Results for “FAILURES” — Page 13

DOE OSTI2025

Failure Analysis–Informed Risk Assessment Framework for Geological Carbon Storage Using Numerical Simulation and Machine Learning

Geological carbon storage (GCS) is recognized as a critical technology for achieving large-scale reductions in anthropogenic carbon dioxide (CO 2 ) emissions. Ensuring long-term containment and safety requires robust risk assessment frameworks that account for geological uncertainty and identify potential failure scenarios. Among various indicators, the area of review (AoR) serves as a key metric for evaluating storage performance, regulatory compliance, and monitoring design, as it delineates the spatial extent impacted by pressure buildup and plume migration. However, conventional AoR-based risk assessments typically perturb parameters within narrow uncertainty bounds, potentially overlooking rare but high-impact events arising from extreme geological conditions. In this study, we present a failure analysis–informed risk assessment framework for large-scale GCS projects to improve site prescreening and monitoring design. A suite of 300 numerical simulations was generated using stochastic geological models that vary five key parameters: net-to-gross ratio, anisotropy azimuth, porosity multiplier, permeability multiplier, and vertical-to-horizontal permeability ratio. Among these, 200 realizations represent normal geological uncertainty, while 100 additional cases explore extreme yet plausible conditions for failure-case analysis. The AoR was simulated and computed from pressure and CO 2 saturation fields, where the baseline AoR boundary, representing the extent predicted under typical geological uncertainty, was defined as the union of 200 normal-range simulations, and failure was identified when extreme-range cases exceeded this baseline. Results show that incorporating broader parameter uncertainty produces significantly larger AoR extents, underscoring the potential underestimation of risk under conventional uncertainty ranges. Furthermore, spatial probability maps derived from failure-induced AoR exceedance identify regions requiring enhanced monitoring attention. Various machine learning (ML)–based classifiers were developed to predict failure occurrence from geological parameters, with the random forest model achieving the highest performance (F1-score of 0.986). Consistent findings from correlation coefficient, feature importance, and Sobol sensitivity analyses reveal that low net-to-gross ratios and permeability multipliers are the dominant risk drivers, reflecting reduced reservoir connectivity and limited pressure dissipation. Altogether, these results provide a novel framework for risk-informed site prescreening and monitoring design that explicitly considers rare but high-impact geological scenarios in GCS projects.

25 ENERGY STORAGE↗

DOE OSTI2025

Strain to failure measurements of Cr coatings with different processing

Differently processed Cr coatings on Zircaloy cladding were assessed for coating strain to failure. This report includes i) quantification of Cr coating strain to failure for several coating variants, ii) incorporation of high throughput analysis developed utilizing artificial intelligence, and iii) develop a hoop direction coating strain to failure testing approach for high temperatures that doesn’t require digital image correlation strain measurements. Room temperature strain to failure measurements indicated that heat treatment had an impact on strain to failure of the coating, with increasing heat treatment temperature increasing the average strain to failure of the Cr coating.

36 MATERIALS SCIENCE↗

DOE OSTI2025

Failure Strain and Related Triaxiality of Aluminum 6061-T6, A36 Carbon Steel, 304 Stainless Steel, and Nitronic 60 Metals, Part I: Experimental Investigation

The objective of this study is to develop failure-limit material models for Aluminum 6061-T6, A36 Carbon Steel, 304 Stainless Steel, and Nitronic 60 metals, based on parameters of plastic equivalent strain (failure strain) and stress triaxiality. The research is conducted in two parts. This paper presents Part One of the study. In Part One, custom-designed test specimens undergo controlled uniaxial tension and compression testing at ambient temperature. These tests are performed at quasi-static speeds using Universal Testing Machines (UTMs) in accordance with ASTM E8 and ASTM E9 standards. Experimental data, specifically engineering stress–strain and force–displacement curves, are recorded from the onset of loading until specimen fracture, or in the case of compression tests, until the capacity of the testing machine is reached. In Part Two, the emphasis shifts to the calibration of Finite Element Analysis (FEA) models of the custom-designed test specimens. Plastic equivalent strain and the corresponding stress triaxiality values at failure are extracted from each test specimen for the given metal. These values are then systematically plotted onto a single graph to construct the failure-limit curve, which delineates the boundary conditions for material failure. This approach will facilitate the development of a comprehensive material property definition that correlates plastic equivalent strain with stress triaxiality at failure for Aluminum 6061-T6, A36 Carbon Steel, 304 Stainless Steel, and Nitronic 60 metals.

Harwell, Ron↗

NASA NTRS1976

Failure detection and isolation investigation for strapdown skew redundant tetrad laser gyro inertial sensor arrays

The degree to which flight-critical failures in a strapdown laser gyro tetrad sensor assembly can be isolated in short-haul aircraft after a failure occurrence has been detected by the skewed sensor failure-detection voting logic is investigated along with the degree to which a failure in the tetrad computer can be detected and isolated at the computer level, assuming a dual-redundant computer configuration. The tetrad system was mechanized with two two-axis inertial navigation channels (INCs), each containing two gyro/accelerometer axes, computer, control circuitry, and input/output circuitry. Gyro/accelerometer data is crossfed between the two INCs to enable each computer to independently perform the navigation task. Computer calculations are synchronized between the computers so that calculated quantities are identical and may be compared. Fail-safe performance (identification of the first failure) is accomplished with a probability approaching 100 percent of the time, while fail-operational performance (identification and isolation of the first failure) is achieved 93 to 96 percent of the time.

Eberlein, A. J.↗

NASA NTRS1980

Ferrographic and spectrographic analysis of oil sampled before and after failure of a jet engine

An experimental gas turbine engine was destroyed as a result of the combustion of its titanium components. Several engine oil samples (before and after the failure) were analyzed with a Ferrograph as well as plasma, atomic absorption, and emission spectrometers. The analyses indicated that a lubrication system failure was not a causative factor in the engine failure. Neither an abnormal wear mechanism, nor a high level of wear debris was detected in the oil sample from the engine just prior to the test in which the failure occurred. However, low concentrations of titanium were evident in this sample and samples taken earlier. After the failure, higher titanium concentrations were detected in oil samples taken from different engine locations. Ferrographic analysis indicated that most of the titanium was contained in spherical metallic debris after the failure.

Jones, W. R., Jr.↗

NASA NTRS1980

Failure analyses of composite bolted joints

The complex failure behavior exhibited by bolted joints of graphite epoxy (Hercules AS/3501) was investigated for the net tension, bearing and shearout failure modes using combined analytical and experimental techniques. Plane stress, linear elastic, finite element methods were employed to determine the two dimensional state of stress resulting from a loaded hole in a finite width, semiinfinite strip. The stresses predicted by the finite element method were verified by experiment to lend credence to the analysis. The influence of joint geometric parameters on the state of stress and resultant strength of the joint was also studied. The resulting functional relationships found to exist between bolted joint strength and the geometric parameters, were applied in the formulation of semiempirical strength models for the basic failure modes. A point stress failure criterion was successfully applied as the failure criterion for the net tension and shearout failure modes.

Wilson, D. W.↗

NASA NTRS1981

Failure of morphology of (0 deg)8 graphite/epoxy as influenced by environments and processing

Optical and scanning electron microscopy were used to investigate the failure morphology of graphite/epoxy specimens which had been tested until tensile failure. Failure morphology was studied as a function of the quality control variables of specimen preparation technique, prepreg batch, and cure condition, and also as a function of the environmental parameters of temperature and moisture content. Defective specimens were found to exhibit a low energy failure morphology. Poor specimen edge preparation and one batch of prepreg when tested at elevated temperature or moisture content also exhibited energy failure morphology. Postcuring had no effect on strength but did slightly alter failure morphology. Temperature or moisture appeared to decrease flaw sensitivity and thus increase strength; however, moisture also appeared to increase interfacial debonding between filament and matrix. When combined moisture and temperature increased interfacial debonding and made the epoxy matrix more prone to fracture.

Clements, L. L.↗

NASA NTRS1981

Failure analysis of a tool steel torque shaft

A low design load drive shaft used to deliver power from an experimental exhaust heat recovery system to the crankshaft of an experimental diesel truck engine failed during highway testing. An independent testing laboratory analyzed the failure by routine metallography and attributed the failure to fatigue induced by a banded microstructure. Visual examination by NASA of the failed shaft plus the knowledge of the torsional load that it carried pointed to a 100 percent ductile failure with no evidence of fatigue. Scanning electron microscopy confirmed this. Torsional test specimens were produced from pieces of the failed shaft and torsional overload testing produced identical failures to that which had occurred in the truck engine. This pointed to a failure caused by a high overload and although the microstructure was defective it was not the cause of the failure.

Reagan, J. R.↗

NASA NTRS1983

Cycle life test and failure model of nickel-hydrogen cells

Six ampere hour individual pressure vessel nickel hydrogen cells were charge/discharge cycled to failure. Failure as used here is defined to occur when the end of discharge voltage degraded to 0.9 volts. They were cycled under a low earth orbit cycle regime to a deep depth of discharge (80 percent of rated ampere hour capacity). Both cell designs were fabricated by the same manufacturer and represent current state of the art. A failure model was advanced which suggests both cell designs have inadequate volume tolerance characteristics. The limited existing data base at a deep depth of discharge (DOD) was expanded. Two cells of each design were cycled. One COMSAT cell failed at cycle 1712 and the other failed at cycle 1875. For the Air Force/Hughes cells, one cell failed at cycle 2250 and the other failed at cycle 2638. All cells, of both designs, failed due to low end of discharge voltage (0.9 volts). No cell failed due to electrical shorts. After cell failure, three different reconditioning tests (deep discharge, physical reorientation, and open circuit voltage stand) were conducted on all cells of each design. A fourth reconditioning test (electrolyte addition) was conducted on one cell of each design. In addition post cycle cell teardown and failure analysis were performed on the one cell of each design which did not have electrolyte added after failure.

Smithrick, J. J.↗

NASA NTRS1984

Failure analysis of plasma-sprayed thermal barrier coatings

Thermally induced failure processes of plasma-sprayed thermal barrier coatings are examined. Cracking processes give rise to noise which was monitored by acoustic emission (AE) techniques. The sequential failure of coatings was examined from samples which were thermally cycled. Coatings of yttria-stabilized zirconia with and without a NiCrAlZr bond coat were plasma-sprayed onto U700 alloy rod. In some cases the substrate was intentionally overheated during deposition of the thermal protection system to check how this process variable influenced the AE response of the specimen. In this way a qualitative appraisal of how process variables affect coating integrity could be discerned in terms of cracking behavior. Results from up to seven consecutive thermal cycles are reported here. Coating failure was observed in all cases. Failure of the thermal protection system is progressive, since cracking and crack growth were observed prior to ultimate failure. Thus castastrophic failure occurs at some stage when there is a transformation from the microcrack to a macrocrack network.

Berndt, C. C.↗

NASA NTRS1983

Failure detection with uncertain models

In this paper the author considers the issue of robust failure detection. In one way or another all failure detection methods generate signals which tend to highlight the presence of particular failures if they have actually occurred. However, if any model uncertainties have effects on the observables which are at all like those of one or more of the failure modes, these will also be accentuated. Consequently the problem of robust failure detection is concerned with generating signals which are maximally sensitive to some effects (failures) and minimally sensitive to other (model errors). Previously announced in STAR as N83-32417

Lou, X.-C.↗

NASA NTRS1983

Cycle life test and failure model of nickel-hydrogen cells

Six ampere hour individual pressure vessel nickel hydrogen cells were charge/discharge cycled to failure. Failure as used here is defined to occur when the end of discharge voltage degraded to 0.9 volts. They were cycled under a low earth orbit cycle regime to a deep depth of discharge (80 percent of rated ampere hour capacity). Both cell designs were fabricated by the same manufacturer and represent current state of the art. A failure model was advanced which suggests both cell designs have inadequate volume tolerance characteristics. The limited existing data base at a deep depth of discharge (DOD) was expanded. Two cells of each design were cycled. One COMSAT cell failed at cycle 1712 and the other failed at cycle 1875. For the Air Force/Hughes cells, one cell failed at cycle 2250 and the other failed at cycle 2638. All cells, of both designs, failed due to low end of discharge voltage (0.9 volts). No cell failed due to electrical shorts. After cell failure, three different reconditioning tests (deep discharge, physical reorientation, and open circuit voltage stand) were conducted on all cells of each design. A fourth reconditioning test (electrolyte addition) was conducted on one cell of each design. In addition post cycle cell teardown and failure analysis were performed on the one cell of each design which did not have electrolyte added after failure. Previously announced in STAR as N83-25038

Smithrick, J. J.↗

NASA NTRS1984

Compression failure characteristics of + or - 45-deg-dominated laminates with a circular hole or impact damage

An investigation of the compression failure characteristics of + or - 45-deg-dominated laminates with a circular hole or impact damage was conducted. Graphite-epoxy laminates consisting of all + or - 45-deg plies and of + or - 45-deg and 90-deg plies were evaluated. Inplane shearing between the fibers and matrix (i.e., matrix shearing) was found to be the primary compression failure mechanism for an all + or - 45-deg laminate with a hole. The failure mechanism for the + or 45-deg-dominated laminates with a hole appeared to be a combination of delamination and matrix shearing. The compression failure stresses and strains for both the all + or - 45-deg laminate and the + or - 45-deg-dominated laminates were unaffected by impact damage at an energy level of 5.5 J when compared to results for undamaged control specimens. The failure stresses and strains for all laminates with impact damage at an energy level of 17.0 J or of 34.3 J were significantly lower than the corresponding data for the undamaged control specimens. The failure stresses and strains for these impact-damaged specimens were independent of the percentage of 90-deg plies in the laminate.

Shuart, M. J.↗

NASA NTRS1984

Effects of Constituent Properties on Compression Failure Mechanisms

Compression failure mechanisms were investigated through the analysis of matrix-embedded fiber bundles. The use of fiber bundle specimens can provide much needed information on compression failure mechanisms because failure of the bundle is well contained and can be monitored during testing. The method can clearly distinguish between buckling-induced failure and shear-induced failure. The results indicate that WY and T300 graphite fibers and E-glass fiber fail in buckling while the high-modulus P75S graphite fiber fails in shear. Buckling-induced failure becomes more evident with low-modulus fiber in softer epoxy.

Hahn, H. T.↗

NASA NTRS1985

The application of the detection filter to aircraft control surface and actuator failure detection and isolation

The performance of the detection filter in detecting and isolating aircraft control surface and actuator failures is evaluated. The basic detection filter theory assumption of no direct input-output coupling is violated in this application due to the use of acceleration measurements for detecting and isolating failures. With this coupling, residuals produced by control surface failures may only be constrained to a known plane rather than to a single direction. A detection filter design with such planar failure signatures is presented, with the design issues briefly addressed. In addition, a modification to constrain the residual to a single known direction even with direct input-output coupling is also presented. Both the detection filter and the modification are tested using a nonlinear aircraft simulation. While no thresholds were selected, both filters demonstrated an ability to detect control surface and actuator failures. Failure isolation may be a problem if there are several control surfaces which produce similar effects on the aircraft. In addition, the detection filter was sensitive to wind turbulence and modeling errors.

Bonnice, W. F.↗

NASA NTRS1986

A geometric approach to failure detection and identification in linear systems

Using concepts of (C,A)-invariant and unobservability (complementary observability) subspaces, a geometric formulation of the failure detection and identification filter problem is stated. Using these geometric concepts, it is shown that it is possible to design a causal linear time-invariant processor that can be used to detect and uniquely identify a component failure in a linear time-invariant system, assuming: (1) The components can fail simultaneously, and (2) The components can fail only one at a time. In addition, a geometric formulation of Beard's failure detection filter problem is stated. This new formulation completely clarifies of output separability and mutual detectability introduced by Beard and also exploits the dual relationship between a restricted version of the failure detection and identification problem and the control decoupling problem. Moreover, the frequency domain interpretation of the results is used to relate the concepts of failure sensitive observers with the generalized parity relations introduced by Chow. This interpretation unifies the various failure detection and identification concepts and design procedures.

Massoumnia, M. A.↗

NASA NTRS1984

Failure analysis of plasma-sprayed thermal barrier coatings

Thermally induced failure processes of plasma-sprayed thermal barrier coatings are examined. Cracking processes give rise to noise which was monitored by acoustic emission (AE) techniques. The sequential failure of coatings was examined from samples which were thermally cycled. Coatings of yttria-stabilized zirconia with and without a NiCrAlZr bond coat were plasma-sprayed onto U700 alloy rod. In some cases the substrate was intentionally overheated during deposition of the thermal protection system to check how this process variable influenced the AE response of the specimen. In this way a qualitative appraisal of how process variables affect coating integrity could be discerned in terms of cracking behavior. Results from up to seven consecutive thermal cycles are reported here. Coating failure was observed in all cases. Failure of the thermal protection system is progressive, since cracking and crack growth were observed prior to ultimate failure. Thus castastrophic failure occurs at some stage when there is a transformation from the microcrack to a macrocrack network.

Berndt, C. C.↗

NASA NTRS1987

A vector-based failure detection and isolation algorithm for a dual fail-operational redundant strapdown inertial measurement unit

A vector-based failure detection and isolation technique for a skewed array of two degree-of-freedom inertial sensors is developed. Failure detection is based on comparison of parity equations with a threshold, and isolation is based on comparison of logic variables which are keyed to pass/fail results of the parity test. A multi-level approach to failure detection is used to ensure adequate coverage for the flight control, display, and navigation avionics functions. Sensor error models are introduced to expose the susceptibility of the parity equations to sensor errors and physical separation effects. The algorithm is evaluated in a simulation of a commercial transport operating in a range of light to severe turbulence environments. A bias-jump failure level of 0.2 deg/hr was detected and isolated properly in the light and moderate turbulence environments, but not detected in the extreme turbulence environment. An accelerometer bias-jump failure level of 1.5 milli-g was detected over all turbulence environments. For both types of inertial sensor, hard-over, and null type failures were detected in all environments without incident. The algorithm functioned without false alarm or isolation over all turbulence environments for the runs tested.

Morrell, Frederick R.↗