Jim Chang, USAF Bolling AFB
Thomas Flournoy, Federal Aviation Administration
D. Gary Harlow, Lehigh Univ
Robert Piascik, NASA Langley Research Center
Robert Wei, Lehigh Univ
Air Force Office of Scientific Research
* Invited paper
8:30 AM INTRODUCTION AND SYMPOSIUM GOALS.
SESSION NN1: AGING AIRCRAFT I
Chairs: Jim C.I. Chang and Robert P. Wei
Monday Morning, November 30, 1998
Clarendon A/B (S)
8:45 AM *NN1.1
DAMAGE DETECTION CONSIDERATIONS FOR JET TRANSPORT STRUCTURES. Ulf G. Goranson , Structures Laboratories & Technology Standards, Boeing Commercial Airplane Group, Seattle, WA.
Criteria and procedures used in commercial jet transport design and manufacture over the last four decades have resulted in structures with a credible safety record. Advancements in the capability to characterize structural performance by analysis have spurred adaptation of traditional fatigue and fracture mechanics technologies with large test and service databases to achieve development of technology standards over the last 25 years. Major Boeing efforts have been focused on capturing lessons learned for future continuous design improvements. The first generation of Boeing jet transports is characterized by structural design principles addressing static strength and fail-safety. The fail-safe design concept is founded on redundancy, which indeed has provided sufficient detection opportunities for accidental damage and corrosion-related deterioration. Such redundancies are not always efficient, based on obvious damage detection considerations for aging airplanes subjected to fatigue damage at multiple-sites. Supplemental fatigue inspection programs were therefore implemented in the early 1980s to ensure sufficient fatigue damage detection opportunities. Development of structural performance data is the key element of the structural improvement process. These data come from full-scale testing, teardown inspections of in-service airplane structure, and fleet survey programs. Such structural performance data provide validation of the design for structural safety as well as for economic service performance. Although recent aging fleet concerns have focused unprecedented attention on such activities, it is important to recognize that such programs have been ongoing for the last 30 years. Timely damage detection is the key element for ensuring structural damage tolerance evolving from the fail-safe approach, which has worked well since the first generation of commercial airplanes. Structural maintenance and damage detection are the cornerstones of continuing airworthiness assurance. Inspection thresholds for most structures are based on traditional fatigue damage assessments, complemented by rogue flaw growth estimates for structural components with visually hidden details and lighted crack arrest capability. Residual strength and fatigue crack growth evaluations are combined with service based crack detection data to produce detection reliability representing multiple type and intervals of inspections in a fleet of airplanes subjected to exploratory inspections. Such data give operators freedom to adjust quantitatively their maintenance program in any manner that is desired as long as the required reliability of damage detection is preserved
9:15 AM NN1.2
AGING AIRFRAME MATERIALS: A MULTI-DISCIPLINARY ISSUE/FRETTING CORROSION TO CRACKING. Robert S. Piascik , Mechanics of Materials Branch, NASA-Langley Research Center, Hampton, VA.
Aircraft are complex structures constructed of aluminum alloys that are exposed to myriad conditions (variable loads and environments) over a life that can typically range from 30 to 50 years. As these structures age, the cumulative effects of load and environment increase the likelihood of damage. One such damage phenomenon is termed wide spread fatigue damage (WFD). Here, the simultaneous presence of fatigue cracks at multiple structural details may compromise the residual strength requirements of the structure. Research has shown that crack nucleation by fretting corrosion can accelerate the formation of fatigue cracks at multiple sites and thus increase the potential for early onset of WFD in the aging fuselage lap joint. Fretting corrosion is a result of repeated relative movement (micro-displacement) of the inner and outer layers of the fuselage lap joint. The highly localized damage is characterized by surface abrasion, oxide debris and surface microcracks that are sites for fatigue crack nucleation. Experimental data on fretting corrosion are presented to elucidate the processes of aging. From these studies, multidisciplinary research was conducted to develop improved detection methods for small fatigue cracks and fracture mechanics methods for predicting fatigue life.
9:30 AM NN1.3
AGING AIRFRAME MATERIALS: A MULTI-DISCIPLINARY ISSUE/FROM PITTING TO CRACKING. Robert P. Wei and D. Gary Harlow, Departament of Mechanical Engineering and Mechanics, Lehigh University, Bethlehem, PA.
Aging of airframe aluminum alloys during service can take the form of wide spread fatigue damage (WFD). In WFD, the simultaneous presence of fatigue cracks at multiple structural details may compromise the residual strength requirements of the structure. Research has shown that crack nucleation from sites of localized corrosion (such as pitting) can accelerate the formation of fatigue cracks at multiple sites and thus increase the potential for early onset of WFD in the aging fuselage lap joint. Severe pits are formed by galvanic corrosion attack of the matrix surrounding clusters of constituent particles in the alloys, which was induced by the particle-matrix couple. Fatigue cracking away from the nucleating severe corrosion pit undergo a regime of chemically-short crack growth, with anomalously higher growth rates than that to be expected from conventional long-crack data. Experimental data on pitting corrosion, crack nucleation and short- and long-crack growth are presented to elucidate the processes of aging. A multidisciplinary approach is outlined, including a preliminary probabilistic model for life prediction that integrates the individual processes. The significance of the early corrosion-related damage on service life and the scale of this damage in relation to NDE capabilities are described. The need for multidisciplinary research to develop improved detection methods for early corrosion damage and small fatigue cracks and mechanistically based probability methods for predicting fatigue life are discussed.
10:15 AM *NN1.4
CORROSION AND DEGRADATION PROCESSES ON METAL SURFACES. William H. Smyrl and Luis Garfias-Mesias, Corrosion Research Center, Department of Chemical Engineering and Materials Science, University of Minnesota, Minneapolis, MN.
Corrosion processes have a major financial impact on aerospace systems. Because of this, the detection of corrosion at early stages and the design of effective countermeasures are recognized as being crucial for protection and life extension of metal structures. Here we will discuss the general character of corrosion reactions on heterogeneous surfaces and the experimental techniques that have been developed to study them. The techniques generally fall into three categories: (1) methods to study the rate of corrosion reactions, (2) methods to locate breakdown sites at early stages or before breakdown occurs, and (3) methods to study the properties of the passive films on metals and how their stability is affected by aggressive environments. We will complete the presentation by describing the microscopy techniques that have been developed in our laboratory to obtain high resolution mapping around corrosion sites. The high resolution techniques may be used either in gas phase or liquid phase in aggressive environments and combine topographic imaging with chemical mapping. Examples of studies on aluminum aircraft alloys and on titanium alloys will be included.
10:45 AM *NN1.5
QNDE TECHNIQUES FOR DETECTION OF CRACKS AND CORROSION IN AIRCRAFT. J.D. Achenbach , Center for Quality Engineering and Failure Prevention, Northwestern University, Evanston, IL.
Methods of quantitative nondestructive evaluation (QNDE) are used to assess material deterioration of structures, and to detect and size discrete flaws. QNDE is particularly important for the in-service inspection of high-cost structures whose failure could have tragic consequences. Examples are aircraft, nuclear reactors and steel bridges. Trustworthy in-service inspection techniques are becoming increasingly needed as this country's civilian infrastructure, its transportation equipment and military platforms are becoming older but may have to be kept in service well into the next century. In this paper we present a brief overview of QNDE techniques with particular attention to problems of detectability and sizing of distributed (corrosion) and localized (cracking) damage. Some applications of ultrasonic techniques to civilian and military aircraft will be discussed in some detail. Particular attention will be devoted to detection and sizing of cracks and corrosion in the second or third layer of multilayered aircraft structures.
11:15 AM NN1.6
FRACTOGRAPHIC EXAMINATION OF CORROSION FATIGUE DAMAGE IN FULL-SCALE (C/KC-135 PANELS). Takao Kobayashi, Donald A. Shockey , Yvonne Wu, SRI International, Poulter Laboratory, Menlo Park, CA; Thomas H. Flournoy, Federal Aviation Administration, William J. Hughes Technical Center, Atlantic City International Airport, NJ.
SRI International is examining the multisite damage in C/KC-135 fuselage panels that were fatigue-tested full scale at the Foster-Miller facility under joint sponsorship by the Federal Aviation Administration and the Oklahoma City Air Logistics Center. Advanced fractographic analysis techniques were applied to selected cracks in attempting to infer post test the crack growth rates, nucleation times, and crack link-up behavior. The topographies of conjugate crack faces were mapped with a confocal optics scanning laser microscope. The topographs were then juxtaposed to approximate the position of the crack faces during the tests, and then incrementally displaced to simulate crack growth. The results were then correlated with the loading history to infer damage kinetics. Fast Fourier transform and wavelet analyses were applied to the elevation data to seek signature representations reflective of the loading conditions and microstructure. We attempt to establish a relationship between load conditions and failure behavior. The results are intended to assist development of computational models for predicting degradation in aging aircraft.
11:30 AM NN1.7
CORROSION AND CORROSION FATIGUE IN AIRFRAME MATERIALS: PROBABILITY OF OCCURRENCE VERSUS PROBABILITY OF DETECTION. D. Gary Harlow and Robert P. Wei, Lehigh Univ, Dept of Mechanical Engr and Mechanics, Bethlehem, PA.
Structural airworthiness requires the assurance of integrity and safety of an aircraft throughout its planned period of operation from its current state to the next scheduled inspection and maintenance. An accurate assessment depends critically on a quantitative approach that integrates information from state-of-the-art nondestructive evaluations with validated methods for prediction of damage accumulation and structural integrity assessments. Mechanistically based probability descriptions for the initiation, evolution, and geometric distribution of damage are not addressed adequately in current methodologies. The criticality, or structural significance, of the damage relative to its probability of detection by nondestructive evaluation techniques, likewise, is not well established. Herein, the probability of occurrence of damage versus its probability of detection is considered. The probability of occurrence is approximated from a model based upon plausible mechanisms for damage accumulation, or aging, in airframe materials. Specifically, a model, based on first principles, for pitting corrosion and corrosion fatigue cracking in aluminum alloys common to airframe structures, e.g. 2024-T3, is used to illustrate the critical functions of modeling and detecting structural damage. The model is employed to illustrate the difference between the probability of occurrence and the probability of detection of damage. The distribution for the probability of detection of damage depends on many factors including choice of nondestructive evaluation method, equipment calibration and sensitivity, acceptance and rejection criteria, component accessibility, surface condition, complexity of the structural component, inspection environment, and inspector expertise. A baseline distribution is used for comparison that reflects recent state-of-the-art nondestructive inspection based on eddy current systems. It is suggested that an appropriate target for nondestructive inspection methods would be flaw sizes of about 0.10 mm with a probability of detection of at least 90% versus the current capability of detecting flaw sizes between 0.75 and 1.27 mm with a probability of detection of only 50%.
1:30 PM *NN2.1
SESSION NN2: AGING AIRCRAFT II
Chairs: Jan D. Achenbach and Thomas H. Flournoy
Monday Afternoon, November 30, 1998
Clarendon A/B (S)
ASSESSMENT OF U.S. AIR FORCE AGING AIRCRAFT. Edward A. Davidson , John Lincoln, Lawrence Butkus, Aeronautical Systems Center, Wright-Patterson AFB, OH.
As a result of the National Research Council's 1997 report on the Aging of U.S. Air Force Aircraft, the Aeronautical Systems Center and the Air Force Research Laboratory have formed an Aging Aircraft Technologies Team (AATT) to address the structures research and management recommendations contained in the report. As a part of this activity, an in-depth survey of 33 Air Force systems was completed to first, determine the ``health of the fleet'' from a structural standpoint, and second, identify current and potential problems where structural research investment may provide reductions in the future maintenance cost of the systems. This briefing will provide an overview of current AATT activities, and a summary of the results of the completed surveys.
2:00 PM *NN2.2
APPLICATION OF NEW NDE TECHNOLOGIES FOR ASSESSING DAMAGE IN AIRCRAFT. William P. Winfree , NASA Langley Research Center, Hampton, VA.
A review is presented of several techniques under developed at NASA Langley Research Center for detection and quantification of flaws in aircraft structures. The techniques focus on the detection of cracks, corrosion and disbonds in thin laminated structures. The techniques range from thermal procedures which give a rapid indication of the regions of concern to eddy current instrumentation for detecting small cracks and multilayer corrosion. Results are presented on specimens with both manufactured defects for calibration of the techniques, specimens removed from aircraft and measurements in the field on aircraft.
2:30 PM *NN2.3
PROBABILITY OF DETECTION CURVES DETERMINED FROM HIT/MISS DATA FOR NONDESTRUCTIVE INSPECTIONS. Floyd W. Spencer , Sandia National Laboratories, Statistics and Human Factors Dept., Albuquerque, NM.
The reliability or effectiveness of a nondestructive inspection technique is often characterized by a probability of detection (POD) curve. The data available to estimate such curves are often in the form of hits and misses from the inspection of test samples with known flaw characteristics. POD curve estimation is based on binary regression, a technique with substantial statistical background. In this paper, we discuss a generalization of the usual model used in NDE characterization. The generalization allows the fitted POD curve more degrees of freedom and is based on observed inspection behaviors. The generalized model incorporates the probability of a false alarm into the POD curve estimation, as well as a probability of miss that is independent of flaw size. The model and the philosophy behind the model will be compared to other generalizations of POD, such as the probability of indication model. In light of the generalized model, we examine the use of confidence statements to characterize an inspection technique. The most commonly used confidence statement is to summarize NDE capability by giving the flaw size that has a 0.90 chance of detection with 95 percent confidence. Such confidence values are derived given a specific model for the data. We show how the use of such values can be misleading, especially if the inspection data are not adequately modeled. We illustrate parameter estimation of the proposed model and compare models using inspection data gathered on various NDE equipment. All such data were gathered using blind tests.
3:30 PM *NN2.4
TECHNOLOGIES FOR LIFE EXTENSION OF ADVANCED TURBINE ENGINE COMPONENTS. Stephan M. Russ , James M. Larsen, Charles F. Buynak, Air Force Research Laboratory-Materials and Manufacturing Directorate, Wright-Patterson Air Force Base, OH; Edmund H. Hindle, Pratt & Whitney, West Palm Beach, FL.
Under increasing economic pressure to reduce maintenance costs of advanced turbine engines, a study is being performed to assess opportunities to extend critical component lives through application of advances in technology. An extension in life could greatly reduce the costs associated with replacement of these expensive parts. Presently, after a certain number of cycles or predetermined amount of use a number of critical components are retired even though there may be no obvious damage. Under a Retirement For Cause (RFC) philosophy, each component would be periodically inspected and returned to service until an indication of damage is observed. There is, however, an inherent risk as more parts are used longer that a flawed or cracked part could pass inspection and fail in service. Therefore, in order to implement an RFC philosophy, without increasing the risk of an in-service failure, a number of technologies are viewed as critical. The technical areas include nondestructive evaluation, life prediction, and engine health monitoring. Each area will briefly be discussed.
4:00 PM *NN2.5
NDE, PROBABILISTICS, AND RISK IN AGING ENGINEERED SYSTEMS. Charles Annis , P.E. Pratt & Whitney, West Palm Beach, FL.
Quantitative NDE - NonDestructive Evaluations with measured efficacy - now permit a priori risk evaluation of differing maintenance schemes. But meaningful prognostication is often hampered by inappropriate use of statistics in anticipating the risks. Computations sometimes use assumptions based more on expediency than reality. For example, there are unspoken (or sometimes, unappreciated, or, worse, unknown) assumptions that the events being modeled are statistically independent, when they are not.
The resulting computations can dangerously underestimate the structural risk. An example of a triple engine failure on an Eastern Airlines L-1011 illustrates some of these ideas.
4:30 PM *NN2.6
LIFE PREDICTION FOR GAS TURBINE ENGINES. R.H. Van Stone , GE Aircraft Engines, Cincinnati, OH.
Life management systems for aircraft engine components are becoming increasing reliant on damage tolerance techniques. The crack propagation behavior of advanced nickel base superalloys used in disk applications can be influenced by the interactions between high stresses, temperatures, and extended application times under these conditions. Fracture mechanics models have been developed to accurately predict the combined influence of hold times, overpeak retardation, and complex mission cycling. The basis of these models will be discussed. Examples of laboratory data validating these life prediction methods will be shown. The difficulties associated with predicting complex missions and the combinations of different mission types will also be discussed.
8:30 AM *NN3.1
SESSION NN3: AGING AIRCRAFT III
Chairs: Ulf Goranson and D. Gary Harlow
Tuesday Morning, December 1, 1998
Clarendon A/B (S)
TAILORING AN NDI SYSTEM FOR APPLICABILITY TO A BROAD CLASS OF AEROSPACE VEHICLES. Charles E. Harris and Edward R. Generazio, National Aeronautics and Space Administration, Langley Research Center, Hampton, VA.
There are several very versatile inspection methods that may be used for a wide variety of inspection needs. These methods include ultrasonic, thermal, electromagnetic, x-ray, and optical. Furthermore, most of these methods can be used to detect fundamentally different types of damage such as fatigue cracks, corrosion, and disbonds. The best system for any given inspection need will provide the required resolution capability with the highest reliability or probability of detection. However, each type of material and structural configuration requires specialization of the inspection method. Thin-sheet aluminum fuselage structure are fundamentally different to inspect than are thick-section aluminum wing structure. In addition, the differences between materials such as titanium, aluminum, steel, and continuous fiber-reinforced polymer matrix composites also require considerable specialization of any inspection method. Specialization includes both changes to the hardware as well as changes to the signal processing algorithms. Those systems that give the best results are typically developed and optimized for a specific inspection requirement. Since the adoption of the damage tolerance philosophy in the 1970's, dramatic improvements in detection capability have been achieved through the development of better sensors and probes, signal processing algorithms, and system optimization processes. Recently, there has been considerable effort to develop rapid, broad-area scanning techniques using concepts such as sensor arrays and robotics. Also, high fidelity systems are being developed that will reliability detect damage in hidden structure such as corrosion between overlapping components in built-up structure. The authors will present numerous examples of advanced inspection techniques being used to insure the safety of aerospace vehicles including aircraft and the NASA space shuttle orbitor, fuel tanks, and main engine. The authors will also describe recent developments in in-situ health monitoring systems.
9:00 AM *NN3.2
PROBABILISTIC REPRESENTATION OF A POLYCRYSTALLINE MICROSTRUCTURE WITH APPLICATION TO INTERGRANULAR CRACK GROWTH. S.R. Arwade , M. Grigoriu, A.R. Ingraffea, P.A. Wawrzynek, E. Iesulauro, School of Civil and Environmental Engineering; P.R. Dawson, M.P. Miller, Sibley School of Mechanical and Aerospace Engineering; C.R. Myers, Cornell Theory Center; Cornell University, Ithaca, NY.
The finite element method can be applied to model microstructural features in a material and analyze the evolution of cracks which may have length of the order of a few grain sizes. The analysis is capable of treating complex geometries and many degrees-of-freedom. Such analysis involves (1) mechanistically based models for the behavior of short fatigue cracks, and (2) algorithms to create digital models of the material microstructure. These algorithms are based upon probabilistic models which capture the inherent variability of microstructural features.
Probabilistic models have been developed for the grain geometry of recrystallized metals, and the associated crystal lattice orientation. The Voronoi tessellation, with nuclei defined by a Poisson point process, was selected as the model for the grain geometry. The tessellation has been shown to produce grain geometries which agree well with measured geometries in terms of grain size and topological class distribution. There remain challenges regarding replication of the grain shapes found in certain materials types, such as rolled plates. A random field model was selected for representing the crystal lattice orientation at each point in a material sample. The model was calibrated to experimental measurements of the lattice orientation of an aluminum sample made using the electron backscatter diffraction technique, and is found to match the orientation distribution function of the experimental sample. We expect that the general methodology of generating digital samples of the material microstructure from probabilistic models should be applicable to other microstructural features such as precipitate particles, or dislocation structures.
An example is used to demonstrate the generation of microstructures consistent with a sample of aluminum and the evolution of cracks along grain boundaries. A simplified mechanical model for grain boundary (GB) toughness is given which correlates the GB fracture toughness to the GB misorientation. The finite element analysis code FRANC2D is used for calculations.
9:30 AM NN3.3
A PHYSICALLY-BASED THEORY OF DAMAGE ACCUMULATION IN ELASTO-PLASTIC MATERIALS INCORPORATING THE EFFECT OF MICROSTRUCTURE. Tony Honein, H-Cube Technology, Fremont, CA; Elie Honein , George Herrmann, Stanford Univ, Div of Mechanics and Computation, Stanford, CA.
This paper presents a physically-based model of damage in elasto-plastic materials. The theory is based on the so-called Conventional Thermodynamics of Irreversible Processes, where the concept of a local thermodynamic state plays a dominant role. A material body prone to damage is regarded as a thermodynamic system characterized by a set of extensive variables that can be defined in both equilibrium and nonequilibrium states and assigned approximately the same values in both the physical space and the abstract state space (i.e., the Gibbsian phase space of constrained equilibria.) The extensive variables introduced include internal parameters that describe the damaged state of the body and which are selected according to the form in which they appear in the thermodynamic potentials. The local state approximation is applied by assigning to the entropy and temperature in physical space local values which can be calculated in the Gibbsian phase space by the traditional methods of equilibrium thermodynamics. This leads to an explicit expression for the entropy production. The rate equations for the damage are then postulated in such a way as to conform to the second part of the second law of thermodynamics. The resulting theory captures many features of real material behavior such as loading/unloading paths, dependence on the straining or loading rates, transition from brittleness to ductility with temperature rise and dependence on some global geometric parameters of the structure. Additionally, the effect of microstructure and the associated toughening mechanisms are incorporated into the theory by considering a power law for the toughness curve. The theoretical results are compared with experimental results under various conditions and an excellent qualitative as well as quantitative agreement is achieved.
10:15 AM NN3.4
MONITORING ANELASTIC RESPONSE DURING ALUMINUM ALLOY MICROSTRUCTURAL EVOLUTION USING LASER ULTRASONICS. J.B. Spicer and S.L. Wallace, Dept of Materials Science and Engineering, The Johns Hopkins University, Baltimore, MD.
Laser ultrasonic methods for materials characterization have been applied to a variety of materials systems ranging from polymer-based composites to single crystal sapphire. The widely recognized characteristics of these methods, most importantly remote and couplant-free generation and detection of ultrasound, are often exploited for materials characterization where the materials system under consideration is in an aggressive environment or is particularly affected by contact with ultrasonic coupling media. Unfortunately, most applications use the laser source as a simple substitute for a conventional ultrasonic source. The unique characteristics of laser source, which transmits both longitudinal and shear waves, are rarely exploited. In this work, the laser ultrasonic monitoring of aluminum alloy microstructural development during thermal processing is demonstrated. This monitoring is based on the ability to measure the differential ultrasonic attenuation in alloy systems. Differential attenuation is related to anelastic physical processes which are active during microstructural evolution. For the aluminum alloys studied in this work, these anelastic effects derive from alloying species solution and mobility; however, other effects that produce differential attenuation might also be monitored successfully.