Table of Contents
13. Damage-Minimizing Construction
This chapter deals with the conditions of the work environment, methods of approach, and self-conception of the design engineer, and additionally provides recommendations for minimizing the risk of damage. It does not enclude detailed specifications for constructive design and configurations, as this is the task of design engineering manuals.
In addition to his creative role, the design engineer should feel responsible for the realization of his designs, and his self-conception should correspond to his central position in the development process (Fig. "Designer responsibility for safety"). This task can be summarized under the term “structural integrity” (Fig. "Planning engine parts safety by structural integrity"). This is the topic of Chapter 13.1; Safe Life Span. The high utilization of the strength of parts in modern engines leads to limited life spans and an increasing influence of flaws. At lower loads in older engine types, these flaws could still be seen as weak points. For this reason, the evaluation of flaws and possible crack growth is becoming ever more important.
Flaws can be prevented most successfully in the design phase. Flaws that are only found later, such as during production or operation, require considerably more effort and expense to remedy (Fig. "Preventing flaws as first phase "). Therefore, it is extremely important to avoid weak points and problems in the product during the design phase.
The term “damage-minimizing design” can be interpreted in different ways. Even the environment of the design engineer influences the probability of later problems. In addition to ability, experience is a requirement for successful design engineering. It can help prevent repeating serious mistakes, promotes evolutionary improvement as a condition for the safety of complex facilities, and minimizes problems in production and serial operation. Personal experience generally costs money, time, and prestige. Therefore, it is beneficial to acquire experience from someone else. Unlike knowledge, experience for practical applied action cannot be gained effectively through reading and studying databanks. A far more effective method of transmitting experience is through spontaneous, trusting, and helpful technical discussions. This method is more likely to ensure a personal connection with the subject, which is an important condition of experience. However, the working conditions of design engineers are becoming more and more of a
hindrance to this type of communication. The first requirement is an organizational structure that supports the spatial and temporal conditions necessary for communication with experienced individuals. To this end, technical consultants and technical teams can be given the task of conducting communication. In the past, design engineering in the form of an outline on a drawing table was practically an invitation to personal technical discussions. In this way, chief design engineers were given the opportunity to use their technical expertise and analyzing abilities to provide support and suggestions in an unconstrained discussion. On the other hand, while the modern computer screen work station has many advantages, such as accessibility of procedure documentation and integration into work processes, it also has serious problems. In spite of the extensive networking capabilities, the computer work station leads to isolation and makes it far more difficult for experience to be passed on. The position of the design engineer with his back to the outside world, sitting in front of and largely blocking a screen that is considerably smaller than a drawing table, is not an atmosphere that promotes the transfer of experience. In addition, overviews are made more difficult by the limited sections of diagrams that are shown on the screens. This makes it hard for a casual observer to spot unique characteristics that might lead to a discussion. The design engineer himself can work to resolve these problems, depending on the specific corporate culture (Fig. "How to use experience").
An additonal aspect of damage-minimizing design engineering is incorporating design rules that may often seem trivial, but are not always obvious and can have catastrophic results if overlooked.
Computer-based design engineering (Fig. "Knowledge based design") is subject to increasing criticism, at least during the concept/design phase. The major criticism is that the required creativity and knowledge base are not being used. As several reported examples demonstrate, experience is often not sufficiently incorporated into projects (Ref. 13-14). This makes an optimal solution, which would also manifest itself in a cost reduction, seem less and less likely. Instead, so-called knowledge-based design engineering is recommended. This method makes better use of the special abilities, knowledge, complete concept, and experience of the design engineer. The abilities of the design engineer as a person become more important than the optimization of processes. This method of design engineering is evidently more similar to that of earlier generations.
Figure "How to use experience": Experience and its transmission are prerequisites for damage-minimizing design. Within the framework of the prescribed environment, the design engineer can contribute greatly to his experience and professional success. In order to defuse the disadvantages of the modern workplace with regard to personal communication, the design engineer should enforce certain methods of behavior. This might be accomplished by accessibly hanging up a printout of the work in progress in regular intervals for his colleagues to examine. Of couse, this requires suitable configuration of the work area and is not always the most immediately cost-effective option (e.g. larger work surface). Communication can also be promoted in a positive motivational way by a technically competent superior. For example, this person could show an interest in the work in progress and bring together various specialists without creating a controlling, competitive atmosphere. Naturally, these behaviors are also determined by the corporate culture, and therefore by the position of technical ability in the technical hierarchy.
This also includes the rank of technical (sustained) advanced training, which should be noticeable in the ambience (e.g. external).
The opportunity to observe assembly and production should certainly be taken. This is especially true if one`s own project is being realized. Feedback from production and operating experiences is extremely important for the design engineer, and should be gathered. Production and assembly should be consulted in a timely manner in order to prevent problems in the run-up.
Specific information in a good technical library, which should be available in a high-tech company, can help prevent “dead ends” and enable one to take promising approaches from the beginning. The question, “how do others do it?”, should be of primary importance. For this reason, visits to the library should be encouraged. During concept design, proof of research could even be made mandatory. Patent research must also not be neglected. Patents represent the state of technology, their documentation shows problems that occur in the inventive step, and prevent the use of unauthorized technologies. Modern electronic communication with databanks and the internet provides a wealth of information. For example, the homepages of aviation safety authorities (NTSB, FAA, etc.) contain publically accessible, often surprisingly detailed technical descriptions of flight accidents and dangerous incidents, making these especially important sources.
Another important factor is a design engineering handbook that ensures the use of company-specific proven technologies and design principles, and plays a crucial role in preserving and transferring these experiences.
One must also remember overhaul manuals and maintenance manuals. These can provide important information regarding part-specific weak points and flaw limits.
A good collection of experiences contains an extensive documentation of damages. They should be searchable based on as many different terms as possible, and should be made easily accessible (preferably online) and clearly arranged for the use of design engineers. It should be ensured that the interested user can easily obtain all relevant documentation and attendant illustrations for specific damage cases. This is the only way to ensure that the incident is actually relevant to the issue at hand, and typical, visually-focused design engineers are given the opportunity to gain their own impressions of the situation.
Figure "Knowledge based design" (Ref. 13-13): The following consideration is based on the cited literature. It supports the hypothesis that the younger generation of design engineers already has an exclusive reliance on CAD (Computer Aided Design), a condition that does not yield optimal results. This trend makes it necessary to understand all possible risks and weaknesses of this method of approach, and to be able to make any necessary corrections.
Can methods of approach of the type that were used by earlier generations of design engineers reveal weak points and possible solutions to a design engineer? Knowledge-based design (KBD) is currently being discussed as an alternative to computer based design (CBD).
In KBD, the design engineer (not a computer) takes into consideration all necessary knowledge that he has collected and compiled, and combines this with the necessary dedication in order to find the right design. KBD is focused on the individual, and is therefore not accurately definable. Data, information, and knowledge that are compiled during the design process also flow into the knowledge base. This makes possible the total overview that is necessary for the critical evaluation of the data.
To ensure an optimal result, the design engineer must use all his abilities, senses, and knowledge. This method of approach differs from others, such as artificial intelligence (AI) and expert systems (ES). With these methods, the decision is ultimately not made by a human.
Unlike KBD, CBD makes use of both AI and ES. The more CBD, with its available computing programs, comes to dominate the typical multidisciplinary surroundings of the design engineer, the more likely erosion of experience-based ability becomes. This can weaken the knowledge base for the future. This trend is evidently being observed in fields such as strength and aerodynamics. CBD is viewed as fast and inexpensive, and is therefore seen as a more effective approach. These expected advantages depend decisively on the reliability of the programmed tools. Proponents of CBD, on the other hand, consider KBD to not be accurate enough for modern design engineering processes. For this reason, it is only given the role of moderator or knowledge manager.
KBD strives to use the design engineer as an individual with special abilities whose most important tasks are creative and inventive. This leads to a clear distinction between a design engineer, who is typically both creative and responsible for system functions, and a person who conducts routine design engineering processes. A similar comparison might be made with playing an instrument in a concert and playing a record on the stereo.
The left diagram shows how strongly the manufacturing costs of a product are influenced by the initial design/concept phase (“influence”). In this phase, the work is determined by basic effects and simple models. At this point, CBD methods have a large disadvantage, which is that evidently only a fraction of available knowledge for solving the specific problem is connected with the programs being used.
Therefore, the cited literature recommends that three conditions are created in order to minimize costs and maximize the results:
- creation of a highly creative and multi-disciplinary design phase
- using KBD in order to ensure the necessary intellectual freedom
- recognition that success depends decisively on the design engineer, not the design process.
It is expected that this approach would positively shift the curves in accordance with the black arrows.
The right diagram serves as an argument for the use of KBD. With simple, easily overseeable tasks (low complexity of knowledge), the curves of KBD and CMB are almost equal. In the case of complex relationships, however, the CBD costs increase dramatically with the computation costs and necessary auxiliary work. KBD can reduce these costs considerably through proper limitations and pre-selection of potentially successful solving methods, which are based on a more complete knowledge-based overview.
Figure "Passing the baton does not work": The Senior Expert
A major problem is the loss of senior experts. This problem is exacerbated if, during a slump, an entire generation of older, experienced experts, such as design engineers and professionals from development, finishing, and testing, are retired. Even apparently minor changes to safety-critical parts such as rotors demand comprehensive verifications, i.e. new ratings. A typical case is the introduction of a new generation of machining tools, such as electron beam welding machines. If a senior expert retires and his task is taken over by an inexperienced successor, a comparable level of safety is the exception rather than the rule. Therefore, this type of change not only increases the risk of cost increases and time delays, but also affects the safety of the parts and thus that of the entire engine. In order to minimize this risk, there should be an available plan for the optimal procedure when a senior expert leaves. This plan, which must be followed, should contain the following elements:
- Sufficiently early naming of a successor. The successor should be given roughly one year to become familiarized.
- Selection of the successor with input from the senior expert. Under no circumstances may a controversial successor be named. Effective “working-in” can only be expected with proper chemistry between the senior expert and successor. This ensures that the knowhow is passed on effectively.
- The successor must be taken to all relevant technical discussions. He or she should be introduced to all future clients and team colleagues. As the successor becomes more familiar with the work, he or she should be given more an more responsibility regarding the technical representation of his or her field. The senior expert must provide support.
- The successor must be taken on trips to see partners and customers, to conferences, etc.
- Technical reports and correspondence should be increasingly composed by the successor with positive assistance and guidance from the senior expert.
- The successor should share an office with the senior expert. This ensures that discussions and information are shared freely. When the senior expert is on vacation or otherwise absent, the successor can gain experience and increasingly assume responsibilities.
- The senior expert should introduce the successor to the collected literature and other important documents, such as regulations and reports, which have generally been collected over years. The successor should be animated to independent use of these documents. This allows the successor to become familiar with the system of the collection, which is usually very individual.
The issues surrounding a senior expert also involve personnel policies and corporate culture:
- Internal promotion: In this case, a senior technical expert is promoted into a managing role within his or her own long-term organizational unit. This type of advancement is hardly possible today, and is often explicitly prohibited. This presents a danger of losing experience and knowhow that has often been obtained at great cost. Experience has shown that a technical hierarchy is not a satisfactory alternative. New structures may form that consciously avoid contact with isolated senior experts, who are typically “undisiplinable”. In the interests of engine safety, cost efficiency, and company prestige, internal promotion with personnel responsibilities for senior technical experts should be desirable.
* Job rotation: Job rotation is the common procedure by which the tasks and position within the company are changed every few years. This type of change can be done with a bit of help from the personnel department. Naturally, this policy prevents senior experts from ever developing. It takes many years, if not an entire working life, to become a senior expert in the typical complex fields of turbine engine design.
- Outsourcing: This is a method of cost reduction in which work is acquired from external suppliers. This procedure contains the danger that the outsourced tasks are those of important senior experts, and the company loses its capability for development tasks in its own core industry. Similar effects can occur if cooperation with universities and technical institutes is not strategically guaranteed. Often, the experience that has been obtained through hard work is not secured for the long term. For example, a long-term employee of the contractor may leave. In order to prevent this type of situation from occurring, a company must specify the tasks that are not to be outsourced. If a company has decided to outsource tasks which are expected to yield important experience, or which use its own senior experts, continuity must be ensured through a binding long-term strategy.
Figure "Designer responsibility for safety": If the design engineer feels responsible for the ultimate success of his or her project, he or she needs a technical horizon that goes beyond the actual design work and is sufficient to understand other specialists who are involved in realizing the project, and also to be able to initiate and evaluate actions. These tasks are summarized under the terms structural integrity and damage tolerant design (Fig. "Planning engine parts safety by structural integrity" and Fig. "Damage tolerance concept").
The design is inseparably connected with the available materials (Chapters 12.1 and 12.2). This applies, for example, to the usable static and dynamic strength properties and the demands of certain operating influences, such as corrosion, wear, or rubbing. The strength properties are closely related to the quality of the serially-implementable testing processes.
Realization of a project requires suitable finishing work. Therefore, it is important to consult with the finishers (work preparation) to determine, for example, whether the part is sufficiently accessible for the finishing work to be successful (Fig. "Accessibility of threaded connections"), or if the required dimensional tolerances, mass tolerances, and surface quality can be assured without excessive cost or difficulty.
Assembly and installation are important for part behavior in serial operation (Fig. "Reworkability and testability of rotor weld seams"). For example, notch-sensitive materials such as titanium alloys should be used in a way minimizes the danger of handling damage. In modular design, it must be ensured that no assembly damage occurs at the segment connections (bearings, labyrinths).
For maintenance, not only good accessibility must be assured. It may also be necessary to specify auxiliary materials, such as cleaning agents, while considering special design characteristics (e.g. coatings). Adhesives and other synthetics (matrix resins, rubber, lacquer) must be selected in the design phase based on their tendency to age and become damaged.
Last but not least, the design engineer must consider repairability. Both the limit of repairable damages and the feasibility of these repairs must be considered (overhaul manual). A hot crack-sensitive labyrinth material, for example, is likely to have serious problems during welding repair of the tips. Sprayed coatings that are made from a very similar material to the substrate are very difficult to remove without damaging the base material (e.g. etching before non-destructive testing).
Figure "Warning signs for design risks": The design engineer will receive various warning signs indicating design risks and flaws from a fairly early point on. These can be divided into preliminary developments and serial applications. Warning signs for serial applications include:
Unusual structural characteristics and design principles: The saying “don`t be the first to try something” can certainly be applied to serial parts. For example, why would an obvious structural design never have been used before in a serial product? It can hardly be assumed that all other competitors are simply less intelligent. It is much more likely that this design has a fundamental problem (perhaps a showstopper, see Fig. "Show stopper during development") that has not yet been recognized.
Chronic quality problems, shipping delays, and high reject rates in a supplier of semi-finished products may indicate permanent problems related to specific requirement combinations of a desing. These could be, for example, overly close tolerances or overly sharp cross-section changes in cast parts. Unusually pronounced hot cracking in part zones that are also exposed to this influence during operation will in all likelihood result in the same damages occurring later in the engine. A typical example is material flaws in compressor rotors made of titanium alloys. These problems seem to affect engine technology since the early 1970s (Fig. "LCF fracture of a fan disk").
If several similar damages occur during the development phase and testing, it is an alarming sign for an engine-specific weak point (Fig. "Weaknesses by design"). If the specific cause of these weak points is not specifically removed in an early stage, but rather merely “defused” by combating contributing influences, they will continue to affect the engine type for its entire operational life (Example "Flawed design principle").
Flaws that can not be sufficiently safely detected with the available, serially-implementable non-destructive testing procedures reveal an unsuitable material selection and/or overly optimistic strength estimations. This situation can become evident as early as the outturn pattern testing or during cyclical centrifugal tests for life span verification.
If untested and/or strange technologies are being used, it should alarm the serial design engineer. Safe application of a technology often requires knowledge of an experience with special properties that do not reveal themselves until very late, in the worst case, after long operating times.
Example "Flawed design principle" (Ref. 13-12):
Excerpt: “…The new HPC ring case design, developed to counter a string of surge issues that have nagged the 2.4m (109in)-fan diameter version for several years, is based on a set of one piece rings that replaces the original segment design. The 2.4m fan…derivative…suffered `an asymmetric distortion of clearance due to case deformation, while the ring case for the 112in version of the engine remained perfectly round, offering much better control'…Better matching of the expansion coefficient ot the rotating stages and the casing has now been achieved for the critical high power, low-altitude conditions that have often led to surges occurring at the take-off, says the company….Between 70% and 75% of the 2,200 affected engines will be retrofitted during overhauls…deadline for fleet completion by the first quarter of 2007…“
Comments: Surges in this engine type have been repeatedly reported since its introduction (Ref. 13-11). The birth defect (Fig. "Weaknesses by design") was evidently the axially segmented housing (Volume 2, Ill. 7.1.3-20). The problems could apparently not be solved by evolutionary methods, so the design principle had to be changed.
Figure "Weaknesses by design": In nature, genes determine the stronger and weaker organisms. Weaknesses can only reveal themselves at an advanced age. Experience has shown that engines behave in a similar fashion. Weaknesses in certain engine types reveal themselves early in the development phase. These weak points cannot always be easily avoided through evolutionary development. This is the case if the main damage cause is connected with a special design characteristic (Example "Flawed design principle"). The weak point must instead be defused by optimizing secondary contributing influences. This can result in a life span that is sufficient for the development phase, but that the problems reappear during later serial operation with its typical long run times and/or in the framework of repairs.
If there is a suspicion that this type of weak point exists, development protocols and/or overhaul manuals can make safer evaluation possible. Typical weak points are shown in the adjacent diagram. These include a tendency to stall (Example "Flawed design principle"), for example, when the cannon is fired (Fig. "Compressor surge influenced by cannons"). Additional examples are ignition problems in the afterburner that cause a stall in certain parts of the flight envelope (Fig. "After burner triggered compressor surge"). Flexure of the engine core (backbone bending) became a special problem with the first generation of large fan engines and was only satisfactorily resolved with the introduction of external stiffening braces (Volume 2, Ill. 7.1.2-17).
Figure "Design based on evolution": Comlex machines and/or large series with the lowest possible failure rates (such as in the automotive industry) demand very careful action during development in order to keept the risk acceptably low. It has been shown to be extremely risky to introduce new or revolutionary technologies without the necessary background of experience (Fig. "Development time of fiber fanblades"). Even technologies that have proven successful in competitors` products for many years, cannot simply be adopted if the experiential background of design and operating and long-term behavior is insufficient. A typical example is large fan engines with two or three shafts. Evidently, the three-shaft design has considerable advantages with increasing engine size (weight, deterioration; Volume 2, Ill. 7.0-2). However, this design is not used by most OEMs, even though they certainly have access to the design tools. It can be assumed that specific problems of the three-shaft design, such as the long-term behavior (today`s normal guaranteed life spans) of the intermediate shaft bearings, may be preventing the accptance of this design principle in the short term.
For this reason, the process in engine construction, especially for civilian engines, is generally evolutionary. It is based on proven, “understood” design principles that were incrementally improved over long periods of time. “Understood” means understanding the design, configuration, production, and operating behavior, even under unusual loads. These designs are generally adhered to even if there are problems, and further improvement steps are usually tried before a new principle is adopted, which will avoid the known problems but might have other weak points.
Figure "Use understood construction principles": Understanding a design principle does not mean simply grasping its function and creating the appropriate configuration. Experienced engine design engineers know that many years of operating experience are necessary to recognize and eliminate the failure mechanisms and weak points of a design principle. These include fretting, coefficients of friction (changes over time), aging, fatigue (LCF behavior), and corrosion. Certain properties, such as stiffness, forces in removable connections, and damping, can influence other components in unforeseen ways.
The selected example is a design detail (Ref. 13-1) that shows the connection of compressor rotor blades with the aid of interference-fitting and radial pins to transfer torque. This seemingly simple and cost-effective solution would require a great deal of experience and could definitely not be adopted for serial operation without the necessary experience. The application limits must be known in order to sufficiently safely ensure proper functioning throughout the life span of the rotor. The notch effect in the pin area can restrict the application considerably due to factors such as an overly short LCF life. The connection influences the stiffness and, therefore, the vibration behavior of the rotor. For this reason, the entire rotor design must be adjusted with regard to this principle. The shrink-fit must naturally retain its function during unsteady operation, such as during temperature changes and expansion due to centrifugal force during startup (see Fig. "Overheating and fusing at shafts by vibrations"). For example, pwerful dynamic forces can act on the connection during compressor surges. Micro-movement must not lead to unallowable wear in the radial bolts and contact surfaces. This raises issues regarding the tolerances of the connection, coatings, and the material combination.
Of course, assembly must be safe and simple.
Even if the conditions of normal operation are mastered, behavior in the case of damage and/or under unusual loads must also conform to regulations. This is also true for the behavior during rotorbow (Volume 2, Ill. 7.1.2-9). Other possibilities include large imbalances following bird strikes or a blade failure.
Repairability must be made clear. For example, within which limits is it possible to repair wear or corrosion. This determines the consequential costs, which can be inacceptably high for the operator and/or, if there is a run time guarantee, for the OEM.
Figure "Why a design did not go into series": An important warning sign, especially when introducing new designs and technologies, is the absence of previous reported serial application (Fig. "Risk as first user of a technology"). We must be wary of assuming that we are the first to discover something that would seem evident to experts. While this may be possible, it is fairly unlikely. In this case, it is very probable that there are unrecognized problems that the competitors were unable to satisfactorily resolve.
The selected example is a fiber-technical containment for fighter aircraft engines. It would be a fatal flaw to assume that, based on the serial implementation in civilian aircraft with nacelle engines on the wing, this technology would be transferable to engines located in the fuselage, even though this use has never been reported. In this case, the first question that must be answered is why the competitors never used the obvious advantage of lighter weight.
Tests in a centrifugal testing rig would rapidly show that utilization of the advantages of fiber-technical containment rings requires sufficient elastic strain in the ring (Volume 2, Ill. 8.2-5). However, there is usually not enough room for this in the tight spaces in the fuselage, or it must be an integral part of the initial concept of the engine and aircraft (installing accessory equipment; Volume 2, Ill. 8.2-18).
In addition, one must determine if other unexpected effects occur. It is possible that the surrounding temperatures are high enough to damage the fiber ring, or containment incidents might damage the parallel engine. One must also examine whether assembly and maintenance (e.g. moistening with fuel) result in any problems.
Remember: Without sufficient personal experience and/or assistance from a licenser, the introduction of new design principles demands comprehensive development work, even if the technology has been proven in other situations.
Figure "Adhering to a life span concept": An OEM has a safety concept that is recognized by the acceptance authorities. This is dependent on the design philosophy, e.g. the utilization of part strength and life span concept, especially where the life span of rotor parts is concerned. This type of concept also incorporates the entire production of primary material and semi-finished products, quality assurance (e.g. non-destructive testing), and specific operation. These individual ancilliary conditions usually do not allow a head design engineer (outside of licensed production) to take over this type of concept rather than his or her own. In extreme cases, such as with multiple projectw with different partners, several application-specific safety concepts can exist in parallel.
It is a fundamental principle that ones own established safety concept is certified and used in all products that one is responsible for.
One example is life span verification of rotor components with the aid of cyclical centrifugal tests. These tests are cost- and time-intensive. In this case, there seems to be a possibility for savings through simplified tests. Under closer consideration, however, it is clear that the design must also be adjusted accordingly. For example, it is thinkable that the acceptance authority demands a lower strength utilization, which results in heavier parts. It is also possible that the simplified procedure requires extensive experience, which is difficult to ensure.
Remember: Ones own certified safety concept must be maintained in parts that one is responsible for.
Avoid mixing safety concepts.
Figure "Fundamental construction concepts": Every OEM has typical design characteristics. Of course, these can be related to the requirements of the operator, such as demands of authorities in military applications.
The top diagram shows the probability of an engine failure due to enemy fire when the auxiliary equipment is located at the bottom of the engine (Ref. 13-2). One can see that damage to the auxiliary equipment is responsible for a high percentage of engine failures in combat situations.
If the auxiliary equipment is located on top of the engine (middle diagram), it has the advantage of decreased vulnerability to ground fire. Equipment located at the bottom of the engine is more easily accessible during overhauls and makes it easier to design the engine suspension.
Many important properties are already specified when the engine is being conceived. Subsequent development can only changes these to a very small degree. If new technologies are to be incorporated into a new engine type, the concept phase is crucial, and all necessary, design-influencing properties of the new technology must be readily available for serial design (see Fig. "Why a design did not go into series"). This makes it vital to have a long-term development strategy for technologies of ones own core products (Fig. "Time periods for series application of a technology").