Table of Contents
13.1 Safe Life Spans
This chapter discusses several fundamental principles regarding damage tolerance and damage-minimizing approaches to configuration and design, as well as the corresponding part behavior. This concept includes potential cracking and its influence on life span (Fig. "Configuration oriented safe life"). The contrast to this is the earlier concept of safe life, which excludes crack initiation during the designed life span. No detailed instructions for action are provided. Instead, the problem mechanisms are explained and further work with these is encouraged. Therefore, this chapter does not replace the information in a design engineering handbook.
Fail safe design occupies a prominent place in this chapter, since it deals with controllable part behavior in case of crack initiation and fracture. It is desirable that crack growth does not lead to loss of operatability before the damage is discovered, i.e. before the cracking makes itself known. This be accomplished through early detection based on charactecteristics such as imbalances and vibrations, unusual sounds, or indicators during boroscopic inspections. The task of a fail safe design is becoming ever more difficult with the higher utilization of strengths, smaller safety margins, higher dynamic loads, higher RPM, and higher temperatures. The reason for this is that smaller flaws become capable of growth, the crack growth rate grows, and the critical crack length, at which residual fractures occur, becomes shorter.
Figure "Planning engine parts safety by structural integrity" (Ref. 13-3): Structural integrity depends on a great number of influences that are encompassed by the entire development phase. The development phase is therefore superordinate to many different fields, such as materials science, strength, and fracture mechanics. Development of safe, complex, and highly-stressed parts demands the iterative integration of the entire surroundings of a design engineer. The design engineer should be required to take a coordinating role. Structural integrity must be planned in the design engineering/configuration phase with consideration given to the following fields (there is no claim to completeness):
- stress, strain, load, and temperature analyses (strength)
- a concept for dealing with fatigue processes
- analysis of cracked components (fracture mechanics)
- flaw criteria (e.g. allowable flaw sizes)
- wear concept
- concept for creep processes
- concept for corrosion loads
- material selection
- material deformation and damage mechanisms
- inspectability of parts (e.g. accessibility)
- non-destructive testing
- manufacturing process (e.g. feasibility)
- maintenance feasibility
Aspects within the framework of structural integrity are safe life design and damage tolerant design (Fig. "Configuration oriented safe life").
These concepts/philosophies take fatigue processes into account.
Safe life design: This approach demands analyses and/or tests to verify that the probability of each type of failure is extremely remote during the assumed life span (Volume 1, Ill. 2-3). It is irrelevant, whether the assumed life span is limited or unlimted. Crack initiation is excluded from these considerations.
Damage tolerant design, including fail safe design (Volume 1, Ill. 2-2): This concept is used especially for taking into account the influence of flaws on the life spans of cyclically stressed parts. It is assumed that acceptable, sufficiently safe operating behavior of a damaged structure must be guaranteed until the damage is discovered and suitable corrective measures can be introduced. This can be accomplished with the aid of the following measures:
- identification of critical components; i.e. critical zones
- multiple load introductions/acceptances (redundancy, Fig. "Damage tolerance concept")
- crack stoppers (Fig. "Remedy for thermal fatigue cracks by repair")
- proof tests: These include overstress tests such as overspeed tests in rotating parts made from brittle materials with no pronounced fatigue behavior (e.g. ceramics)
- methods of estimating crack growth
- inspections with boroscopes, visual inspections during maintenance, non-destructive testing (e.g. penetrative testing, eddy flow testing).
In damage tolerant design, one assumes, for simplicity`s sake, that there is already a crack-like notch (also see Ill. 18.104.22.168-19). In order to estimate the safe life span (incubation and crack growth, Fig. "Factors influencing crack growth"), however, a great number of influences must be considered in order to determine the parameters necessary for estimation (Chapter 12.2):
- failure criteria (damages that indicate the end of part life, risk analysis)
- crack properties:
- type, size, location, and orientation in the stress field (see Fig. "Controllability of cracks by stress gradient" and Fig. "Detecting and controlling cracks"),
- detectability during inspections, method of spreading (crack opening types, Fig. "Understanding fracture processes"),
- accuracy of fracture mechanics (e.g. cross-section thickness, Fig. "Forced fracture behavior by section thickness").
- loads: Sizes, directions, and types (single or multiple axis) of forces, deformations, and temperatures; temporal progress.
- crack growth mechanism: Subcritical and critical, Paris function (Fig. "Characteristic crack growth")
- behavior of the material and part (e.g. stress relocation, Fig. "Design philosophy influences part failure"),
- fracture toughness
- influence of surrounding atmosphere (Ill. 12.2-11),
- load spectrum (Fig. "Dynamic fatigue life span estimations (Miner rule)"), temporal load progression.
Figure "Life span estimation" (Ref. 13-9): These diagrams give an impression of the procedures and influences that must be considered when estimating life spans for a safe life design concept (Fig. "Configuration oriented safe life"). In this case, the life span-determining criterion is the first crack. The life span lies within the incubation period (Fig. "Material changes during crack development"). One can see that experience (bottom diagram) is an important prerequisite for a successful, i.e. sufficiently safe, life span estimation.
Figure "Damage tolerance concept": The damage tolerant concept requires consideration of a large number of influences on crack frequency (Fig. "Failure mechanisms by operation loads"), crack size, and crack growth (also see Chapter 12.2), and assumptions must be made regarding their effects (top diagram, Ref. 13-3).
The bottom diagram (Ref. 13-8) shows the influence of surrounding media on crack growth. This must be considered if the data for a sensitive material were determined using material specimens in an atmosphere that does not correspond to the one during operation (Fig. "Minimum crack initiation flaw (threshold)" and Fig. "Types of crack growth under corrosion").
Figure "Failure mechanisms by operation loads": If, during design, one only considers the strength values that were determined in the usual tests (e.g. resting air) on relatively small specimens (Fig. "Background of material data for design"), the demand to utilize as much of the available material strength as possible can result in dangerous situations.
The higher the utilized strength, the smaller the allowable flaw sizes. This means that one may cross a threshold beyond which the manufacturer of the semi-finished parts can no longer guarantee the necessary quality and/or the serially implementable quality assurance measures are no longer sufficient. This is especially dangerous if this situation only becomes known at a point where only very elaborate measures will be successful.
Changes during operation can also unallowably affect part properties. The selected example is high-strength titanium alloys.
Based on their potential high-temperature strength, titanium alloys could be used well above 500°C. If one utilizes this potential, the result is new, safety-threatening effects.
For compressor blades, after the dynamic material strength, the friction conditions are most important (Volume 2, Ill. 6.1-12). The formation of oxide coatings at high operating temperatures is logarithmically plotted in the top diagram. One can see the sharp increase in the oxide coating thickness between 500 °C and 600°C. Thick, brittle oxide coatings should act to reduce the usable strength at the blade root (e.g. cracking under LCF loads) while increasing the FOD sensitivity in the blade leaf (crack initiation and HCF).
Above 500°C, the risk of stress corrosion cracking under the influence of sea salt increases sharply in titanium alloys (bottom diagram; Volume 1, Ill. 22.214.171.124-8). Because this corrosive media must always be expected in compressors, this damage form is also to be expected.
Therefore, the design engineer must ensure that the strength configuration takes into account this type of operating influence.
Figure "Configuration oriented safe life": There are two different life span design concepts. One is safe life design, and the other is damage tolerant design.
In safe life design, the allowable (configured) life span is limited to the incubation time (Fig. "Material changes during crack development"). This means that no crack growth occurs during the life span. Due to the long incubation time, a limiting of the cyclical life span is generally not provided for. This demands correspondingly low loads and was therefore a fundamental design principle in older engine types. If the loads are considerably higher, flaws can grow early, and the crack growth phase must be incorporated into the design in order to attain a sufficiently long life span (top left diagram).
For this reason, newer engine types with highly stressed parts must use damage tolerant design. This limits the cyclical life span under consideration of a growth-capable flaw size. Because the incubation life span is otherwise too short for the usual applications, (possible) crack growth is incorporated into the life span design concept (configuration). The top right diagram shows that, in a damage tolerant design with the same stress levels, larger flaws that are capable of growth can be allowable. The prerequisite is an acceptably long safe life span.
Therefore, the damage tolerant design concept concerns itself primarily with the influence of flaw size and crack initiation on part life.
Naturally, these concepts also influence maintenance and repair. This means that the inspection intervals in highly-stressed parts of modern engines are shorter in acute cases, and the danger of damage occurring is greater than in older engine types
Repairs on parts that are subjected to low stress levels allow greater play, although any changes in repair procedures must also undergo a verification.
Figure "Probability of a disk fracture": A material flaw in a part that unallowably affects its life span is only possible if several unfortunate factors combine. These include the production process for semi-finished parts, non-destructive testing, and the configuration (loads).
Naturally, the probability depends decisively on the load levels in the part (Fig. "Trends of materials applications"). Higher stress means that small flaws can become dangerous, and also reduces the probability of detecting flaws with non-destructive testing.