequent three stations. One of the forms appeared to be catching on the leading edge of the next die block as the strip was fed forward, which, in turn, induced a bow in the part just before part separation.
The obvious solution seemed to be to lower the die block on which the formed tab was catching. The observation process needed to be completed, however, before any action was taken. The reason that the out-of-flat condition did not present itself in the first run of 100,000 parts still had to be investigated.
4. Process variable review. The earlier strip was found to be flat to the last station. At this point, all process variables were reviewed. The material was checked for coil condition, width, and thickness. The setup of the press line, from coil dereeler through the straightener and the feed and into the die, also was checked.
Everything appeared in order except for two minor conditions that did not seem to be significant at first: The material strip had a slight camber, and the strip was rubbing the back stock rail at the entry into the die and also rubbing the front stock rail at the exit of the die. The slight pressure exerted on the edge of the strip caused the part to bow enough so that the formed tab stumbled on the die block, causing the bow to be increased enough to produce an unacceptable part.
A closer investigation of the setup revealed a slight misalignment of the die to the material strip. The combination of the slight camber in the material and the slight misalignment of the die was what had changed in the process from the previous run and was sufficient to produce the unacceptable parts. The die was realigned and the problem was solved.
This example demonstrates the importance of thorough observation. It also shows how important it is to avoid drawing premature conclusions. As is often the case, if action had been taken only on the obvious problem, and the observation process had not been completed, the real problem would not have been found, and die damage could have resulted.