An Overview Of Prognosis Health Management Research At GRC .
Annual Conference of the Prognostics and Health Management Society, 2009An Overview of Prognosis Health Management Research at GRCfor Gas Turbine Engine Structures With Special Emphasis onDeformation and Damage ModelingSteven M. Arnold1, Robert K. Goldberg1, Bradley A. Lerch1,and Atef F. Saleeb21NASA Glenn Research Center, Cleveland, Ohio, 44135, sa.govBradley.A.Lerch@nasa.gov2University of Akron, Akron, Ohio, 44325, CTHerein a general, multimechanism, physicsbased viscoelastoplastic model is presented inthe context of an integrated diagnosis andprognosis methodology which is proposed forstructural health monitoring, with particularapplicability to gas turbine engine structures.In this methodology, diagnostics andprognostics will be linked through stateawareness variable(s). Key technologies whichcomprise the proposed integrated approachinclude 1) diagnostic/detection methodology,2) prognosis/lifing methodology, 3) diagnostic/prognosis linkage, 4) experimentalvalidation and 5) material data informationmanagement system. A specific prognosislifing methodology, experimental characterization and validation and data informationmanagement are the focal point of currentactivities being pursued within this integratedapproach. The prognostic lifing methodologyis based on an advanced multi-mechanismviscoelastoplastic model which accounts forboth stiffness and/or strength reductiondamage variables. Methods to characterizeboth the reversible and irreversible portions ofthe model are discussed. Once the multiscalemodel is validated the intent is to link it toappropriate diagnostic methods to provide a1structuralhealthmonitoringINTRODUCTIONThe desire for higher performance (e.g., increasedthrust to weight ratio) and efficiencies in gas turbineengines, along with the availability of advancedtechnologies, has resulted in designers pushing thedesign envelop to the limit in such areas as, forexample, increasing the pressure ratio, decreasing tipclearance, and increasing operational temperatures.These advanced designs have in turn caused significantincreases in overall operational and maintainabilitycosts, due to associated uncertainties in component lifeand reliability. Similarly, as technology advances,component costs typically increase proportionately –thus enhancing the desire for better clarification ofuseful (safe) remaining life, to decrease direct operatingcosts. This need explains the heightened desire for andincreased research activity in the area of conditionmonitoring with an eye toward diagnosis and prognosisof various critical components. To date, in commercialgas turbine engines continuous health monitoring islimited to generic “global” system measurements likeshaft vibrations and EGT (exit gas temperatures) andlocalized measurements are only taken at windows of*Steven M. Arnold et al. This is an open-access articledistributed under the terms of the Creative CommonsAttribution 3.0 United States License, which permitsunrestricted use, distribution, and reproduction in anymedium, provided the original author and source are credited.1
Annual Conference of the Prognostics and Health Management Society, 2009opportunity rather than on a continuous basis. Theselocal measurements (which involve expensive teardown) are still limited; in general, to visual/opticalsurface inspection techniques (measurements), whichthemselves fall short in their ability to detect criticaldefects. Conversely, in military engines whichexperience far more intense and often unexpectedloadings, condition-based monitoring has recently beenundertaken as the standard; wherein generic engine data(environment, temperature, pressure, vibration) arecollected and monitored, and integrated and correlatedwith past mission histories. However, prognosticmethodologies, as defined herein, still remain elusive,as clear linkage between damage (critical defects, lifelimiting events) and response signature(s), be theyrefined or crude, have yet to be established.Further structural life is known to be extremelysensitive to preexistent damage such as manufacturingflaws and/or service induced damage that can causeimmediate fracture or serve as sources for early agement approaches rely on a combination ofpredictive models and periodic inspections both on-lineand off-line. The approaches primarily in use today are(Grandt, 2004): 1) Safe-Life Design - requires that thecomponent be retired before the initiation of cracks andis susceptible to the presence of unanticipated structuralor material damage that greatly reduce the crackinitiation portion of the fatigue process; 2) DamageTolerant Design - assumes a structure contains initialcracks (typically assumed equal to the largestundetected defect size), and defines the ability of thestructure to resist fracture from cracks of a given sizefor a specified time period; and 3) Retirement for Cause- utilizes periodic inspection intervals to locatedamaged components that are then either repaired orreplaced. The inspection intervals are based on the timefor an undetected crack to grow to fracture. All threeapproaches demonstrate the interdependence ofdiagnostic (inspection) and prognosis (life prediction)methods and stress the importance of developing thefundamental scientific causal relationships of failure toprovide the key diagnostic/prognostic linkage fordifferent material systems and structures. Largeimprovements in safety, life extension, and overall lifecycle costs can be attained by employing a newphilosophy to propulsion health management.However, in many of the approaches currently used(Adams, 2004, Springer, 2004), the prognosis of futureevents is based on an extrapolation of events whichhave occurred previously. If the future loading is variedsignificantly from the previously applied loadingprofiles, the classical prognosis methods will notaccurately predict the future response, which explainswhy the accuracy of these approaches is limited to onlyshort time extrapolations. Similarly, many successeshave been documented (Fatemi and Yang, 1998,Grandt, 2004) in the realm of prognosis (analyticalmodeling predictions); given the damage initiation siteand subsequent known loading conditions. However,for truly accurate predictions of future events (that varysignificantly from prior events), a physics-basedprognostic model is required in which, given an initialload/damage condition, an arbitrary set of futureloading profiles can be applied (in an off-line approach)to determine the future damage and ultimate life of thestructure. Linkages to diagnostics methods through theuse of state-awareness variables are still required,however, in order to provide information on the currentload/damage state at a given point in time, so that aphysics based prognostics model can independentlydetermine the future response of the structure.Consequently, the overall objective of the presentstructural health management program at NASA GlennResearch Center (GRC) (which is supported by theNASA IVHM (Integrated Vehicle Health Management)Project within the Aviation Safety Program, is todevelop, implement and experimentally verify a lifing(prognosis) methodology for health monitoring ofstructural components operating at high temperatures,which is tightly coupled (through corresponding stateawareness variables) to detection (diagnosis)techniques. In this project, NASA Glenn researchersare focusing on the development/characterization/ andexperimental verification of a fully associative,physics-based, viscoelastoplastic - damage modelwhich can be utilized as a prognostic tool within a hotstructural life management system, primarily aimed atthe propulsion system environment; wherein, inducedlocalized softening mechanisms (which are stronglyinfluenced by geometry, loading conditions, inherentdefect distributions and material anisotropy) areconsidered. As implied above, in this work, the termprognostics refers to the ability to predict remaining lifegiven the current state of the material. A key feature ofthe viscoelastoplastic model which constitutes the basisof our prognostics system is the incorporation ofadvanced features of the response of metallic alloys atelevated temperatures. As engine structures willencounter elevated temperatures, classical analysismethods and constitutive equations will not be able toaccurately simulate the material response. The keyfeature of our methodology is that the time-dependentaspect of the material response (i.e., viscoelastic andviscoplastic) dominant at elevated temperatures, areaccounted for within the model along with theappropriate local and global failure criteria. Therefore,the physics-based prognosis methodology proposedwill be established for high temperature structuralcomponents under general loading conditions(multiaxial loading with and without overloads, cycliceffects, thermomechanical, etc.) and experimentally2
Annual Conference of the Prognostics and Health Management Society, 20092FULLY INTEGRATEDDIAGNOSTIC/PROGNOSTIC LIFEMANAGEMENT SYSTEM (FILMS)The life of a component is dependent not only upon itspast history (i.e., the loads imposed on it, its initial andcurrent physical condition) but also upon the ntly, if one desires to maximize the life of agiven component one must always know the currentstate of the component (state awareness, obtained viadetection techniques), and how future events will affectits life (prognosis or life prediction), given its presentcondition. Given the above considerations, a two-prongapproach is adopted wherein developed FEAppropriate ScaleResponse SignatureMeasurementsFuture EventsTimectionIncurred History- Past EventsPrognosis/Predivalidated with the design of a prognosticallychallenging test matrix. This test matrix will consist ofbi-axial experimental demonstration problems atambient and elevated temperatures that will be suitablefor use in characterizing, evaluating and rankingprognostic (and detection) methods. While research isbeing conducted on both metallic and compositestructures, and the overall methodology describedherein is applicable to a variety of material systems, thespecific results presented in this paper are focused ontitanium engine components made of Ti-6-4. Thegeneral framework, once validated for the Ti-6-4system, will be applied in the future using otherprototypical alloys used in gas turbine engines, such aspowder metallurgy nickel-based disk superalloys.The paper begins with an overview of a proposedfully-integrated diagnostic/prognosis life managementsystem, with special emphasis on how the developedprognosis (constitutive) models can fit within tive model which is the basis of the prognosisframework is then described in detail. This is thenfollowed by the efforts to date on the characterizationand validation of the specific prognostic component ofthe proposed integrated framework. Specifically, thecharacterization and validation of the reversible timedependent (viscoelastic) behavior of a representativetitanium alloy (i.e., Ti-6-4) is discussed as well as thecurrent progress achieved in understanding theirreversible (viscoplastic) response. While admittedlythere is significant amounts of work that have yet to becarried out so as to enable one to fully predict thedeformation and damage response of a metallicstructure, even the preliminary results presented hereinwill demonstrate the capability of the proposedprognosis model to analyze key features of the responseof metallic alloys that classical analysis methods cannotcapture (such as the time- and rate-dependence of thelow-strain reversible response).State Awarenessis Connecting LinkMulti-mechanismDeformationandLife ModelsRequires Knowledge Of1.Material Behavior2.Physics of Failure3.Specific Geometry4.Anticipated Loads/Constraints5.State (material orstructure )Figure 1: Depicts the distinction between Diagnosticand Prognostic Methodologies and their integration.methods will be fully linked with complementarydetection tools, illustrated in Figure 1. As discussedearlier, it is believed that only when these two viewsare consistently integrated (the connecting link beingthe current state of the material/structure) can a rationaland viable health management system be established.Consequently, for the proposed prognostic model to beviable, a detection scheme (of the direct type) must beestablished with sufficient resolution to 1) detect thepresence of defect(s), 2) locate the defect(s) and 3) sizethe extent of damage throughout the history of acomponent. Given this information, establishing astructurally meaningful connection (based on thephysics of the failure or defect) with a physics-basedcoupled deformation and damage model is essential topredicting reliably the available remaining life givenmultiple future event scenarios.Figure 2 illustrates how the diagnostic andprognostic tools need to be linked to form a fullstructural health monitoring system. The diagnosticstools (utilized in the “past history” portion of the chart)are required to provide the damage state at a currentpoint in time. The prognostics tools are then utilized inthe “future” portion of the chart to predict the futuredamage and life of the structure. As mentioned earlier,without the use of advanced prognostics tools onlyextrapolations of future events based on previouslyoccurring trends (the “future past” line in the chart)can be predicted. However, with the use of a fullphysics-based prognostic tool, the damage and liferesulting from any potential future loading profile canbe examined. Note that the “critique window”identified in the figure represents the period of timewhere the damage progression is examined in thestructure in order to either extrapolate the futurematerial response (in order to generate the“Future Past” type curve) or to provide the initialconditions for a more physics-based type of prognosismethodology.3
Annual Conference of the Prognostics and Health Management Society, 2009DcriticalDAre PastDiagnostic Viewpoint –Enablesknowledge ofcurrent state.FutuDamage (e.g. crack size)State awareness sensorsMicrostructurally-based stochasticbehavior – enables confidence levels.Modified Future MissionCritiqueWindowDA-1Prognosis Viewpoint – Enableslife managementby modifyingfuture loading.FutureATime (cycles)Past HistoryFigure 2: Depiction of an integrated diagnostic andprognosis approach to incrementally updating a lifemanagement system.Consistent with this vision is the concept of scalespecific resolution – that is the understanding that eachsuccessive level of fidelity on the prognosis sidedemands a commensurate level on the diagnostic side.Consequently, many researchers have attempted toassign discriminators to allow the use of rank methods.For example, consider the four levels of damageidentification (LODI) put forth by Rytter (Rytter,1993): Level 1: Determination that damage is present inthe structureLevel 2: Level 1 plus determination of thegeometric location of the damageLevel 3: Level 2 plus quantification of the severityof the damageLevel 4: Level 3 plus prediction of the remainingservice-life of the structureNote many of the popular global diagnostictechniques provide no more than LODI 1 fidelity,when the ultimate desire is to reach level 4.Furthermore, understanding that failure is a process andnot an event empowers a detection/prognosisphilosophy, which can be exploited to manage theevent that is developing, not just react to it. In thiscontext, failure being a process means that the ultimatefailure of a structure is due to the accumulation ofdamage to a significant enough degree that thestructural integrity is compromised. If the initiation andprogression of the lower levels of damage are notaccurately detected and predicted, determination of theultimate structural life is questionable at best.Furthermore, the higher up in the failure processhierarchy (i.e. the closer to ultimate failure) that a faultis detected, the less manageability remains and the lesstime exists before functionality is compromised beyondthe usable state. Ideally damage should be detected wellbefore imminent failure is present, and the prediction ofremaining life should be able to take place well withinFigure 3: Temporal interrogation of a given component.the zone of safe operation, and well before the partshould be removed from service. This has been referredto as the predictive horizon (Hess, 2002).This concept is especially significant whendeveloping a viable health monitoring system forcomplex “real-world” applications. In constructingsuch a system one must fully recognize the cornucopiaof available sensor techniques and their practicallimitations (e.g., inability to operate in elevatedtemperature environments, limited accessibility, fearthat failure initiates at sensor sites, etc). Consequently,any on-line monitoring device is likely to be quitecoarse in nature (i.e., scalar, e.g., pressure or EGT,) andin its spatial distribution (e.g., at very few specificlocations in the engine). Alternatively, off-linediagnostic techniques are often rich in their spatialcontent, but ideally limited in their temporaldistribution, that is, taken at intermittent windows ofopportunity rather than on a continuous basis, seeFigure 3. Thus the ultimate measure of success for anycomprehensive health monitoring endeavor (forinstance like FILMS, proposed herein) will be theability to discriminate between these two distinctdiagnostic scales (that is, a very spatially-crude rally-rare). Furthermore, the overarching goalmust be to provide diagnostic data in sufficient quantityand quality such that the prognostic tool can makereasonable predictions.With the above ideas in mind there are fiveoverarching technology areas that must be addressedconcurrently to establish such a robust fully-integrated,multi-scale, multi-mechanism diagnostic and prognosislife management system (FILMS). These are: 1)diagnostic/detection methodology, 2) prognosis/lifingmethodology, 3) diagnostic/prognostic linkage, 4)experimental validation and 5) reasoning methodologyand material data information management. In thispaper, the areas of prognosis/lifing methodology andexperimental validation will be concentrated on asthese components will be developed in the greatestdetail and most likely provide the most unique4
Annual Conference of the Prognostics and Health Management Society, 2009contributions to the development of an integrateddiagnostic/prognostic system. The material datainformation management topic will also be covered insome detail as the proper capturing, analysis,dissemination and maintaining of material data can playa significant role in facilitating the characterization andutilization of the sophisticated material model which isthe heart of the proposed prognostics approach. Theareas related to diagnostics and diagnostics/prognosticslinkage will be described briefly, to give insight as tohow the current prognostic model can be integratedwithin the context of a full health monitoring system.Furthermore, in the area of detection, multipleapproaches (some complementary/some competitive)will be considered as resources allow as part of theresearch program, with an eye toward down selectionand the development of a hierarchy system withvarying scale specific resolution2.1 Detection MethodologyDetection techniques may be classified as global orlocal. Global methods attempt to assess simultaneouslythe condition of the whole structure (e.g., vibrationmeasurements using an accelerometer), whereas localmethods provide information about a relatively smallregion of the system by using localized measurements(e.g., strain gages
applicability to gas turbine engine structures. In this methodology, diagnostics and prognostics will be linked through state awareness variable(s). Key technologies which comprise the proposed integrated approach include 1) diagnostic/detection methodology, 2) prognosis/lifing methodology, 3) diag nostic/prognosis linkage, 4) experimental
Figure 2: End of Life Planning: Details of Care Provision 7 Prognosis of less than 1 year 8 Prognosis of less than 6 months 11 Prognosis of a 'few weeks' 12 Prognosis of days or hours 13 Care after death 16 Additional services to meet individual needs 16 Specialist Palliative Care 16 Audit 16 Care Quality Commission and End of Life Care 17 .
those in classical ILC with relatively abundant, eosino-philic cytoplasm. Classical ILC has a more favorable prognosis than the pleomorphic form [4]. In invasive ductal carcinoma, the prognosis differs widely according to molecular subtype. Unfortunately, there is no evidence of any difference in prognosis between ILC and IDC with similar .
The aim of our study was to determine the influence of apoptotic variations on the carcinogenesis and prognosis of colorectal carcinomas . Higher AIs are associated with more aggressive tumours and a poorer prognosis for patients with CRC. Keywords: Apoptosis, Colorectal carcinoma, Prognosis . As well 26 sam-ples belonging to patients with .
A of Medicare and be certified as being terminally ill. An individual is considered to be terminally ill if the medical prognosis is that the individual’s life expectancy is six months or less if the illness runs its normal course. * CMS IOM Publication 100-02, Medicare Benefit Policy Manual, Chapter 9, Section 10, “Requirements- General” 9
IQVIA MARKET PROGNOSIS 2017-2021 . continue to battle chronic overcrowding. Effective implementation of healthcare reforms will be hampered by funding issues, staff and equipment shortages, and poor management. . branded products by hospitals may be reduced from 2018. Changes could also be introduced for
doctor and patient. If the Claimant has died then the expert has to comment on what the prognosis would have been. Pre-Action it helps the instructing solicitor to determine whether there is a potential claim and also to give an indication of the likely va
Key words: Renal cell carcinoma, Early relapse, Prognosis, Biomarker, mRNA signatur e Introduction Renal cell cancer (RCC) is one of the worldwide common carcinomas, with approximately 403,262 new cases and 175,098 deaths expected in 2018[1]. The overall prognosis of RCC cou
Click 500 series – Customizable devices built on our Click 500 platform This user guide covers the Click 500 series. For the Click 100–400 series, please see the Click 100–400 Series User Guide. Using this Manual This manual is divided into two parts: Part I: Introduction to the Click Series – This part contains information common to the Click line, beginning with basic module .
