Sriram Pattabhiraman

Skipping unnecessary structural airframe maintenance using an on-board structural health monitoring system

Structural airframe maintenance is a subset of scheduled maintenance, and is performed at regular intervals to detect and repair cracks that would otherwise affect the safety of the airplane. It has been observed that only a fraction of airplanes undergo structural airframe maintenance at earlier scheduled maintenance times. But, intrusive inspection of all panels on the airplanes needs to be performed at the time of scheduled maintenance to ascertain the presence/absence of large cracks critical to the safety of the airplane. Recently, structural health monitoring techniques have been developed. They use on-board sensors and actuators to assess the current damage status of the airplane, and can be used as a tool to skip the structural airframe maintenance whenever deemed unnecessary. Two maintenance philosophies, scheduled structural health monitoring and condition-based maintenance skip, have been developed in this article to skip unnecessary structural airframe maintenances using the on-board structural health monitoring system. A cost model is developed to quantify the savings of these maintenance philosophies over scheduled maintenance.

Effects of Uncertainty Reduction Measures by Structural Health Monitoring on Safety and Lifecycle Costs of Aircrafts

In aircrafts, fuselage inspections are performed regularly to remove large damages that threaten the safety of the structure. Traditionally preventive inspections have been scheduled and performed leading to high costs over the lifecycle of an airplane. Recently, structural health monitoring techniques have been developed that uses sensors and actuators to detect damages on structures paving way for progressive inspection. In this paper, the cost effectiveness of progressive inspection over scheduled inspection is analyzed. The lifecycle of an airplane was modeled as blocks of damage propagation interspersed with inspection. The Paris model with random parameters is used to model damage growth, and detection probability during inspections is modeled by Palmberg expression. Monte Carlo Simulations delineate the process. SHM based progressive inspection are found to be 50% more cost effective than schedule-based preventive inspections. The sensitivity of the lifecycle cost to the inspection parameters has been studied. To accommodate critical panels which must be manually inspected, a hybrid model of inspection is also proposed. The hybrid model is found to have sufficient cost savings over scheduled inspection model. Nomenclature a = Half damage size p = Pressure differential t = Thickness of the fuselage r = Fuselage radius m = Paris Law exponent C = Paris Law constant ah-man = Palmberg parameter for manual inspection w man

Bayesian Approach for Fatigue Life Prediction From Field Data

Due to uncertainty in design, manufacturing and operating processes, the initial prediction of a machine‟s useful life is often quite different from that of the actual machine. In this paper, we utilize the Bayesian technique to incorporate the field data with the initial predictions in order to improve the prediction. The field data is interpreted in terms of the probability of having defective hardware, and then the likelihood function is generated from the binomial distribution. Since the predictions incorporate field experience, as time progresses and more data becomes available the probabilistic predictions are continuously updated. This results in a continuous increase of confidence and accuracy of the prediction. The resulting distributions can then be used directly in risk analysis, maintenance scheduling, and financial forecasting by both manufacturers and operators of heavy-duty gas turbines. This presents a quantification of the real time risk for direct comparison with the volatility of the power market.

Effect of Inspection Strategies on the Weight and Lifecycle Cost of Airplanes

Inspections of aircrafts fuselage panels are performed periodically, at scheduled intervals, to repair damage that can threaten the safety of the structure. Recently, structural health monitoring techniques have been developed that use sensors and actuators to detect damage, paving way for condition-based maintenance. This paper quantifies the effect of inspection strategies on the safety and lifetime cost of an airplane. The lifecycle of an airplane was modeled as blocks of damage propagation interspersed with inspection. The Paris model with uncertain parameters is used to model fatigue damage growth, and the detection process by inspections is modeled by the Palmberg equation. Two inspection strategies are compared based on their effect of optimal weight and lifecycle cost of the panel, while maintaining a desired level of reliability. Direct integration procedure computes the reliability for a given set of maintenance parameters. It is found that the better inspection model leads to about 15% savings in weight of airplane and about 35% savings in lifecycle cost of an airplane over scheduled maintenance.

Modeling Average Maintenance Behavior of Fleet of Airplanes using Fleet-MCS

In aircrafts, fuselage inspections are performed regularly to remove large damages that threaten the safety of the structure. Recently, structural health monitoring techniques have been developed that uses sensors and actuators to detect damages on structures paving way for progressive inspection. The average maintenance hangar trips per airplane and the average number of panels replaced on it have a direct bearing on the cost of progressive inspection. The lifecycle of an airplane was modeled as blocks of damage propagation interspersed with inspection. The Paris model with random parameters is used to model damage growth, and detection probability during inspections is modeled by Palmberg expression. Conventionally. Monte Carlo Simulations delineate the process. In this paper, a fleet-MCS procedure is presented that predict the average behavior of a fleet of airplanes using simple analytical expressions. Fleet-MCS procedures reduce the high computational cost of Monte Carlo simulations in predicting the average fleet behavior while maintaining similar level of accuracy. Monte Carlo simulations involve random sampling and would require multiple simulations to predict the fleet average. Fleet-MCS procedure predicts the fleet average with a single run of the simulation reducing the computational burden. The fleet average from the regular MCS and the fleetMCS has been compared in this paper and has been found to be in accordance with reasonable accuracy.

Synchronizing Condition-based Maintenance with Necessary Scheduled Maintenance

Structural airframe maintenance is a subset of scheduled maintenance, and is performed at regular intervals to detect and repair cracks that would otherwise affect the safety of the airplane. It has been observed that only a fraction of airplanes undergo structural airframe maintenance at earlier scheduled maintenance times. But, intrusive inspection of all panels on the airplanes need to be performed at the time of scheduled maintenance to ascertain the presence/absence of large cracks critical to the safety of the airplane. Recently, structural health monitoring (SHM) techniques have been developed. They use onboard sensors and actuators to assess the current damage status of the airplane, and can be used as a tool to skip the structural airframe maintenance, whenever deemed un-necessary. Two maintenance philosophies, Sched-SHM and CBM-skip, have been developed in this paper, to skip unnecessary structural airframe maintenances using on-board SHM system. A cost model is developed to quantify the savings of these maintenance philosophies over scheduled maintenance.

Bayesian Approach for Fatigue Life Prediction from Field Inspection

In the design considering fatigue life of mechanical components, uncertainties arising from the materials and manufacturing processes should be taken into account for ensuring reliability. Common practice in the design is to apply safety factor in conjunction with the numerical codes for evaluating the lifetime. This approach, however, most likely relies on the designers experience. Besides, the predictions often are not in agreement with the real observations during the actual use. In this paper, a more dependable approach based on the Bayesian technique is proposed, which incorporates the field failure data with the prior knowledge to obtain the posterior distribution of the unknown parameters of the fatigue life. A matter of prior knowledge is also considered since the posterior distribution is influenced by it. Posterior predictive distributions and associated values are estimated afterwards, which represents the degree of our belief of the life conditional on the observed data. As more data are provided, the values will be updated to more confident information. The results can be used in various needs such as a risk analysis, reliability based design optimization, maintenance scheduling, or validation of reliability analysis codes. In order to obtain the posterior distribution, Markov Chain Monte Carlo (MCMC) technique is employed, which is a modern statistical computational method which draws effectively the samples of the given distribution. Field data of turbine components are exploited to illustrate our approach, which counts as a regular inspection of the number of failed blades in a turbine disk.

Bayesian Technique for Reducing Uncertainty in Fatigue Failure Model

In this paper, Bayesian statistics is utilized to update uncertainty associated with the fatigue life relation. The distribution for fatigue strain at a constant load cycle is determined using the initial uncertainty from analytical prediction and likelihood functions associated with test data. The Bayesian technique is a good method to reduce uncertainty and at the same time provides a conservative estimate, given the distribution of analytical prediction errors and variability of test data. First, the distribution of analytical fatigue model error is estimated using Monte Carlo simulation with uniformly distributed parameters. Then the error distribution is progressively updated by using the test variability as a likelihood function, which is obtained from field test data. The sensitivity of estimated distribution with respect to the initial error distribution and the selected likelihood function is studied. The proposed method is applied to estimate the fatigue life of turbine blade. It is found that the proposed Bayesian technique reduces the scatteredness of fatigue life by almost 50%, while maintaining the conservative life estimate at a given fatigue strain. In addition, a good conservative estimate of fatigue life prediction has been proposed using a knockdown factor that is obtained from the distribution of lowest test data.

Conservative Fatigue Life Estimation using Bayesian Update

In this paper, Bayesian update is utilized to reduce uncertainty associated with the fatigue life relation. The distribution for fatigue strain at a constant life cycle is determined using the initial uncertainty from analytical prediction and likelihood functions from test data. The Bayesian technique is a good method to reduce uncertainty and at the same time, provides a conservative estimate, given the distribution of analytical prediction errors and variability of test data. First, the distribution of fatigue model error is estimated using Monte Carlo simulation with uniformly distributed parameters. Then the error distribution is progressively updated by using the test variability as a likelihood function, which is obtained from field test data. The sensitivity of estimated distribution with respect to the initial error distribution and the selected likelihood function is studied. The proposed method is applied to estimate the fatigue life of turbine blade. It is found that the proposed Bayesian technique reduces the scatteredness in life by almost 50%, while maintaining the conservative life estimate at a given fatigue strain. In addition, a good conservative estimate of fatigue life prediction has been proposed using a knockdown factor that is obtained from the distribution of lowest test data. Moreover, Bayesian update has been utilized to update the parameters of strain – life curve for a case of constant strain amplitude