Read the pages and make abrief summary of them with your own words, please. It is what you understand. Don't make copy-paste, please. Mention important parts only. Also, you will put your comments and ideas about the topic. Please don't write item by item. Write the summary in paragraph form.
Extracted text: Failure Mode and Effects Analysis way, the designers will be able to develop a design that has fewer potential failures, and those that cannot be avoided can be made less severe. Also, by using FMEA concurrently with the design activity, it is more likely that test and inspec- tion methods will be able to catch the problems before they Design FMEA. FMEA applied during the design phase of a product or service to ensure that potential failure modes of the new product or service have been addressed. Failure mode and effects analysis (FMEA) tries to identify all possible potential failures of a product or process, priori- tize them according to their risk, and set in motion action to eliminate or reduce the probability of their occurrence. FMEA cannot by itself bring about this happy ending, since it is an an- alytical tool, not a problem solver. But it will point to the prob- lems that must be solved through the use of the other tools. Failure mode and effects analysis-the name itself is enough to scare off the unfamiliar. So you don't give up on FMEA before we get into it, let's simplify the concept. FMEA just tries to identify all the possible types (modes) of failures that could happen to a product or a process before they hap- pen. Once the possible "failure modes" have been identified, the "effects analysis" kicks in and studies the potential con- sequences of those failures. Next, the consequences of each potential failure are ranked by Process FMEA. FMEA applied to a process (as in a factory or office) to ensure that potential failure modes of the process have been addressed. get to the customer. A second version is process FMEA. In this case, FMEA is looking at the potential failures (errors, miscues) of a pro- cess. The process might be that of an accounting firm, a hos- pital, a factory, a governmental agency, or any other entity. One can imagine that in a hospital there are many processes Risk assessment factors. . Severity (S): A number from 1 to 10, depending on the severity of the potential failure mode's effect: 1 = no effect, 10=maximum severity. that can have lots failure modes, some probably not too important, but some as severe as they come. One would hope that FMEA is in every hospitaľ's tool kit. Ford Motor Company uses FMEA even before it gets to the design stage of a vehicle. As the concept for a new vehicle is being developed, FMEA is employed to make sure that the vehicle will not bring problems related to the concept into the design and production stages. FMEA can also be used after the fact (as in the case of a product repeatedly failing in the hands of the customer). This may lead to a retrofit or recall of the product if the problem is severe or simply to a design change for future production if the problem is not critical. The procedure is essentially the same for every kind of FMEA. FMEA is not new, although until recently its use was mainly associated with military and aerospace programs. It was developed by the U.S. military in 1949 and has seen increasing use in industry, especially since the 1980s, its importance being driven by the worldwide quality movement under TQM and ISO 9000 and by litigation in the United States against com- panies whose products are involved in customer injuries or deaths. FMEA is now considered an invaluable quality tool. • Probability of occurrence (O): A number from 1 to 10, depending on the likelihood of the failure mode's occur- very unlikely to occur, 10 = almost certain to rence: 1 occur. • Probability of detection (D): A number from 1 to 10, depending on how unlikely it is that the fault will be de- tected by the system responsible (design control process, quality testing, etc.): 1 impossible to detect. • Risk Priority Number (RPN): The failure mode's risk is found by the formula RPN = SXO X D. Said another way, RPN = Severity x Probability of Occurrence X Probability of Detection. RPN will be a number between 1 (virtually no risk) and 1,000 (extreme risk). The auto industry considers an RPN of 75 to be ac- ceptable, although in light of some recent manufacturer recalls for safety-related failures, we anticipate that may change. Seriousness/Criticality to the customer nearly certain detection, 10 = Probability of the fault's occurrence Probability of the fault's detection by the systems re- sponsible for defect prevention or detection Seriousness of consequence, likelihood of occurrence, and difficulty of detection all work together to determine the criticality of any specific failure mode. Comparing the criticality of all the identified potential failure modes estab- lishes the priority for corrective action. That is the objective of FMEA. FMEA tells the organization where its resources should be applied, and this is very important because all possible failures are not equal and the organization should always deploy its resources to correct the problems that are most critical. Without the benefit of FMEA, it is doubtful that an organization could identify its most critical failure modes very accurately. Remember, usually FMEA addresses prob- lems that have not yet happened. Next time you are cruising at 600 miles per hour at 35,000 feet, consider whether the designers of your airliner should have used FMEA-or the next time you really, really need your brakes to work (Am I going to go over the cliff?) or when you buy that new $2,000 high-definition TV (So much technology-is it going to be reliable?). We might also consider it if we have to go to a hospital-or when we ship our original, no-copy-available manuscript to our publisher by overnight express. Looking at it from another viewpoint, had FMEA been available, could it have prevented the Titanic's disaster? Given what we now know about the ship's collision with the iceberg, we are convinced that had FMEA been employed, the Titanic might have plied the seas through most of the twentieth century. Of course, FMEA did not come along until four decades later. There are several kinds of FMEA. Design FMEA is employed during the design phase of a product or service, hopefully starting at the very beginning of the project. In this FMEA Illustration Let's consider a simplified FMEA to illustrate how the process works. We will assume we manu- facture bicycles and we are designing a new bike that will be made largely of composite materials. Since this is a new technology for our company, we are using design FMEA to make sure we've considered all the possible problem areas of the design before we go into production. The FMEA team has listed several potential failure modes, one involving sud- den, unwarned breakage of the front fork. It is obvious that should the fork fail, the effect on the customer could be se- The Language of FMEA FMEA has its own unique set of terms. We have captured most of them in the following list: Failure mode. The way in which something might fail. For example, a race car's tire might fail by puncture from a sharp object. It might also fail from a blowout resulting from wear. Puncture and blowout are two (of many) tire failure modes. Failure effect. The failure's consequence in terms of operation, function, or status of the item. Effects analysis. Studying the consequences of the various failure modes to determine their severity to the customer. Of the two tire failure modes mentioned earlier, the blowout is likely to have the most serious vere. Since the rider will probably have no warning before the fork breaks, we rate the severity a 10 (S = 10). We then identify possible causes of the fork failure and conclude that the probability of the most likely cause oc- curring is moderate, with five occurrences per thousand bikes. We rate the probability of occurrence a 6 (O = 6). Of consequence, since when a tire suddenly explodes, the speeding race car usually goes out of control, often with dire consequences. On the other hand, a puncture usu- ally allows the tire pressure to decrease gradually, allow- ing the driver time to sense the problem before he or she loses control. Neither failure mode is something the driver wants, but of the two, the puncture is preferred. course, it is our intention to detect the defective forks and discard them before they are attached to the bicycle fra After we examine our fork testing methods, we conclude that the probability of detecting the failure mode flaw in the fork is low. We assign it a 6 (D = 6). Plugging these numbers RPN = SX O X D, we have into our equation, Failure mode analysis (FMA). An analytical tech- nique used to evaluate failure modes with the intent to eliminate the failure mode in future operations. RPN = 10 x 6 X 6 = 360
Extracted text: function because the total quality envelope is constantly being expanded and there will always be the need for a few to be on the leading edge and to bring the others along. It is management's responsibility to ensure that the people who are solving the problems have the proper train- ing and facilitation. It is also management's responsibility to make sure the problems being attacked are of interest to the enterprise and not trivial. Management must populate the problem-solving team with the cross-functional expertise the problem requires. The team must be given the power and support necessary to see the effort brought to its conclusion, Management must be vigilant that data used in problem solving are valid, which is a function that usually falls to the fa- cilitator. Especially when teams are immature in total quality, they have a tendency to grab at the first set of data that comes along. Management must ensure that the data and the statistical techniques employed are appropriate for the problem at hand. Finally, management must ensure that there are results. Too many problem-solving, process improvement, and re- lated efforts take on a life of their own and go on forever. This cannot be allowed. People are watching. Especially in the early stages, some people will hold the view that "This shall pass." If results do not come rather quickly, the detrac- tors will be given the ammunition they need to subvert the whole total quality effort. For this reason, it is important that the first projects attempted have a high probability of success, and management must monitor them closely, even to the point of being involved in the activity. As the process matures and successes are tallied, an occasional failure will not be an All other failure modes result in RPNS in the range of 40 to 70, so our focus should be on eliminating, or drastically reducing, this potential fork failure mode. We could redesign the fork so that it is more robust, thereby lowering the occur- rence value (O), or change the test process so that it is much more likely to detect a fork that might fail, thereby lowering the detection value (D). Notice that if our fork testing process gave us complete assurance of detecting the fault-say, at the D 1 level- RPN would be 60, and we probably wouldn't need to put a lot of resources on this fault mode. The same could be said of process experimentation allows multiple factor adjustment simultaneously, shortening the total process, but equally as im- but also they embraced them and their philosophy (whereas portant, revealing complex interaction among the factors. A in the United States we were abandoning their teaching amid well-designed experiment can be concluded on a process such as wave soldering in 30 to 40 runs and will establish the opti- Japan developed its own quality gurus (Ishikawa, Taguchi, mum setting for each of the adjustable parameters for each of Shingo, and others) who expanded the work of Deming and the selected factors. For example, optimal settings for conveyor Juran. For 30 years, into the 1980s, Japanese manufactur- speed, conveyor angle, wave height, preheat temperature, sol- ers perfected their quality and production methods. The der temperature, and flux specific gravity will be established 1980s found Japan ahead of the rest of the world, not just for each PC board type, solder alloy, and so on. DOE also shows which factors are critical and which are decade, companies in the United States began to wake up not. This information enables you to set up control charts for to the fact that Japan's products were the best in the world those factors that matter, while saving the effort that might and that they were running roughshod over U.S. companies have been expended on the ones that don't. While design of not only in the world markets, but also right here at home. experiments is beyond the scope and intent of this book, the Whole markets were conceded to Japan as U.S. companies DOE work of Deming, Taguchi, and others may be of help found they could not compete. to you. Remember that DOE is available as a tool when you start trying to improve a complex process. Not only did the Japanese listen to Deming and Juran, a seemingly insatiable market for manufactured goods). the United States, in product quality and value. During that if D remained at 6, but the probability of occurrence of the fault mode turned out to be remote (O = 1). When to Use FMEA FMEA should be employed at the following points: During the design or redesign of a process, product, or service The survival mentality finally surfaced. We woke up to the fact that not only our industrial survival but perhaps even our national survival was at stake. Either we became When improvements are needed or planned for existing processes, products, or services When existing processes, products, or services are to be used in a new way competitive in the global marketplace, or we lost the first war fought without bullets since the invention of gunpowder. Now that the wake-up call has been received, many peo- MANAGEMENT'S ROLE IN TOOL DEPLOYMENT too Management's role is changing from one of directing to one of ple have come to realize that we have been managing poorly facilitating. Since the Industrial Revolution, management has for a very long time-say, since 1945. We (those of us who supplied the place of work, the machinery and tools, and the have heard the alarm) have come to understand that manage- work instructions. The concept has been that management ment's proper role is to facilitate, not to direct. Management knows what the job is and needs only to hire the muscle power provides the place of work and the machines and tools as to get it accomplished. The workers were there only because before, but in addition, management could not get the job done without their labor. employees do the job. That means training. It means listen- Workers were not expected to think about doing things differ- ing to their thoughts and ideas-more than that, it means ently but simply to follow the boss's orders. Work was typically seeking their thoughts and ideas. It means acting on them. It divided into small tasks that required minimal training, with means giving them the power to do their jobs without man- little or no understanding on the part of laborers as to how their contribution fit into the mosaic of the whole. During much of the twentieth century, and certainly It means accepting every employee as a valued member of since World War II, changes have been creeping into the management-labor relationship. Some people think that the the corporate team. labor unions were responsible for these changes, and they did help obtain better pay, shorter hours, workplace improve- cates its responsibility to set the direction for the enterprise, ments, and other benefits for workers. However, the relation- ship changes between management and labor have happened the enlistment of all the brain power that had formerly gone largely in spite of the unions. Unions have had at least as dif- untapped, even this job becomes easier than it was before. ficult a time as management has had in dealing with employee involvement. Nor has management at large been responsible also intellectual tools. The tools discussed in this chapter for the changes sweeping across the industrial world today. should eventually be used by most employees-eventually Certainly, there are champions representing management, but the changes are coming about for one reason and one reason only: They are necessary in order for businesses to survive in an increasingly competitive marketplace. After World War II, when Deming went to Japan to teach industrialists about quality and the use of statistics for tools. When a group of people is ready to put some of the achieving it, Japan had just lost the war. Its industrial base tools into practice, that is when the group should be trained. was a shambles. The Japanese needed to resurrect their fac- tories and put people to work quickly. That meant they had of time until everyone has the need. Train them as required. to be able to sell their products abroad-to the same people who had defeated them. To do that, it was essential that their products be of high quality. Their survival depended on it. velop their expertise. Facilitation is probably a never-ending You know the rest of the story. During after-the-fact failure analysis When safety or health is an issue This is intended to be a brief introduction to FMEA. do everything we can to help our Going into it more thoroughly is beyond the scope of this text. Should you find that you need more, the Internet is a good source of information, and there are many books dedi- cated to the subject. issue. In fact, people must be given the chance to fail, and fail- ure must be free of repercussions for the team or its members. agement interference. It means giving them time to think and discuss and suggest and experiment. It means commu- nicating-fully and honestly. No secrets, no smoke screens. Design of Experiments Design of experiments (DOE) is a very sophisticated method for experimenting with processes with the objective of op- timizing them. If you deal with complicated processes that have multiple factors affecting them, DOE may be the only practical way of bringing about improvement. For example, such a process might be found in a wave soldering machine. Wave solder process factors include the following: This approach does not mean that management abdi- to establish the corporate vision, to steer the course. But with Solder type Flux specific gravity Conveyor angle Conveyor speed Solder temperature It is management's responsibility use not only physical tools (and that is very important) but train employees to Wave height Preheat temperature PC board layer count because it is a mistake to schedule all employees for training on the tools if they will not be using them very soon. You would not train a person on a new machine a year before the machine arrives because without putting the training to practice, its effect will be lost. So it is with the total quality Flux type PC board groundplane mass These 10 factors influence the process, often interacting with one another. The traditional way to determine the proper selection or setting was to vary one factor while holding all others fixed. That kind of experimentation led to making hun- dreds of individual runs for even the simplest processes. With that approach, it is unusual to arrive at the optimum setup be- cause a change in one factor frequently requires adjustment of one or more of the other factors for best results. The DOE method reduces the number of runs from hun- As the total quality concept takes root, it will be only a matter Management must also provide the internal experts, often called facilitators, to help the new teams get started and to de- dreds to tens as a rule, or by an order of magnitude. This means