Failure analysis and life assessment of structural components and Equipment Introduction LIFE ASSESSMENT of structural components is used to avoid catastrophic failures and to maintain safe and reliable functioning of equipment. The articles in the"Structural Life Assessment Methods"Section in this Volume are written to provide an overview of the prevalent life assessment methodologies for structural components. Because the failure analyst is often asked questions concerning remaining life, fitness-for-service inspection intervals, and reliability of structural components and equipment, it is necessary that the failure analyst be aware of life assessment methodologies to address the questions and concerns of the industry he or the serves. Life assessment method advances and changes in technologies for structural components and equipment will require the investigator to adapt to the need of the industry. Furthermore, the failure investigator role has expanded from providing accurate identification of life-limiting failure mechanisms and degradation phenomena to also providing the time for degradation or damage, and crack growth rate to be used in life assessment estimates. Thus, the failure investigator's input is essential for meaningful life assessment of structural components. This article provides an overview of the structural design process, the failure analysis process, the failure investigator's role, and how failure analysis of structural components integrates into determination of remaining life, fitness-for-service, and other life assessment concerns. The topics discussed in this article include Industry perspectives on failure and life assessment of components Structural design philosophies Life-limiting factors The role of the failure analyst in life assessment The role of nondestructive inspection Fatigue life assessment Elevated-temperature life assessment Fitness-for-service life assessment Probabilistic and deterministic approaches Industry Perspectives on Failure and life Assessment of Components As noted previously, life assessment of structural components is a means to avoid catastrophic failures and to maintain safe and reliable functioning of equipment. Catastrophic failures of structural components occur rather infrequently, but when they do, they take a heavy toll on human lives in addition to the cost of repairs, replacement power, and litigation costs. In 1982, the National Bureau of Standards commissioned a study to determine the direct and indirect cost of fracture in the United States. It was estimated that $120 billion are spent annually to cover direct costs and costs associated with fracture-related accidents(Ref 1). The estimates took into account the necessity of overdesigning to prevent failure, added inspections, repairs, and replacement of degraded materials. Needless to say, the costs and stakes for failure are high. In addition, the cost savings are reat if failure can be mitigated or prevented For the failure investigator, a failure is often defined as the rupture, fracture, or cracking of a structural member The industrial definition of failure is often quite different from the textbook definition. A component, in practice, is deemed to have failed when it can no longer perform its intended function safely, reliably, and economically. Any one of these criteria can constitute failure. For example, a steam turbine blade whose tip has eroded affects turbine efficiency and hence affects the economics of operation adversely. The blade should therefore be replaced even though it can continue to operate. Component failures are thus defined in terms of functional"rather than"structural failures. Replacement of parts can be based on economic considerations, reliability, and material properties. In the discipline of life assessment, equipment and structures are evaluated
Failure Analysis and Life Assessment of Structural Components and Equipment Introduction LIFE ASSESSMENT of structural components is used to avoid catastrophic failures and to maintain safe and reliable functioning of equipment. The articles in the “Structural Life Assessment Methods” Section in this Volume are written to provide an overview of the prevalent life assessment methodologies for structural components. Because the failure analyst is often asked questions concerning remaining life, fitness-for-service, inspection intervals, and reliability of structural components and equipment, it is necessary that the failure analyst be aware of life assessment methodologies to address the questions and concerns of the industry he or she serves. Life assessment method advances and changes in technologies for structural components and equipment will require the investigator to adapt to the need of the industry. Furthermore, the failure investigator role has expanded from providing accurate identification of life-limiting failure mechanisms and degradation phenomena to also providing the time for degradation or damage, and crack growth rate to be used in life assessment estimates. Thus, the failure investigator's input is essential for meaningful life assessment of structural components. This article provides an overview of the structural design process, the failure analysis process, the failure investigator's role, and how failure analysis of structural components integrates into determination of remaining life, fitness-for-service, and other life assessment concerns. The topics discussed in this article include: · Industry perspectives on failure and life assessment of components · Structural design philosophies · Life-limiting factors · The role of the failure analyst in life assessment · The role of nondestructive inspection · Fatigue life assessment · Elevated-temperature life assessment · Fitness-for-service life assessment · Probabilistic and deterministic approaches Industry Perspectives on Failure and Life Assessment of Components As noted previously, life assessment of structural components is a means to avoid catastrophic failures and to maintain safe and reliable functioning of equipment. Catastrophic failures of structural components occur rather infrequently, but when they do, they take a heavy toll on human lives in addition to the cost of repairs, replacement power, and litigation costs. In 1982, the National Bureau of Standards commissioned a study to determine the direct and indirect cost of fracture in the United States. It was estimated that $120 billion are spent annually to cover direct costs and costs associated with fracture-related accidents (Ref 1). The estimates took into account the necessity of overdesigning to prevent failure, added inspections, repairs, and replacement of degraded materials. Needless to say, the costs and stakes for failure are high. In addition, the cost savings are great if failure can be mitigated or prevented. For the failure investigator, a failure is often defined as the rupture, fracture, or cracking of a structural member. The industrial definition of failure is often quite different from the textbook definition. A component, in practice, is deemed to have failed when it can no longer perform its intended function safely, reliably, and economically. Any one of these criteria can constitute failure. For example, a steam turbine blade whose tip has eroded affects turbine efficiency and hence affects the economics of operation adversely. The blade should therefore be replaced even though it can continue to operate. Component failures are thus defined in terms of “functional” rather than “structural” failures. Replacement of parts can be based on economic considerations, reliability, and material properties. In the discipline of life assessment, equipment and structures are evaluated
to determine if they are suitable, reliable, and economical for continual service. If deemed unreliable, the equipment or structure may be repaired, refurbished, or replaced any component the failure criteria need to be defined and established. Failure does not always involve acture or rupture. Progressive damage of structure and components under operating conditions leads to exhaustion of life, thus leading to failure. Damage may be defined as a"progressive and cumulative change acting to degrade the structural performance of the load-bearing component or components which make up the plant"(Ref 2). Life may be defined as the"period during which a component can perform its intended function safely, reliably, and economically"(Ref 3). With some modifications, the definitions used by Viswanathan and Dooley for fossil-fuel power steam plant components(Ref 3)can be used to define failure and life of components; that is, component life is expended when Design life has elapsed Calculations predict life exhaustion Service time has reached some arbitrarily chosen fraction of calculated or experimental failure life Previous failure statistics indicate high probability of failure Frequency of repair renders continued operation uneconomical Nondestructive inspection reveals cracking Surface degradation from corrosion, including coating degradation, is excessive Grain-boundary attack and/or pitting by oxidation/hot corrosion, is excessive Foreign object damage Destructive sampling and testing indicate life exhaustion Excessive deformation has occurred due to creep, causing distortion and unfavorable changes in Sudden and complete fracture occurs References cited in this section 1. J.J. Duga et al, The Economic Effects of fracture in the United States, Report to the National Bureau of Standards. Battelle Columbus Laboratories. March 1983 2. L.F. Coffin, Damage Evaluation and Life Prediction for High-Temperature Gas Turbine Materials 2382-3, Electric Power Research Institute, April 1986,p 1. 1-1/e Moterials,EPRI AP-4477,Project Proc. Conf on Life Prediction for High Temperature Gas Turbin 3. R. Viswanathan andR. B Dooley, Creep Life Assessment Techniques for Fossil Power Plant Boiler Pressure Parts, Proc. Conf on Life Prediction for High Temperature Gas Turbine Materials, EPRI AP 4477, Project 2382-3, Electric Power Research Institute, April 1986, p 2. 1-2.28 Structural Design Philosophies Historic Failures. It is often stated that history repeats itself. Yet, when it comes to structural components and equipment, structural designers, original equipment manufacturers (OEMs), and users do not want a repeat of history. The consequences and costs of fractured, cracked, corroded, and malfunctioned equipment are unwanted. Through the years, history has demonstrated that failures occur; history has also shown that the engineering communities have responded to prevent failure from occurring again. Table 1(Ref 4, 5, 6,7,8,9,10, 11, 12, 13, 14, 15)identifies some of the historic structural failures that have occurred in the 20th century. These historic failures as well as other failures have revolutionized design philosophies, inspection techniques and practices, material development, and material processing and controls and have redefined the criteria for failure. Furthermore, the pursuit of understanding how and why these ailures occurred have resulted in the development of structural-integrity programs, enhanced analytical modeling and prediction techniques, accurate life assessment methods, and a fortified commitment to avoid the recurrence of these failures through improved designs. The examples cited in Table I were serious and often tragic failures that had a great impact on structural designs and life assessment developments. However, not all failures or malfunctions of equipment is as pivotal in history as those mentioned in Table 1. Yet, it is emphasized that any failure, no matter how seemingly asignificant, should be investigated and the findings used to improve the design and increase the life and reliability of at component or equipment Thefileisdownloadedfromwww.bzfxw.com
to determine if they are suitable, reliable, and economical for continual service. If deemed unreliable, the equipment or structure may be repaired, refurbished, or replaced. In any component the failure criteria need to be defined and established. Failure does not always involve fracture or rupture. Progressive damage of structure and components under operating conditions leads to exhaustion of life, thus leading to failure. Damage may be defined as a “progressive and cumulative change acting to degrade the structural performance of the load-bearing component or components which make up the plant” (Ref 2). Life may be defined as the “period during which a component can perform its intended function safely, reliably, and economically” (Ref 3). With some modifications, the definitions used by Viswanathan and Dooley for fossil-fuel power steam plant components (Ref 3) can be used to define failure and life of components; that is, component life is expended when: · Design life has elapsed. · Calculations predict life exhaustion. · Service time has reached some arbitrarily chosen fraction of calculated or experimental failure life. · Previous failure statistics indicate high probability of failure. · Frequency of repair renders continued operation uneconomical. · Nondestructive inspection reveals cracking. · Surface degradation from corrosion, including coating degradation, is excessive. · Grain-boundary attack and/or pitting by oxidation/hot corrosion, is excessive. · Foreign object damage is severe. · Destructive sampling and testing indicate life exhaustion. · Excessive deformation has occurred due to creep, causing distortion and unfavorable changes in clearances. · Sudden and complete fracture occurs. References cited in this section 1. J.J. Duga et al., The Economic Effects of Fracture in the United States, Report to the National Bureau of Standards, Battelle Columbus Laboratories, March 1983 2. L.F. Coffin, Damage Evaluation and Life Prediction for High-Temperature Gas Turbine Materials, Proc. Conf. on Life Prediction for High Temperature Gas Turbine Materials, EPRI AP-4477, Project 2382-3, Electric Power Research Institute, April 1986, p 1.1–1.17 3. R. Viswanathan and R.B. Dooley, Creep Life Assessment Techniques for Fossil Power Plant Boiler Pressure Parts, Proc. Conf. on Life Prediction for High Temperature Gas Turbine Materials, EPRI AP- 4477, Project 2382-3, Electric Power Research Institute, April 1986, p 2.1–2.28 Structural Design Philosophies Historic Failures. It is often stated that history repeats itself. Yet, when it comes to structural components and equipment, structural designers, original equipment manufacturers (OEMs), and users do not want a repeat of history. The consequences and costs of fractured, cracked, corroded, and malfunctioned equipment are unwanted. Through the years, history has demonstrated that failures occur; history has also shown that the engineering communities have responded to prevent failure from occurring again. Table 1 (Ref 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15) identifies some of the historic structural failures that have occurred in the 20th century. These historic failures as well as other failures have revolutionized design philosophies, inspection techniques and practices, material development, and material processing and controls and have redefined the criteria for failure. Furthermore, the pursuit of understanding how and why these failures occurred have resulted in the development of structural-integrity programs, enhanced analytical modeling and prediction techniques, accurate life assessment methods, and a fortified commitment to avoid the recurrence of these failures through improved designs. The examples cited in Table 1 were serious and often tragic failures that had a great impact on structural designs and life assessment developments. However, not all failures or malfunctions of equipment is as pivotal in history as those mentioned in Table 1. Yet, it is emphasized that any failure, no matter how seemingly insignificant, should be investigated and the findings used to improve the design and increase the life and reliability of that component or equipment. The file is downloaded from www.bzfxw.com
Table 1 Historic failures and their impact on life assessment concerns Failure I Year Reason for failu Life assessment developments Titanic(Ref 4) 1912 Ship hits iceberg and watertight Improvement in steel grades compartments rupture Safety procedures established for lifeboats Warning systems established for Molasses Tank Failures 1919, Brittle fracture of the tank as a result of poor Design codes for storage tanks (Ref 5) 1973 ductility and higher loads Consideration give to causes for Tacoma Bridge Failure 1940 Aerodynamic instability and failure caused Sophisticated analytical models (Ref 6, 7 by wind vortices and bridge desi developed for resonance Bridge design changed to account for World War II Liberty 1942- 1289 of the 4694 warships suffered brittle Selection of increased toughness ships (ref 8) 1952 fracture or structure failure at the welded mater Improved fabrication practices Development of fracture mechanics Liquefied natural gas 1944 Failure and explosion of an LNG pressure Selection and development of (LNG) storage tank vessel due to a possible welding defect and materials with improved toughness at (Ref 9) improperly heat treated material resulting in the service temperature of-160C( obsequent fatigue crack growth 250°F) Comet aircraft failures 1950s Fatigue crack initiation in pressurized skins Development of the fatigue"safe-life" (Ref 10) due to high gross stresses and stress concentration effects from geometric features Evaluation of the effects of geometry Evaluation of the effects of stiffeners on stress distribution Establishment of aircraft structural gram(ASIP)in 1958 F-111 Aircraft No 94 1969 Fatigue failure due to material defect in Improved inspection techniques wing pivot fitting(Ref high-strength steel l1) damage-tolerant design philosophy relopment of materials with Seam-welded high- 1986- Cavitation and creep voids in welds Development of elevated-temperature energy piping failures 2000 resulting in catastrophic high-energy rupture life assessment techniques for (Ref 12 cavitation and creep failure Aloha Incident, Boeing 1988 Accelerated corrosion and multiple fatigue Improved aircraft maintenance and 737(Ref13) crack-initiation sites in riveted fuselage skin inspection procedures Life assessment methods developed Sioux City Incident 1989 Hard alpha case present in titanium fan disk Increased process controls on (Ref 14) resulted in fatigue crack initiation and processing of titanium ingots catastrophic failure
Table 1 Historic failures and their impact on life assessment concerns Failure Year Reason for failure Life assessment developments Titanic (Ref 4) 1912 Ship hits iceberg and watertight compartments rupture. Improvement in steel grades Safety procedures established for lifeboats Warning systems established for icebergs Molasses Tank Failures (Ref 5) 1919, 1973 Brittle fracture of the tank as a result of poor ductility and higher loads Design codes for storage tanks developed Consideration given to causes for brittle fracture Tacoma Bridge Failure (Ref 6, 7 1940 Aerodynamic instability and failure caused by wind vortices and bridge design Sophisticated analytical models developed for resonance Bridge design changed to account for aerodynamic conditions World War II Liberty ships (Ref 8) 1942– 1952 1289 of the 4694 warships suffered brittle fracture or structure failure at the welded steel joints. Selection of increased toughness material Improved fabrication practices Development of fracture mechanics Liquefied natural gas (LNG) storage tank (Ref 9) 1944 Failure and explosion of an LNG pressure vessel due to a possible welding defect and improperly heat treated material resulting in subsequent fatigue crack growth Selection and development of materials with improved toughness at the service temperature of -160 °C (- 250 °F) Comet aircraft failures (Ref 10) 1950s Fatigue crack initiation in pressurized skins due to high gross stresses and stress concentration effects from geometric features Development of the fatigue “safe-life” approach Evaluation of the effects of geometry and notches on fatigue behavior Evaluation of the effects of stiffeners on stress distribution Establishment of aircraft structural integrity program (ASIP) in 1958 F-111 Aircraft No. 94 wing pivot fitting (Ref 11) 1969 Fatigue failure due to material defect in high-strength steel Improved inspection techniques Change from fatigue “safe-life” to damage-tolerant design philosophy Development of materials with improved toughness Seam-welded highenergy piping failures (Ref 12) 1986– 2000 Cavitation and creep voids in welds resulting in catastrophic high-energy rupture Development of elevated-temperature life assessment techniques for cavitation and creep failure Aloha Incident, Boeing 737 (Ref 13) 1988 Accelerated corrosion and multiple fatigue crack-initiation sites in riveted fuselage skin Improved aircraft maintenance and inspection procedures Life assessment methods developed for multiple-site damage (MSD) Sioux City Incident (Ref 14) 1989 Hard alpha case present in titanium fan disk resulted in fatigue crack initiation and catastrophic failure. Increased process controls on processing of titanium ingots
Failure Year Reason for failure Life assessment developments Development of probabilistic design and assessment using dedicated computer rograms for titanium disks Earthquakes in Kobe 1994, Failure occurred in I-beams and columns Development of earthquake resistant City, Japan, and 1995 due to joint configuration and welding structures Northridge. California practices that resulted in low ductility of the (Ref 15) Improved joint designs and weldi practices for structural steels ding Improved controls on steel Overview of the Design Process. Because of failures similar to those in Table 1, predicting performance and assessing the remaining life with greater confidence becomes increasingly important as costs for manufacturers and operators need to be reduced. Furthermore, the cost of failure is progressively greater as systems become more complex, downtime costs increase, and liability for failure increases. a brief discussion follows on the design process because it is important for failure investigators and life assessment engineers to understand some of the design issues. Each structure has unique design requirements, but all structures are designed using some basic design principles. Figure 1 illustrates the relationship among the design phase, testing, systematic failure analysis, and life assessment of components Thefileisdownloadedfromwww.bzfxw.com
Failure Year Reason for failure Life assessment developments Development of probabilistic design approach and analytical life assessment using dedicated computer programs for titanium disks Earthquakes in Kobe City, Japan, and Northridge, California (Ref 15) 1994, 1995 Failure occurred in I-beams and columns due to joint configuration and welding practices that resulted in low ductility of the steel. Development of earthquake resistant structures Improved joint designs and welding practices for structural steels Improved controls on steel manufacture Overview of the Design Process. Because of failures similar to those in Table 1, predicting performance and assessing the remaining life with greater confidence becomes increasingly important as costs for manufacturers and operators need to be reduced. Furthermore, the cost of failure is progressively greater as systems become more complex, downtime costs increase, and liability for failure increases. A brief discussion follows on the design process because it is important for failure investigators and life assessment engineers to understand some of the design issues. Each structure has unique design requirements, but all structures are designed using some basic design principles. Figure 1 illustrates the relationship among the design phase, testing, systematic failure analysis, and life assessment of components. The file is downloaded from www.bzfxw.com
Materials and full-scale testing Design criteria Design analysis Material selection and life predictions Manufacturing and fabrication Structural component in-service In-service failure Structural aging or malfunction fit service concern Nondestructive Field perating inspections examinations conditions Mechanical Microscopic Fracture Laboratory Deposit and and physical examinations aminations surface chemical property test analysIs analysis Materials characterized Failure mechanism identified Environmental factors established ComponentFracture Engineering StructuralIOperations testing mechanics analyses analysis evaluation Root cause determined Fit for service Not fit for service Corrective Identify next Repair or Remove Replace with nspection interval I restrict service from service new structure ig. 1 Flow diagram showing the relationship between the design phase and the investigative tasks for in-service failure structural aging and fitness-for-service of structural components One alternative for avoiding failures used in the past was to overdesign and to operate overconservatively. The economic nalties for both are increasingly significant; however, the economic penalties for failures are significant as well. It is
Fig. 1 Flow diagram showing the relationship between the design phase and the investigative tasks for in-service failure, structural aging, and fitness-for-service of structural components One alternative for avoiding failures used in the past was to overdesign and to operate overconservatively. The economic penalties for both are increasingly significant; however, the economic penalties for failures are significant as well. It is
necessary, then, to pay more attention to predicting and ensuring performance Predicting and ensuring performance is fundamentally a part of the design process for buildings, power plants, aircraft, refineries, and ships( Ref 16) For any given design, the mission and the intended use are established Predicting the performance and design life of a component depends on defining what life or performance is required for a given duration while the component operates generally in combinations of mechanical and chemical environments. Defining performance may involve defining end points such as: acceptable length of propagating cracks, maximum depth of propagating pits, acceptable remaining thickness of corroding pipes, maximum number of fatigue cycles or extent of cumulative damage, maximum number of plugged tubes, maximum number of failed circuits, maximum leakage, or appearance of a maximum area or number of rust spots. Defining such end points is a critical part of predicting life since prediction defines when these end points will be reached and therefore when"failure occurs Defining failure is also related to what is meant by the"design life. For example, for the aerospace industry, an airplane may be designed for 8000 flight hours and analyzed for two lifetimes or 16,000 flight hours. For the power industry, the design life of components is sometimes taken as 40 years. This means that the equipment is expected to perform satisfactorily at its rated output for 40 years. This is not to say that some maintenance is not necessary. However, to assert to a customer that a component has a 40 year design life it is necessary to develop bases for such a claim. Such bases are usually provided by analyses and by accelerated testing in the laboratory and with prototype and model testing As part of the life assessment process, it is important to understand how a structural component-whether a pressure vessel, shaft, or structural member-is designed in order to understand how it may fail and to perform meaningful life assessment. For example, the first step in the design of any pressure vessel is to select the proper design code based on its intended use. For example, a pressure vessel may be a power or heating boiler, a nuclear reactor chamber, a chemical process chamber, a hydrostatic test chamber used to test underwater equipment, or a pressure vessel for human occupancy. Once the intended use is identified, the appropriate design code can be selected. For example, pressure vessels use codes provided by many organizations and certifying agencies, such as the American Society of Mechanical Engineers(ASME), the American Bureau of Shipping(ABS), and European agencies have pressure vessel design codes Strict adherence to these codes for the design, fabrication, testing, and quality control and assurance allows the finished pressure vessel to be certified by the appropriate authorizing agency( ref 17) One of the first incentives to develop a pressure vessel code occurred after the Boston Molasses tank incident in 1919 when the tank failed by overstress, consequently releasing more than 2 million gallons of molasses and resulting in the loss of life and property(Ref 5 ). Even after that catastrophic failure and understanding the nature of the failure, another molasses tank failure occurred in New Jersey in 1973. Figure 2 shows the destruction caused by the molasses tank incident. These molasses tank incidences demonstrate how important it is to prevent failures, and it underscores that good designs consider the operating conditions and limitations of materials of construction Thefileisdownloadedfromwww.bzfxw.com
necessary, then, to pay more attention to predicting and ensuring performance. Predicting and ensuring performance is fundamentally a part of the design process for buildings, power plants, aircraft, refineries, and ships (Ref 16). For any given design, the mission and the intended use are established. Predicting the performance and design life of a component depends on defining what life or performance is required for a given duration while the component operates generally in combinations of mechanical and chemical environments. Defining performance may involve defining end points such as: acceptable length of propagating cracks, maximum depth of propagating pits, acceptable remaining thickness of corroding pipes, maximum number of fatigue cycles or extent of cumulative damage, maximum number of plugged tubes, maximum number of failed circuits, maximum leakage, or appearance of a maximum area or number of rust spots. Defining such end points is a critical part of predicting life since prediction defines when these end points will be reached and therefore when “failure” occurs. Defining failure is also related to what is meant by the “design life.” For example, for the aerospace industry, an airplane may be designed for 8000 flight hours and analyzed for two lifetimes or 16,000 flight hours. For the power industry, the design life of components is sometimes taken as 40 years. This means that the equipment is expected to perform satisfactorily at its rated output for 40 years. This is not to say that some maintenance is not necessary. However, to assert to a customer that a component has a 40 year design life, it is necessary to develop bases for such a claim. Such bases are usually provided by analyses and by accelerated testing in the laboratory and with prototype and model testing. As part of the life assessment process, it is important to understand how a structural component—whether a pressure vessel, shaft, or structural member—is designed in order to understand how it may fail and to perform meaningful life assessment. For example, the first step in the design of any pressure vessel is to select the proper design code based on its intended use. For example, a pressure vessel may be a power or heating boiler, a nuclear reactor chamber, a chemical process chamber, a hydrostatic test chamber used to test underwater equipment, or a pressure vessel for human occupancy. Once the intended use is identified, the appropriate design code can be selected. For example, pressure vessels use codes provided by many organizations and certifying agencies, such as the American Society of Mechanical Engineers (ASME), the American Bureau of Shipping (ABS), and European agencies have pressure vessel design codes. Strict adherence to these codes for the design, fabrication, testing, and quality control and assurance allows the finished pressure vessel to be certified by the appropriate authorizing agency (Ref 17). One of the first incentives to develop a pressure vessel code occurred after the Boston Molasses tank incident in 1919 when the tank failed by overstress, consequently releasing more than 2 million gallons of molasses and resulting in the loss of life and property (Ref 5). Even after that catastrophic failure and understanding the nature of the failure, another molasses tank failure occurred in New Jersey in 1973. Figure 2 shows the destruction caused by the molasses tank incident. These molasses tank incidences demonstrate how important it is to prevent failures, and it underscores that good designs consider the operating conditions and limitations of materials of construction. The file is downloaded from www.bzfxw.com
Fig 2 Failed molasses tank, which fractured suddenly in New Jersey in March 1973. This catastrophic and sudden brittle fracture resulted in the release of the molasses in the tank similar to the boston molasses tank disaster in 1919 The next step in the design process is to identify the design parameters, such as configuration, design pressure, and so forth, Table 2 presents an example of a design parameter list applicable for a chemical process chamber(Ref 17). These design parameters are the same parameters considered when conducting a pressure vessel failure investigation and life Table 2 Pressure vessel design parameters Required design code) Penetration and location requirements Basic chamber configuration(Cylindrical or Contents and/or process within the pressure vessel spherical; flat, spherical, or elliptical end details, Internal volume capacity Estimated operational pressure and temperature cycle history (number of cycles at what pressures and temperatures over the Minimum inside diameter Piping, external and internal attachment requirements Minimum inside length Test chamber surroundings(enclosed in building or exposed to Chamber orientation(for cylindrical chambers, Test chamber physical geographical location longitudinal axis vertical or horizontal Support configuration(saddle supports, bottom Vessel special material requirements(vessel material other than cylindrical skirt, legs, etc Maximum internal operating pressure Vessel protection requirements(painted surfaces, stainless steel overlays at seals, cathodic protection, etc Maximum external operating pressure(vacuum,Fabrication requirements Design operating temperature range Material selection (a) AsMe Boiler and Pressure Vessel Code, Section VIl, Div. 1, 2, or 3; ABS; other Given the design parameters, the proper material(s)is selected for the structural component. Safety and economy are often the governing factors when selecting a material for pressure vessels. The material is selected based on its mechanical, corrosion, creep, toughness, and thermal properties as applicable. If necessary, the appropriate weld material is selected based on the chosen base material. Material is assigned an allowable stress value based on its ultimate and yield strengths and operating temperature range. This allowable stress value is then used in design equations or compared to results obtained from detailed analyses The design process then proceeds with the determination of the sizes and/or thickness of the various components. The design process is completed with the creation of the engineering and fabrication drawings. These drawings should include the dimensional information, but also specify materials, weld identification, weld procedures, and required weld inspections. Other helpful information to include on the drawings is basic parameters such as design pressure, design temperature range, design code, and other information deemed necessary for the particular structural component Structural Design Approaches. The criteria of failure are determined by the strength of materials, fracture toughness, creep resistance, fatigue behavior, and the corrosion resistance of materials. These are briefly discussed in this section Strength of Materials. In the strength-of-materials design approach one typically has a specific structural geometry (assumed to be defect free) for which the load-carrying capacity must be determined. To accomplish this, a calculation is first made to determine the relation between the load and the maximum stress that exists in the structure. The maximum stress so determined is then compared with the strength of the material. An acceptable design is achieved when the maximum stress is less than the strength of the material, suitably reduced by a factor of safety It can be assumed that failure will not occur unless omax exceeds the yield strength of the material, oY. To ensure this, a factor of safety (S) can be introduced to account for material variability and/or unanticipated greater service loading. The strength-of-materials approach is a good approach for materials with no defects and simple structures. Figure 3 shows the strength-of-materials approach and the engineering design regime based on a factor of safety
Fig. 2 Failed molasses tank, which fractured suddenly in New Jersey in March 1973. This catastrophic and sudden brittle fracture resulted in the release of the molasses in the tank similar to the Boston Molasses tank disaster in 1919. The next step in the design process is to identify the design parameters, such as configuration, design pressure, and so forth. Table 2 presents an example of a design parameter list applicable for a chemical process chamber (Ref 17). These design parameters are the same parameters considered when conducting a pressure vessel failure investigation and life assessment. Table 2 Pressure vessel design parameters Required design code(a) Penetration and location requirements Basic chamber configuration. (Cylindrical or spherical; flat, spherical, or elliptical end details; etc.) Contents and/or process within the pressure vessel Internal volume capacity Estimated operational pressure and temperature cycle history (number of cycles at what pressures and temperatures over the vessel's lifetime) Minimum inside diameter Piping, external and internal attachment requirements Minimum inside length Test chamber surroundings (enclosed in building or exposed to elements) Chamber orientation (for cylindrical chambers, longitudinal axis vertical or horizontal) Test chamber physical geographical location Support configuration (saddle supports, bottom cylindrical skirt, legs, etc.) Vessel special material requirements (vessel material other than carbon steel, internal cladding, etc.) Maximum internal operating pressure Vessel protection requirements (painted surfaces, stainless steel overlays at seals, cathodic protection, etc.) Maximum external operating pressure (vacuum, etc.) Fabrication requirements Design operating temperature range Material selection (a) ASME Boiler and Pressure Vessel Code, Section VII, Div. 1, 2, or 3; ABS; other Given the design parameters, the proper material(s) is selected for the structural component. Safety and economy are often the governing factors when selecting a material for pressure vessels. The material is selected based on its mechanical, corrosion, creep, toughness, and thermal properties as applicable. If necessary, the appropriate weld material is selected based on the chosen base material. Material is assigned an allowable stress value based on its ultimate and yield strengths and operating temperature range. This allowable stress value is then used in design equations or compared to results obtained from detailed analyses. The design process then proceeds with the determination of the sizes and/or thickness of the various components. The design process is completed with the creation of the engineering and fabrication drawings. These drawings should include the dimensional information, but also specify materials, weld identification, weld procedures, and required weld inspections. Other helpful information to include on the drawings is basic parameters such as design pressure, design temperature range, design code, and other information deemed necessary for the particular structural component. Structural Design Approaches. The criteria of failure are determined by the strength of materials, fracture toughness, creep resistance, fatigue behavior, and the corrosion resistance of materials. These are briefly discussed in this section. Strength of Materials. In the strength-of-materials design approach one typically has a specific structural geometry (assumed to be defect free) for which the load-carrying capacity must be determined. To accomplish this, a calculation is first made to determine the relation between the load and the maximum stress that exists in the structure. The maximum stress so determined is then compared with the strength of the material. An acceptable design is achieved when the maximum stress is less than the strength of the material, suitably reduced by a factor of safety. It can be assumed that failure will not occur unless σmax exceeds the yield strength of the material, σY. To ensure this, a factor of safety (S) can be introduced to account for material variability and/or unanticipated greater service loading. The strength-of-materials approach is a good approach for materials with no defects and simple structures. Figure 3 shows the strength-of-materials approach and the engineering design regime based on a factor of safety
Elastic-plastie Linear elastic fracture mechanics fracture mechanics behavior 、 regime 51/s Engineering design Strength-of-materials behavior 1/s Stress ratio(omaxoy) Fig.3A general plot of the ratios of the toughness and stress showing the relationship between linear elastic fracture mechanics and strength of materials as it relates to fracture and structural integrity(ref 18) Linear Elastic Fracture Mechanics. Brittle fractures. similar to those mentioned in Table 1. are avoided using a linear elastic fracture mechanics design approach. This approach considers that the structure, instead of being defect-free, contains a crack(Ref 18). The governing structural mechanics parameter when a crack is present, at least in the linear approach, is an entity called the stress-intensity factor. This parameter, which is conventionally given the symbol K,can be determined from a mathematical analysis similar to that used to obtain the stresses in an uncracked component. For a relatively small crack in a simple structure, an analysis of the flawed structure beam would give to a reasonable approximation K=1.12√ma (Eq1) where a is the depth of the bracket and omax is the stress that would occur at the crack location in the absence of the crack The basic relation in fracture mechanics is one that equates K to a critical value. This critical value is often taken as a property of the material called the plane-strain fracture toughness, conventionally denoted as Kle. When equality is achieved between K and Klc, the crack is presumed to grow in an uncontrollable manner. Hence, the structure can be designed to be safe from fracture by ensuring that K is less than Kle. The liberty warship fractures are a classic example of a structural failure caused by Klc exceeding K and uncontrolled crack growth The essential difference from the strength of materials approach is that the fracture mechanics approach explicitly introduces a new physical parameter: the size of a(real or postulated)cracklike flaw. In fracture mechanics the size of a crack is the dominant structural parameter. It is the specification of this parameter that distinguishes fracture mechanics from conventional failure analyses The generalization of the basis for engineering structural-integrity assessments that fracture mechanics provides is portrayed in terms of the failure boundary shown in Fig. 3. Clearly, fracture mechanics considerations do not eliminate the traditional approach. Structures using reasonably tough materials(high Klc)and having only small cracks (low K)will lie in the strength of materials regime. Conversely, if the material is brittle(low Klc)and strong(high oY), the presence of even a small crack is likely to trigger fracture. The fracture mechanics assessment is then the crucial one The special circumstances that would be called into play in the upper right-hand corner of Fig. 3 are worth noting. In this regime, a cracked structure would experience large-scale plastic deformation prior to crack extension. Additional information is provided in the article Failure Assessment Diagrams"in this Volume and in Ref 19 Damage Tolerance Approach Life assessment of aircraft and power-plant equipment stems largely from the development of the damage-tolerance philosophy based on fracture mechanics. Damage tolerance is the philosophy used for maintaining the structural safety of commercial transport, military aircraft, structures, and pressure vessels. The use of fracture mechanics and damage tolerance has evolved into the design program for structures that are damage tolerant, that is, designed to operate with manufacturing and in-service-induced defects(Ref 20) Thefileisdownloadedfromwww.bzfxw.com
Fig. 3 A general plot of the ratios of the toughness and stress showing the relationship between linear elastic fracture mechanics and strength of materials as it relates to fracture and structural integrity (Ref 18) Linear Elastic Fracture Mechanics. Brittle fractures, similar to those mentioned in Table 1, are avoided using a linear elastic fracture mechanics design approach. This approach considers that the structure, instead of being defect-free, contains a crack (Ref 18). The governing structural mechanics parameter when a crack is present, at least in the linear approach, is an entity called the stress-intensity factor. This parameter, which is conventionally given the symbol K, can be determined from a mathematical analysis similar to that used to obtain the stresses in an uncracked component. For a relatively small crack in a simple structure, an analysis of the flawed structure beam would give to a reasonable approximation: max K a = 1.12s p (Eq 1) where α is the depth of the bracket and σmax is the stress that would occur at the crack location in the absence of the crack. The basic relation in fracture mechanics is one that equates K to a critical value. This critical value is often taken as a property of the material called the plane-strain fracture toughness, conventionally denoted as KIc. When equality is achieved between K and KIc, the crack is presumed to grow in an uncontrollable manner. Hence, the structure can be designed to be safe from fracture by ensuring that K is less than KIc. The Liberty warship fractures are a classic example of a structural failure caused by KIc exceeding K and uncontrolled crack growth. The essential difference from the strength of materials approach is that the fracture mechanics approach explicitly introduces a new physical parameter: the size of a (real or postulated) cracklike flaw. In fracture mechanics the size of a crack is the dominant structural parameter. It is the specification of this parameter that distinguishes fracture mechanics from conventional failure analyses. The generalization of the basis for engineering structural-integrity assessments that fracture mechanics provides is portrayed in terms of the failure boundary shown in Fig. 3. Clearly, fracture mechanics considerations do not eliminate the traditional approach. Structures using reasonably tough materials (high KIc) and having only small cracks (low K) will lie in the strength of materials regime. Conversely, if the material is brittle (low KIc) and strong (high σY), the presence of even a small crack is likely to trigger fracture. The fracture mechanics assessment is then the crucial one. The special circumstances that would be called into play in the upper right-hand corner of Fig. 3 are worth noting. In this regime, a cracked structure would experience large-scale plastic deformation prior to crack extension. Additional information is provided in the article “Failure Assessment Diagrams” in this Volume and in Ref 19. Damage Tolerance Approach. Life assessment of aircraft and power-plant equipment stems largely from the development of the damage-tolerance philosophy based on fracture mechanics. Damage tolerance is the philosophy used for maintaining the structural safety of commercial transport, military aircraft, structures, and pressure vessels. The use of fracture mechanics and damage tolerance has evolved into the design program for structures that are damage tolerant, that is, designed to operate with manufacturing and in-service-induced defects (Ref 20). The file is downloaded from www.bzfxw.com
Damage-tolerance evaluation has been interpreted in the past as a means to allow continued safe operation in the presence of known cracking. This interpretation is incorrect. No regulations allow the strength of the structure to be knowingly degraded below ultimate strength(1.5 x limit). The damage-tolerance evaluation is merely a means of providing an inspection program for a structure that is not expected to crack under normal circumstances, but may crack in service due to inadvertent circumstances. If cracks are found in primary structure, they must be repaired. The only allowable exception is through an engineering evaluation, which must show that the strength of the structure will never be degraded below ultimate strength operations or in-service conditions After many major fatigue failures in the 1950s on both military and commercial aircraft, the most notable of which were the DeHavilland Comet failures in early 1954, the U.S. Air Force (USAF) initiated the Aircraft Structural Integrity Program(ASIP)in 1958(Ref 10). The fatigue methodology adopted in the AsIP was the reliability approach, which became known as the"safe-life method This safe-life approach, used in the development of USaF aircraft in the 1960s involved analysis and testing to four times the anticipated service life. On the commercial scene, another philosophy, " fail safety, was introduced in the early 1960s, and a choice between safe -life and fail-safe methods was allowed by commercial airworthiness requirements. However, it was found that the safe-life method did not prevent fatigue cracking o thin the service life, even though the aircraft were tested to four lifetimes to support one service life(i.e, scatter factor of 4 ) One notable example is the F-lll aircraft 94, which crashed in 1969(Ref 11). The F-11l aircraft had a safe- life of 4000 flight hours. However, a material defect caused the F-lll aircraft, which used high-strength steel (ultimate tensile strength of 1655 to 1793 MPa, or 240 to 260 ksi, toughness of about 66 MPa Vm, or 60 ksi vin )for the wing box(Fig 4). The defect was not observed during inspection, and a fatigue crack initiated and grew for only about 0.38 mm(0.015 in. ) The aircraft was flown for 107 flights safely, at which time catastrophic failure occurred, causing the destruction of the aircraft stayed with wing stayed with airplane Wing pivot fitting Area of anomaly (b 0.905in. Brittle fracture 0.236 in deer 0.282 in thick 0.015 in. fatigue crack growth Material anomaly
Damage-tolerance evaluation has been interpreted in the past as a means to allow continued safe operation in the presence of known cracking. This interpretation is incorrect. No regulations allow the strength of the structure to be knowingly degraded below ultimate strength (1.5 × limit). The damage-tolerance evaluation is merely a means of providing an inspection program for a structure that is not expected to crack under normal circumstances, but may crack in service due to inadvertent circumstances. If cracks are found in primary structure, they must be repaired. The only allowable exception is through an engineering evaluation, which must show that the strength of the structure will never be degraded below ultimate strength operations or in-service conditions. After many major fatigue failures in the 1950s on both military and commercial aircraft, the most notable of which were the DeHavilland Comet failures in early 1954, the U.S. Air Force (USAF) initiated the Aircraft Structural Integrity Program (ASIP) in 1958 (Ref 10). The fatigue methodology adopted in the ASIP was the reliability approach, which became known as the “safe-life” method. This safe-life approach, used in the development of USAF aircraft in the 1960s, involved analysis and testing to four times the anticipated service life. On the commercial scene, another philosophy, “fail safety,” was introduced in the early 1960s, and a choice between safe-life and fail-safe methods was allowed by commercial airworthiness requirements. However, it was found that the safe-life method did not prevent fatigue cracking within the service life, even though the aircraft were tested to four lifetimes to support one service life (i.e., scatter factor of 4). One notable example is the F-111 aircraft 94, which crashed in 1969 (Ref 11). The F-111 aircraft had a safe-life of 4000 flight hours. However, a material defect caused the F-111 aircraft, which used high-strength steel (ultimate tensile strength of 1655 to 1793 MPa, or 240 to 260 ksi, toughness of about 66 MPa m , or 60 ksi in ) for the wing box (Fig. 4). The defect was not observed during inspection, and a fatigue crack initiated and grew for only about 0.38 mm (0.015 in.). The aircraft was flown for 107 flights safely, at which time catastrophic failure occurred, causing the destruction of the aircraft
Fig. 4 Fatigue cracking in an aircraft wing fitting for the f-lll Aircraft 94 that crashed in 1969.(a)and (b)Location of the left wing-pivot box fitting. The 22 mm(0.91 in material defect was not observed during inspection, and a fatigue crack initiated and grew for only about 0.38 mm(0.015 in. before unstable brittle fracture occurred The leak-before-break design approach is prevalent in pressure-containing equipment such as pressure vessels and piping used in the nuclear and fossil-fuel power-generation plants, refineries, and chemical plants. Failure analysis and life assessment of pressure-containing systems is essential. Although leak-before-break failures are not catastrophic, they are costly and can affect plant operations. Therefore, analyses are often performed to predict when the next internal or external inspection should be performed. Typical life-limiting mechanisms include stress-corrosion cracking, fatigue, and thermal fatigue. Welded structures that could initiate a crack are often susceptible to these mechanisms The leak-before-break concept generally refers to a pressure-contaminant system failure in which a part-through wall crack extends to become a through-wall crack, thus allowing fluid to escape. If no further crack extension occurs, then the loss of the fluid is detected and no further crack growth occurs. Alternatively, when a through-wall crack propagates along the wall, a catastrophic event can occur(Ref 18) Pow plant piping materials that are ductile, such as stainless steel and nickel-base alloys, often leak before break. Figure 5 shows small-bore, socket-welded piping that will initiate fatigue cracks at either the toe of the weld or the root of the weld. These ductile socket-welded pipes leak before catastrophic failure occurs(Ref 21) Axial leg clal weld toe Socket weld Fig. 5 Stainless steel piping such as small-bore piping is designed to leak before break. a fatigue crack either initiates at the toe or the root of the weld(a) Typical socket fitting with a fillet weld. (b) Micrograph of a cross section through a socket-welded joint showing fatigue crack that initiated from the weld root and extended through the weld.(c) Micrograph of through the pipe axia weld toe crack showing a fatigue-induced crack that extended Elevated-Temperature Concerns For elevated-temperature equipment and structures subjected to steady-state or cyclic stresses, the principal design considerations are creep control, oxidation prevention through the use of oxidation-resistant materials or coatings, and selection of materials that have good stress-rupture and creep properties. The criteria for failure is(1)to not go below a minimum stress-rupture strength for a given operating stress and temperature and(2)to not operate above a certain temperature that alters the microstructure or oxidizes the material Corrosion Allowances. Designs are configured such that the operating or loading stresses can be minimized for safe operations. It is necessary to consider the effects that an environment will have on the material; it is just as important as considering structural loads on a component. It is important that environments are known and controlled in such a way that corrosion is minimized on all surfaces. This means that designs consider effects of crevices, galvanic couples, flows stresses, and temperatures to ensure that all the surfaces of materials will be minimally degraded within the design life (Ref 16) A common design approach for pressure vessels and tanks to deal with corrosion is to provide a"corrosion allowance, which takes the form of additional thickness based on available information on rates of general corrosion over the design life. For example, a carbon steel vessel designed for 25 years of service in sulfuric acid at a corrosion rate of 5 mils/yr (0.005 in. yr) would have a corrosion allowance of 125 mils. However, such allowances cannot deal with stress-corrosion cracking, pitting, intergranular cracking, or effects of long-range cells. Use of a corrosion allowance can be disastrously misleading since its use suggests that all corrosion problems have been solved. It should be pointed out that exceeding the corrosion allowance does not necessarily mean the vessel would fail or is unsuitable for service. It is an indication that the vessel should be evaluated for continued service Thefileisdownloadedfromwww.bzfxw.com
Fig. 4 Fatigue cracking in an aircraft wing fitting for the F-111 Aircraft 94 that crashed in 1969. (a) and (b) Location of the left wing-pivot box fitting. The 22 mm (0.91 in.) material defect was not observed during inspection, and a fatigue crack initiated and grew for only about 0.38 mm (0.015 in.) before unstable brittle fracture occurred. The leak-before-break design approach is prevalent in pressure-containing equipment such as pressure vessels and piping used in the nuclear and fossil-fuel power-generation plants, refineries, and chemical plants. Failure analysis and life assessment of pressure-containing systems is essential. Although leak-before-break failures are not catastrophic, they are costly and can affect plant operations. Therefore, analyses are often performed to predict when the next internal or external inspection should be performed. Typical life-limiting mechanisms include stress-corrosion cracking, fatigue, and thermal fatigue. Welded structures that could initiate a crack are often susceptible to these mechanisms. The leak-before-break concept generally refers to a pressure-contaminant system failure in which a part-through wall crack extends to become a through-wall crack, thus allowing fluid to escape. If no further crack extension occurs, then the loss of the fluid is detected and no further crack growth occurs. Alternatively, when a through-wall crack propagates along the wall, a catastrophic event can occur (Ref 18). Power-plant piping materials that are ductile, such as stainless steel and nickel-base alloys, often leak before break. Figure 5 shows small-bore, socket-welded piping that will initiate fatigue cracks at either the toe of the weld or the root of the weld. These ductile socket-welded pipes leak before catastrophic failure occurs (Ref 21). Fig. 5 Stainless steel piping such as small-bore piping is designed to leak before break. A fatigue crack either initiates at the toe or the root of the weld. (a) Typical socket fitting with a fillet weld. (b) Micrograph of a cross section through a socket-welded joint showing fatigue crack that initiated from the weld root and extended through the weld. (c) Micrograph of axial weld toe crack showing a fatigue-induced crack that extended through the pipe wall Elevated-Temperature Concerns. For elevated-temperature equipment and structures subjected to steady-state or cyclic stresses, the principal design considerations are creep control, oxidation prevention through the use of oxidation-resistant materials or coatings, and selection of materials that have good stress-rupture and creep properties. The criteria for failure is (1) to not go below a minimum stress-rupture strength for a given operating stress and temperature and (2) to not operate above a certain temperature that alters the microstructure or oxidizes the material. Corrosion Allowances. Designs are configured such that the operating or loading stresses can be minimized for safe operations. It is necessary to consider the effects that an environment will have on the material; it is just as important as considering structural loads on a component. It is important that environments are known and controlled in such a way that corrosion is minimized on all surfaces. This means that designs consider effects of crevices, galvanic couples, flows, stresses, and temperatures to ensure that all the surfaces of materials will be minimally degraded within the design life (Ref 16). A common design approach for pressure vessels and tanks to deal with corrosion is to provide a “corrosion allowance,” which takes the form of additional thickness based on available information on rates of general corrosion over the design life. For example, a carbon steel vessel designed for 25 years of service in sulfuric acid at a corrosion rate of 5 mils/yr (0.005 in./yr) would have a corrosion allowance of 125 mils. However, such allowances cannot deal with stress-corrosion cracking, pitting, intergranular cracking, or effects of long-range cells. Use of a corrosion allowance can be disastrously misleading since its use suggests that all corrosion problems have been solved. It should be pointed out that exceeding the corrosion allowance does not necessarily mean the vessel would fail or is unsuitable for service. It is an indication that the vessel should be evaluated for continued service. The file is downloaded from www.bzfxw.com