Table of Contents
17. Quality Assurance
What is quality? Quality is the totality of characteristics and properties of a product or task that are related to its suitability to fulfill prescribed requirements (Ref. 17.1-4). In other words: quality is determined by the agreement of “is” and “should be”. If one assumes that quality is primarily the result of skilled work, an optimal environment is a prerequisite.
In engine technology, the operating safety of the part is central. For example, the failure probability of a rotor disk due to fracturing must be 10-9 per flight hour to ensure that the probability of a flight accident is less than 10-7 (Volume 1, Ill. 2-4).
Relative to safety, other important aspects such as low scrap rates (i.e. low production costs, prestige, and reliable delivery) are secondary, although they are important for the survival of a company. The flaw probability in the production process is considerably greater than in the finished, installed part. This difference is ensured by suitable measures of quality control. Turbine engine parts are extremely costly. For example, a compressor housing or turbine disk can have the same value as several mid-sized cars.
In accordance with previous considerations, this chapter is primarily concerned with the prevention of safety-relevant production problems. At the same time, the primary focus will be on problems that have already caused parts to fail, albeit in very rare instances.
This means that the quality requirements for engine parts, and therefore especially for the finishing processes and control measures such as non-destructive testing, are highly demanding. The sequence and entire environment of the production process must be optimally designed as a prerequisite for the quality required of turbine engines. Various strategies with systematic approaches have been developed for this (Refs. 17.1-1 Refs. 17.1-2).
With regard to the flawlessness and safety of new parts, it must be remembered that quality ultimately depends on the persons involved in the finishing process, i.e. the human factor (Fig. "Failure mode and effects analysis (FMEA)").
Figure "Acceptable failure probability": The safety of an engine is the product of the failure probability of its individual parts. Therefore, the safety of the individual engine parts must be very high in order to ensure the safety of the entire engine (Volume 1, Chapter 2). If, for example, a compressor disk has a flaw several tenths of a milimeter in size in a critical area, it can result in a crack that will shorten its operating life considerably. Because there are very different types of flaws that are caused by different production steps (raw part production, machining, heat treatment, and handling), each of these flaws must be prevented with an even greater degree of reliability than is required for the part as a whole (Fig. "Segregation in forged disk").
3 ppm means that 3 flaws occur per one million parts. This means, for example, that on a grain field measuring 30×30 meters only a single seed grain would be allowed to fail to grow.
These safety levels can only be expected from a combination of tested and stable production processes and suitably quality assurance measures and tests. This makes the role of the technical worker especially important (Fig. "Systematic analysis of problems i").
Figure "Relation of quality and work conditions": The quality of a product, and therefore its operating safety, are ultimately determined and ensured by skilled work. The technical personnel who directly “create quality” include primarily
- design engineers
- process testers and developers
- work preparers
- technicians and operators
An optimal environment in which to complete their tasks should be provided in order to support their quality-determining characteristics such as
- realizability (ability).
However, realizing this is a challenge for the responsible decision-maker. Often, apparently more important priorities preclude the creation of a desirable environment. This is environment is determined by factors that also influence one another reciprocally. They include primarily:
Recognition can take many different forms. Although compensation plays an important role and must primarily be in clear relation to comparable tasks and responsibilities, personal and documented appreciation, such as in a technical hierarchy and/or special capacity, is often underestimated. Appreciation includes technical and background information (Ill. 17-5) that is necessary for understanding ones own importance to part safety (Fig. "Action needs understanding the problem"). Early involvement in technical decision making also signifies recognition. This also includes conceiving equipment or adapting facilities. Work reduction is not always motivating. The explicit utilization of experience, on the other hand, can have a motivating effect.
A workplace requires not only a motivating environment but also equipment that is as suitable as possible for accomplishing the task at hand. In this context, the human factor plays an especially important role in aviation. It refers to the effects of the environment on the risk of flaws and prerequisites for flaw-free work (Ref. 17.1-8). The U.S. Federal Aviation Administration is especially engaged in this field.
Quality assurance includes the activities in the framework of self-testing (e.g. for dimensional accuracy). This includes the technical evaluation of the state of the part before and after the work step for which one is responsible (Fig. "Quality assurance by process monitoring"). This is an absolute necessity for safe parts with low production costs. Experience is necessary in order to recognize signs of potential problems. This involves the detection of unusual changes such as chip formation (Fig. "Observation during the finishing process" ), tarnishing, or shiny part surfaces (Fig. "Warning signs of unusable baths").
Documents and regulations: The clearer and more understandable these are, the more they will be able to prevent misinterpretations or false assessments. In this case, as well, information regarding the backgrounds and requirements is a prerequisite for their optimal and engaged application. In contrast, confusing or easily misinterpreted work papers will compromise part safety. Translated documents, such as in the case of licensed production, can make their implementation more difficult. Experience has shown that this usually involves assessment characteristics and technical terms that were not clearly translated.
Work plans are of the greatest importance for product quality. For example, the sequence of work steps influences the probability of problems and the detectability of flaws in parts.
In many cases, in the context of minimizing problems and damages, the industry-specific and company-specific technical education and training is in need of improvement. This is especially clear when compared with the typical training options for the use of computer applications for clerical and administrative work. For the areas of production and finishing technology, especially, experienced-based continual training would appear to be a plausible endeavor, but it is only offered and used to an insufficient degree, if at all. When there is a lack of technicians and engineers, a suitable training program for technicians and skilled workers who are responsible for creating quality is very difficult to find (Ref. 17.1-5).
Figure "Action needs understanding the problem" (Ref. 17.1-5): If one does not clearly understand the consequences of ones own actions, there is an increased risk of errors. If the necessary background information is missing and/or the connection with later operating behavior of the parts is not sufficiently understood, the potential of the greatest possible safety and effectiveness will not be fulfilled.
The mousetrap is intended to make this situation more clear. It shows that inspite of broad knowledge, insufficient experience creates an increased risk. The reason for this is mistaken estimates of the consequences of ones actions. This applies, for example, for unilateral decisions regarding the allowability of, and solutions for, deviations.
It is correct to involve knowledgeable superiors and technical departments sufficiently early. However, this requires a situation of mutual trust in which the bearer of bad news is not punished in a medieval manner (Fig. "Negative motivation").
The same is true for the operating suitability of parts, which is decisively influenced by a large number of apparently minor factors. Correct assessment and realization of these “minor factors” with a proper attitude is a decisive step in assuring quality. Therefore, a prerequisite for quality-conscious action is knowledge of important relationships and influence that can affect parts. These include processes and technical personnel with decision-making power who can enable suitable solutions and evaluations (Fig. "Systematic analysis of problems i"; Fig. "Minimizing scrap rates throuch reworking").
Figure "Quality assurance by process monitoring" (Ref. 17.1-5): The contribution of technicians to ensure the quality of products should not be underestimated. With sufficient experience, training, and motivation, they provide especially important, reliable continual process monitoring (Fig. "Observation during the finishing process").
For example, during the production process a technician sees the behavior of the part, the process sequence, and the tools. The technician can notice deviations that indicate problems. These include different shaving types (Ref. 17-6) or unusually intense tarnishing, which may indicate problems with the stability of the production process. They can indicate serious problems that can no longer be reliably detected with subsequent testing methods. Therefore, motiviation and continued training (Fig. "Maintenance penetrant inspection part two") are a worthwhile investment, especially in the production process.
Motivation (Ill. 17-2) includes a suitable configuration of the place of work and further training that is matched with the demanding task at hand. This training should be especially designed to develop an understanding of technical relationships. Every production step is tied into the process, and has a history and after-effects. Therefore, it is important to have and provide knowledge about how ones own work can be affected by previous production steps, and how it affects subsequent steps and especially the operating characteristics of the part.
Figure "Quality ensurend by communication" (Ref. 17.1-5): Work at computer screens and in largely automated production processes increases the danger of a certain distance from reality and isolation due to a lack of personal contact. This risk is not only present in the “desk work” during design, but also in the production area. Work preparation and the evaluation of work results in the production process are increasingly affected.
This trend presents a risk that should not be underestimated. On the one hand, experiences that are best conveyed in face-to-face conversations are no longer being properly transmitted.
In addition, the opportunity of unforced and, therefore, especially creative interaction with colleagues, which has a certain controlling and correcting function, is lost.
Direct personal contact can also be an important factor for the operating safety of the parts.
Therefore, technical communication should be sought out in line with the proverbial saying that “four eyes see more than two.”
Figure "Necessity of part-specific testing" (Ref. 17.1-7): If the same new car model is repeatedly seen broken down along the side of the road, it will have a negative influence on the purchasing behavior of potential customers regarding this brand. It would have been better for the manufacturer to invest sufficient resources and time into testing and development in the first place, rather than risk expensive retrofitting and the creation of a negative image.
This scenario can also be applied to engines. In this case, there is also the special aspect of safety. Especially with regard to the influence of production processes, part-specific testing is an absolute necessity. This requires considerable time and is usually quite costly. In comparison, a computer simulation is considerably faster and cheaper. However, experience has shown that it is unable to adequately replace hardware tests in many cases.
The developed production parameters must then be observed in serial production in order to guarantee the verified operating safety.
Therefore, it is necessary to realistically estimate and incorporate the required costs and time for testing new or altered production processes into the total framework. For this reason, sufficient medium-term process development, including the operating behavior of the parts, is required before reaching serial maturity (Volume 3, Ill. 14-36). Processes that have not been sufficiently tested carry a high risk.
Figure "Fundamentals for describing quality" (Refs. 17.1-1 and 17.1-.2): The stability of a production process is very important to part safety and cost-effective production (Fig. "Steps minimizing flaws"). An important sign of stability is the statistical distribution of the deviation of important characteristic values or quality indicators (top diagrams) over time (bottom diagrams). The scattering of the deviations (is-should be) must be safely within the allowable tolerances, i.e. between a lower (LTL) and upper (UTL) tolerance limit. The process range is defined as 6 x standard deviation (s ). This means that 99.73% of all values can be expected to be within the process range. The process range is symmetrically split by the nominal dimension. The process capability cp is the relationship of the tolerance range to the process range. The pattern of the frequency curves can be used to draw important conclusions regarding the production process (Fig. "Frequency distribution curves (histograms)").
The top left diagram shows a symmetrical distribution that is also symmetrical to the tolerance range. This is not always the case (top right diagram). The bottom right diagram shows such a case. The distance of the process mean to the tolerance limit is also an important characteristic.
The process capability cpk is the ratio between the smaller distance from the nominal dimension to the tolerance limit and half of the process range.
Of course, a symmetrical distribution relative to the tolerance range is desirable (bottom left diagram). This provides the greatest degree of safety for keeping the process within the tolerance range. It is possible to implement stabilizing measures by observing the measured values over time. One example is controlled setting of a profile grinding disk in order to compensate for disk wear.
Unlike detecting dimensional anomalies, detecting the production flaws that are the focus of interest in this case is often only possible indirectly through tracing of the relevant process parameters. In order for this to be successful, the relationship between the causes of the observed flaw type and the process parameters must be sufficiently understood. If, for example, grinding cracks or signs of overheating are only detected by a suitable subsequent testing process, statistical considerations can still be applied in a useful manner. They can make it possible to limit the flaw frequency to individual process parameters and draw conclusions based on the distribution.
Figure "Frequency distribution curves (histograms)" (Ref. 17.1-1): The pattern of the frequency curves permits conclusions regarding the production process and/or its assessment. It can be very helpful for targeted subsequent testing of potentially affected parts, as well as for determining causes.