A. Doom, Elena Hollingsworth, Riley Bishop, Wesley Fisher, Brian Mazzoni, Chidurala Manohar
This paper discusses the design, validation, and verification of a wind tunnel data acquisition system that measures the pressure distribution around a cylinder and airfoil and calculates the resulting lift and drag forces acting on the object in cross flow. Typical aerodynamic research is conducted using wind tunnels with large test sections and high air speeds, which can be expensive and unattainable for many universities. A traceable path of validation can improve the results of experimentation on small test section and low speed wind tunnels by providing important information on the accuracy, uncertainty, and error involved in the system. It is concluded that CFD verification and validation not only improves experimentation but is the key to conducting quality research with limited resources.
{"title":"Verification and Validation of a Small Wind Tunnel Data Acquisition System","authors":"A. Doom, Elena Hollingsworth, Riley Bishop, Wesley Fisher, Brian Mazzoni, Chidurala Manohar","doi":"10.1115/imece2021-71806","DOIUrl":"https://doi.org/10.1115/imece2021-71806","url":null,"abstract":"\u0000 This paper discusses the design, validation, and verification of a wind tunnel data acquisition system that measures the pressure distribution around a cylinder and airfoil and calculates the resulting lift and drag forces acting on the object in cross flow. Typical aerodynamic research is conducted using wind tunnels with large test sections and high air speeds, which can be expensive and unattainable for many universities. A traceable path of validation can improve the results of experimentation on small test section and low speed wind tunnels by providing important information on the accuracy, uncertainty, and error involved in the system. It is concluded that CFD verification and validation not only improves experimentation but is the key to conducting quality research with limited resources.","PeriodicalId":146533,"journal":{"name":"Volume 13: Safety Engineering, Risk, and Reliability Analysis; Research Posters","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129971317","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Probabilistic failure risk analysis is important in recent development of airworthiness. In the fracture mechanics module of probabilistic failure risk analysis, efficient and accurate stress intensity factor (SIF) solutions are vital. Universal weight function (UWF) is a method that has a remarkable computational efficiency and high accuracy in SIF calculation. However, the concrete coefficients in the UWF remain unknown, and the rules of the geometric parameters and these coefficients have not been clarified. This limitation hinders the subsequent application of the method. This article discusses general off-center embedded cracks. The response surface method (RSM), as a means of database establishment, is used to construct the relations between the coefficients in the UWF and geometric parameters. The mean absolute percentage errors of the SIFs between UWF and finite element method are less than 4% within a certain range. Gaussian process regression is adopted to better fit the data, especially when the crack is near the plate boundary and the R-square is over 0.96. In addition, the effect of the offset distance on SIFs is discussed for a certain embedded crack in a given plate. Results show that changes in the SIFs are dependent on the plate boundary in the uniform stress field. However, stress predominates the changes in the SIFs when the crack is not very close to the plate boundary in the nonuniform distributed stress fields. The databases and discussion that followed can provide some reference and inspiration for further research.
{"title":"Establishment of the Off-Center Embedded Crack Stress Intensity Factor Database for Probabilistic Risk Assessment Based on Universal Weight Function","authors":"Tongge Xu, S. Ding, Guo Li","doi":"10.1115/imece2021-70198","DOIUrl":"https://doi.org/10.1115/imece2021-70198","url":null,"abstract":"\u0000 Probabilistic failure risk analysis is important in recent development of airworthiness. In the fracture mechanics module of probabilistic failure risk analysis, efficient and accurate stress intensity factor (SIF) solutions are vital. Universal weight function (UWF) is a method that has a remarkable computational efficiency and high accuracy in SIF calculation. However, the concrete coefficients in the UWF remain unknown, and the rules of the geometric parameters and these coefficients have not been clarified. This limitation hinders the subsequent application of the method. This article discusses general off-center embedded cracks. The response surface method (RSM), as a means of database establishment, is used to construct the relations between the coefficients in the UWF and geometric parameters. The mean absolute percentage errors of the SIFs between UWF and finite element method are less than 4% within a certain range. Gaussian process regression is adopted to better fit the data, especially when the crack is near the plate boundary and the R-square is over 0.96. In addition, the effect of the offset distance on SIFs is discussed for a certain embedded crack in a given plate. Results show that changes in the SIFs are dependent on the plate boundary in the uniform stress field. However, stress predominates the changes in the SIFs when the crack is not very close to the plate boundary in the nonuniform distributed stress fields. The databases and discussion that followed can provide some reference and inspiration for further research.","PeriodicalId":146533,"journal":{"name":"Volume 13: Safety Engineering, Risk, and Reliability Analysis; Research Posters","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130215157","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Recently we introduced a new model-based fault injection method implemented as a highly customizable Simulink block called FIBlock. It supports the injection of typical faults of essential heterogeneous components of Cyber-Physical Systems (CPS), such as sensors, computing hardware, and network. The FIBlock allows to tune a fault type and configure multiple parameters to tune error magnitude, fault activation time, and fault exposure duration. The FIBlock is able to generate various types of highly adjustable CPS faults. We demonstrated the performance of the FIBlock on a Simulink case study representing a lower-limb EXOLEGS exoskeleton, an assistive device for the elderly in everyday life. In particular, we discovered the spatial and temporal thresholds for different fault types. Upon exceeding said thresholds, the Dynamic Movement Primitives-based control system could no longer adequately compensate errors. In this paper, we proposed a new Deep Learning-based approach for system failure prevention. We employed the Long Short-Term Memory (LSTM) network for error detection and mitigation. Error detection is achieved using the prediction approach. The LSTM models are mitigating the detected errors with computed predictions only when they were subject to the imminent failure (i.e., exceeded the aforementioned thresholds). To compare our approach with previous findings, we trained two LSTM models on angular position and angular velocity signals. For evaluation, we performed fault injection experiments with varying fault effect parameters. The ‘Sensor freeze’ fault was injected into the angular position sensor, and the ‘Stuck-at 0’ fault was injected into angular velocity sensor. The presented Deep Learning-based approach prevented system failure even when the injected faults were substantially exceeding thresholds. In addition, reasoning for data access point choice has been evaluated. We compared two options: (i) the input data for LSTM is provided from the sensor output and (ii) from the controller output. In the paper, the pros and cons for both options are presented. We deployed the trained LSTM models on an Edge Tensor Processing Unit. For that, the models have been quantized, i.e. all the 32-bit floating-point numbers (such as weights and activation outputs) were converted to the nearest 8-bit fixed-point numbers and converted to the TensorFlow Lite models. The Coral USB Accelerator was coupled with a Raspberry Pi 4B for signal processing. The result proves the feasibility of the proposed method. Because the LSTM models were converted to the 8bit integer TensorFlow Lite models, it allowed firm real-time error mitigation. Furthermore, the light weight of the system and minimal power consumption allows its integration into wearable robotic systems.
{"title":"Deep Learning-Based Error Mitigation for Assistive Exoskeleton With Computational-Resource-Limited Platform and Edge Tensor Processing Unit","authors":"T. Fabarisov, A. Morozov, I. Mamaev, K. Janschek","doi":"10.1115/imece2021-70387","DOIUrl":"https://doi.org/10.1115/imece2021-70387","url":null,"abstract":"\u0000 Recently we introduced a new model-based fault injection method implemented as a highly customizable Simulink block called FIBlock. It supports the injection of typical faults of essential heterogeneous components of Cyber-Physical Systems (CPS), such as sensors, computing hardware, and network. The FIBlock allows to tune a fault type and configure multiple parameters to tune error magnitude, fault activation time, and fault exposure duration.\u0000 The FIBlock is able to generate various types of highly adjustable CPS faults. We demonstrated the performance of the FIBlock on a Simulink case study representing a lower-limb EXOLEGS exoskeleton, an assistive device for the elderly in everyday life. In particular, we discovered the spatial and temporal thresholds for different fault types. Upon exceeding said thresholds, the Dynamic Movement Primitives-based control system could no longer adequately compensate errors.\u0000 In this paper, we proposed a new Deep Learning-based approach for system failure prevention. We employed the Long Short-Term Memory (LSTM) network for error detection and mitigation. Error detection is achieved using the prediction approach. The LSTM models are mitigating the detected errors with computed predictions only when they were subject to the imminent failure (i.e., exceeded the aforementioned thresholds). To compare our approach with previous findings, we trained two LSTM models on angular position and angular velocity signals. For evaluation, we performed fault injection experiments with varying fault effect parameters. The ‘Sensor freeze’ fault was injected into the angular position sensor, and the ‘Stuck-at 0’ fault was injected into angular velocity sensor. The presented Deep Learning-based approach prevented system failure even when the injected faults were substantially exceeding thresholds. In addition, reasoning for data access point choice has been evaluated. We compared two options: (i) the input data for LSTM is provided from the sensor output and (ii) from the controller output. In the paper, the pros and cons for both options are presented.\u0000 We deployed the trained LSTM models on an Edge Tensor Processing Unit. For that, the models have been quantized, i.e. all the 32-bit floating-point numbers (such as weights and activation outputs) were converted to the nearest 8-bit fixed-point numbers and converted to the TensorFlow Lite models. The Coral USB Accelerator was coupled with a Raspberry Pi 4B for signal processing. The result proves the feasibility of the proposed method. Because the LSTM models were converted to the 8bit integer TensorFlow Lite models, it allowed firm real-time error mitigation. Furthermore, the light weight of the system and minimal power consumption allows its integration into wearable robotic systems.","PeriodicalId":146533,"journal":{"name":"Volume 13: Safety Engineering, Risk, and Reliability Analysis; Research Posters","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128729865","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Edwar Romero-Ramirez, Charisma Clarke, Sanna F. Siddiqui, Gerardo Carbajal
In this paper, the effect of various additive manufacturing processing parameters on thermoplastic polyurethane’s mechanical properties and the durometer hardness testing were evaluated. Additive manufacturing of thermoplastics is a rapidly growing field in many areas, from hobbyists to manufacturing industries. Parts are built layer by layer following the slicing software settings that convert a 3D object into a group of 2D regions that resembles CNC code. There is a need to understand how the mechanical properties are affected by the process parameters compared to traditional injection molding manufacturing. This thermoplastic has use in many automotive industry applications and consumer products because of its elasticity and high abrasion resistance. A generic thermoplastic polyurethane, also known as flexible TPU, with Shore A durometer hardness of 95 was used for this task. It was found that the number of solid layers impacts the mechanical properties higher than other printing parameters. Maximum tensile strength of up to 51.4±6.0 MPa for 84.1±3.8 Shore A durometer hardness was measured.
{"title":"Mechanical Properties and Durometer Testing Relationship of Thermoplastic Polyurethane","authors":"Edwar Romero-Ramirez, Charisma Clarke, Sanna F. Siddiqui, Gerardo Carbajal","doi":"10.1115/imece2021-69648","DOIUrl":"https://doi.org/10.1115/imece2021-69648","url":null,"abstract":"\u0000 In this paper, the effect of various additive manufacturing processing parameters on thermoplastic polyurethane’s mechanical properties and the durometer hardness testing were evaluated. Additive manufacturing of thermoplastics is a rapidly growing field in many areas, from hobbyists to manufacturing industries. Parts are built layer by layer following the slicing software settings that convert a 3D object into a group of 2D regions that resembles CNC code. There is a need to understand how the mechanical properties are affected by the process parameters compared to traditional injection molding manufacturing. This thermoplastic has use in many automotive industry applications and consumer products because of its elasticity and high abrasion resistance. A generic thermoplastic polyurethane, also known as flexible TPU, with Shore A durometer hardness of 95 was used for this task. It was found that the number of solid layers impacts the mechanical properties higher than other printing parameters. Maximum tensile strength of up to 51.4±6.0 MPa for 84.1±3.8 Shore A durometer hardness was measured.","PeriodicalId":146533,"journal":{"name":"Volume 13: Safety Engineering, Risk, and Reliability Analysis; Research Posters","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126002909","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Undercarriage impact occurs when vehicle’s ground clearance is incompatible with obstacle on the road. This kind of accidents are particularly dangerous to electric vehicles as battery pack is usually integrated into the vehicle floor. In case of an undercarriage collision, the battery pack could be ploughed through by the obstacle on the road, which could cause damage to battery cells in the pack and occurrence of internal short circuit and thermal runaway. To tackle this new problem, damage assessment method is needed to guide design of protective structure of battery pack. The present paper documents an analysis of battery damages in undercarriage collisions and development of a test method for evaluating undercarriage collisions. We first used a simplified finite element model of battery pack and conducted computer simulations to understand the influences of key parameters. The battery pack model included blade battery cells that had no module-level assemblies. Based on the learning from the finite element analysis, we then developed a generic test method for evaluating undercarriage collisions. The test method is robust in terms of repeatability and the device is adaptable for evaluating a wide range of electric vehicle undercarriage collision. We hope that this test method can become a unified and standard method for evaluating battery pack damage in undercarriage collisions and guide design of protection structures for enhancing safety performance of battery packs.
{"title":"Damage Assessment Method of Battery Pack of Electric Vehicle in Undercarriage Collision","authors":"Powen Chen, Yong Xia, Qing Zhou, Yunlong Qu, Xinqi Wei","doi":"10.1115/imece2021-69776","DOIUrl":"https://doi.org/10.1115/imece2021-69776","url":null,"abstract":"\u0000 Undercarriage impact occurs when vehicle’s ground clearance is incompatible with obstacle on the road. This kind of accidents are particularly dangerous to electric vehicles as battery pack is usually integrated into the vehicle floor. In case of an undercarriage collision, the battery pack could be ploughed through by the obstacle on the road, which could cause damage to battery cells in the pack and occurrence of internal short circuit and thermal runaway. To tackle this new problem, damage assessment method is needed to guide design of protective structure of battery pack. The present paper documents an analysis of battery damages in undercarriage collisions and development of a test method for evaluating undercarriage collisions. We first used a simplified finite element model of battery pack and conducted computer simulations to understand the influences of key parameters. The battery pack model included blade battery cells that had no module-level assemblies. Based on the learning from the finite element analysis, we then developed a generic test method for evaluating undercarriage collisions. The test method is robust in terms of repeatability and the device is adaptable for evaluating a wide range of electric vehicle undercarriage collision. We hope that this test method can become a unified and standard method for evaluating battery pack damage in undercarriage collisions and guide design of protection structures for enhancing safety performance of battery packs.","PeriodicalId":146533,"journal":{"name":"Volume 13: Safety Engineering, Risk, and Reliability Analysis; Research Posters","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131268770","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Anil U. Peerapur, Mangesh N. Dhavalikar, S. Dingare, B. Patle
Pneumatics presents an alternative to electric and hydraulic actuators used in manufacturing industries. Pneumatic components provide a safe and reliable opportunity to automate a production line. However, pneumatic systems are notoriously expensive to operate as a result of system losses and inefficiencies. Typical systems have an output efficiency of 10%–12%. The Problems are — there is a fluctuating system pressure, which can cause air tools and other air-operated equipments to function less efficiently, affecting production; excess compressor capacity, resulting in higher than necessary costs due to no tracking of air consumption; decreased service life and increased maintenance of supply equipment due to unnecessary cycling and increased run time. This offers a significant improvement opportunity to meet targets concerning energy consumption. Leakages and excessive pressures are commonly identified as major sources of waste. Leaks or chokes in a compressed air system can be in various types due to diversified reasons. The proposed simulation method aims at the detection of faults based on signatures obtained. It has significance in terms of potential energy savings, decrease in breakdowns and increase in the service life of components. The Transfer function of a system is obtained by mathematical modelling of a system. The fault conditions are simulated in MATLAB SimulinkR to get results.
{"title":"A Proposed Method for Online Condition Monitoring of Pneumatic Systems Under Different Operating Conditions and Parameters for Optimal Energy Consumption","authors":"Anil U. Peerapur, Mangesh N. Dhavalikar, S. Dingare, B. Patle","doi":"10.1115/imece2021-69942","DOIUrl":"https://doi.org/10.1115/imece2021-69942","url":null,"abstract":"\u0000 Pneumatics presents an alternative to electric and hydraulic actuators used in manufacturing industries. Pneumatic components provide a safe and reliable opportunity to automate a production line. However, pneumatic systems are notoriously expensive to operate as a result of system losses and inefficiencies. Typical systems have an output efficiency of 10%–12%. The Problems are — there is a fluctuating system pressure, which can cause air tools and other air-operated equipments to function less efficiently, affecting production; excess compressor capacity, resulting in higher than necessary costs due to no tracking of air consumption; decreased service life and increased maintenance of supply equipment due to unnecessary cycling and increased run time. This offers a significant improvement opportunity to meet targets concerning energy consumption. Leakages and excessive pressures are commonly identified as major sources of waste. Leaks or chokes in a compressed air system can be in various types due to diversified reasons. The proposed simulation method aims at the detection of faults based on signatures obtained. It has significance in terms of potential energy savings, decrease in breakdowns and increase in the service life of components. The Transfer function of a system is obtained by mathematical modelling of a system. The fault conditions are simulated in MATLAB SimulinkR to get results.","PeriodicalId":146533,"journal":{"name":"Volume 13: Safety Engineering, Risk, and Reliability Analysis; Research Posters","volume":"49 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131711946","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this paper a summary of the differences between structured query language (SQL) based traditional relational database management systems (RDBMS) and recently popular NoSQL based database are presented. The importance of selecting a NoSQL database for the implementation of digital Twin (DT) framework for nuclear reactor predictive maintenance has been discussed. Example of commercially available MongoDB based NoSQL database implementation with storing data from various sources: such as from virtual time-series temperature measurements (based on finite element-based heat transfer models) at thousands of nodal locations and from real sensor measurements (from continuous online monitoring based active ultrasonic sensors and from non-continuous nondestructive testing (NDT) based eddy current measurements is demonstrated.
{"title":"A Review of SQL vs NoSQL Database for Nuclear Reactor Digital Twin Applications: With Example MongoDB Based NoSQL Database for Digital Twin Model of a Pressurized-Water-Reactor Steam-Generator","authors":"S. Mohanty, T. Elmer, S. Bakhtiari, R. Vilim","doi":"10.1115/imece2021-73153","DOIUrl":"https://doi.org/10.1115/imece2021-73153","url":null,"abstract":"\u0000 In this paper a summary of the differences between structured query language (SQL) based traditional relational database management systems (RDBMS) and recently popular NoSQL based database are presented. The importance of selecting a NoSQL database for the implementation of digital Twin (DT) framework for nuclear reactor predictive maintenance has been discussed. Example of commercially available MongoDB based NoSQL database implementation with storing data from various sources: such as from virtual time-series temperature measurements (based on finite element-based heat transfer models) at thousands of nodal locations and from real sensor measurements (from continuous online monitoring based active ultrasonic sensors and from non-continuous nondestructive testing (NDT) based eddy current measurements is demonstrated.","PeriodicalId":146533,"journal":{"name":"Volume 13: Safety Engineering, Risk, and Reliability Analysis; Research Posters","volume":"51 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132194128","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
A component will not reliable unless it is designed with the required reliability. Since the Advisory Group on Reliability of Electronic Equipment (AGREE) published the “Reliability of Military Electronic Equipment” in 1957, engineering reliability has gradually become an engineering branch to serve engineering design. But implementation process of reliability in mechanical components design is slow. There are maybe two main reasons: (1) There is a lack of statistical descriptions of material mechanical properties; and (2) There is a lack of appropriate reliability design approaches for mechanical component reliability design, especially for a fatigue design under cyclic loads. A thin-wall vessel is a typical mechanical component under combined stresses. It is extremely hard to establish the limit state function for a thin-wall vessel structure under cyclic loads by using the P-S-N curve approach. For the fatigue design issue, the author proposes to use the probabilistic component fatigue strength index and fatigue damage index model, which has been proposed by the author, for establishing the limit state function for the thin-wall vessel structure under cyclic loads, and then determining its reliability under cyclic loads. This paper presents and explains how to establish the limit state functions of a thin-wall vessel structure under different typical static loads and cyclic loads. Two case study examples are provided for implementing the proposed approaches to conduct the reliability calculation of a thin-wall vessel structure under static load and a cyclic load.
{"title":"Implementation of Reliability Design Theory on a Thin-Wall Vessel Structure","authors":"Xiaobin Le","doi":"10.1115/imece2021-67665","DOIUrl":"https://doi.org/10.1115/imece2021-67665","url":null,"abstract":"\u0000 A component will not reliable unless it is designed with the required reliability. Since the Advisory Group on Reliability of Electronic Equipment (AGREE) published the “Reliability of Military Electronic Equipment” in 1957, engineering reliability has gradually become an engineering branch to serve engineering design. But implementation process of reliability in mechanical components design is slow. There are maybe two main reasons: (1) There is a lack of statistical descriptions of material mechanical properties; and (2) There is a lack of appropriate reliability design approaches for mechanical component reliability design, especially for a fatigue design under cyclic loads. A thin-wall vessel is a typical mechanical component under combined stresses. It is extremely hard to establish the limit state function for a thin-wall vessel structure under cyclic loads by using the P-S-N curve approach. For the fatigue design issue, the author proposes to use the probabilistic component fatigue strength index and fatigue damage index model, which has been proposed by the author, for establishing the limit state function for the thin-wall vessel structure under cyclic loads, and then determining its reliability under cyclic loads. This paper presents and explains how to establish the limit state functions of a thin-wall vessel structure under different typical static loads and cyclic loads. Two case study examples are provided for implementing the proposed approaches to conduct the reliability calculation of a thin-wall vessel structure under static load and a cyclic load.","PeriodicalId":146533,"journal":{"name":"Volume 13: Safety Engineering, Risk, and Reliability Analysis; Research Posters","volume":"163 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133659544","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Safety and reliability are two critical factors of modern Cyber-Physical Systems (CPS). However, the increasing structural and behavioral complexity of modern automation systems significantly increases the possibility of system errors and failures, which can easily lead to economic loss or even hazardous events. Anomaly Detection (AD) techniques provide a potential solution to this problem, and conventional methods, e.g., Autoregressive Integrated Moving Average model (ARIMA), are no longer the best choice for anomaly detection for modern complex CPS. Recently, Deep Learning (DL) and Machine Learning (ML) anomaly detection methods became more popular, and numerous practical applications have been presented in many industrial scenarios. Most of the modern DL-based anomaly detection methods use the prediction approach and LSTM architecture. The Transformer is a new neural network architecture that outperforms LSTM in natural language processing. In this paper, we show that the Transformer-based deep learning model, which has received much attention recently, can be applied to the anomaly detection of industrial automation systems. Specifically, we collect time-series data from a system of two industrial robots using a Simulink model. Then, we feed these data into our Transformer-based model and train it to be a time-series data predictor. The paper presents the experimental results that show the comparison of precision and speed of a Long-Short Time Memory (LSTM) predictor and our Transformer-based predictor.
{"title":"Anomaly Detection for Cyber-Physical Systems Using Transformers","authors":"Yuliang Ma, A. Morozov, Sheng Ding","doi":"10.1115/imece2021-69395","DOIUrl":"https://doi.org/10.1115/imece2021-69395","url":null,"abstract":"\u0000 Safety and reliability are two critical factors of modern Cyber-Physical Systems (CPS). However, the increasing structural and behavioral complexity of modern automation systems significantly increases the possibility of system errors and failures, which can easily lead to economic loss or even hazardous events. Anomaly Detection (AD) techniques provide a potential solution to this problem, and conventional methods, e.g., Autoregressive Integrated Moving Average model (ARIMA), are no longer the best choice for anomaly detection for modern complex CPS. Recently, Deep Learning (DL) and Machine Learning (ML) anomaly detection methods became more popular, and numerous practical applications have been presented in many industrial scenarios. Most of the modern DL-based anomaly detection methods use the prediction approach and LSTM architecture. The Transformer is a new neural network architecture that outperforms LSTM in natural language processing.\u0000 In this paper, we show that the Transformer-based deep learning model, which has received much attention recently, can be applied to the anomaly detection of industrial automation systems. Specifically, we collect time-series data from a system of two industrial robots using a Simulink model. Then, we feed these data into our Transformer-based model and train it to be a time-series data predictor. The paper presents the experimental results that show the comparison of precision and speed of a Long-Short Time Memory (LSTM) predictor and our Transformer-based predictor.","PeriodicalId":146533,"journal":{"name":"Volume 13: Safety Engineering, Risk, and Reliability Analysis; Research Posters","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121172805","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Due to the complexity and challenging nature of predicting the effects of impact on a vehicle, much of the crash-testing industry remains one which operates at full-scale. Automobiles and even aircraft are often crashed at full size, which is an expensive and time-consuming process. This paper details a novel approach to crash-testing that combines early use of simulation, scaling of the appropriate parameters using similitude, and limited testing. The method was used to predict the crashworthiness of a small, commercial unmanned aerial vehicle (UAV). The authors were involved in an undergraduate design project to develop a system capable of safely decelerating the RQ-7 Shadow Unmanned Aerial System (UAS) from an operational flight without the use of a runway. For contingency reasons or tactical purposes, the use of a larger runway for the UAS was not feasible. The team developed an airbag-like system that used controlled release of air to decelerate the vehicle. However, due to testing constraints the team was unable to test the airbag system at full-scale with an actual RQ-7 vehicle. The system is already far in it’s life-cycle so there were no prototypes available and the team’s available facilities would not facilitate a large, high-speed test. As a result, the team built a model of the existing vehicle, focusing on key components and materials. The team then conducted testing at both geometric and dynamic scale and used the results of those tests to determine if actual loading would potentially damage the vehicle. The paper demonstrates the utility of such an approach using the RQ-7 case study. Key aspects of the approach were the use of a model of the vehicle to determine the likely loading conditions that would lead to material failure. An analysis of the important scaling characteristics was conducted and a novel, non-dimensional ratio was developed. Scaled testing was conducted using instrumentation to determine the unknown scaling parameters. The results were then compared to the actual vehicle through the use of the non-dimensional ratio. The ratio compared the size, the approach speed, and the mass of the model to the actual air vehicle with the ultimate goal of determining whether the decelerations experienced by the model when impacting the airbag, would result in damage to key components on the vehicle when decelerated at full-scale. Testing showed that the model airbag was capable of adequately decelerating the UAV, although improvements needed to be made for greater reliability.
{"title":"Scaled Crash Testing Using Modeling, Similitude, and Experimentation","authors":"R. Melnyk, Olivia Beattie, B. Waller","doi":"10.1115/imece2021-66606","DOIUrl":"https://doi.org/10.1115/imece2021-66606","url":null,"abstract":"\u0000 Due to the complexity and challenging nature of predicting the effects of impact on a vehicle, much of the crash-testing industry remains one which operates at full-scale. Automobiles and even aircraft are often crashed at full size, which is an expensive and time-consuming process. This paper details a novel approach to crash-testing that combines early use of simulation, scaling of the appropriate parameters using similitude, and limited testing. The method was used to predict the crashworthiness of a small, commercial unmanned aerial vehicle (UAV).\u0000 The authors were involved in an undergraduate design project to develop a system capable of safely decelerating the RQ-7 Shadow Unmanned Aerial System (UAS) from an operational flight without the use of a runway. For contingency reasons or tactical purposes, the use of a larger runway for the UAS was not feasible. The team developed an airbag-like system that used controlled release of air to decelerate the vehicle. However, due to testing constraints the team was unable to test the airbag system at full-scale with an actual RQ-7 vehicle. The system is already far in it’s life-cycle so there were no prototypes available and the team’s available facilities would not facilitate a large, high-speed test.\u0000 As a result, the team built a model of the existing vehicle, focusing on key components and materials. The team then conducted testing at both geometric and dynamic scale and used the results of those tests to determine if actual loading would potentially damage the vehicle. The paper demonstrates the utility of such an approach using the RQ-7 case study.\u0000 Key aspects of the approach were the use of a model of the vehicle to determine the likely loading conditions that would lead to material failure. An analysis of the important scaling characteristics was conducted and a novel, non-dimensional ratio was developed. Scaled testing was conducted using instrumentation to determine the unknown scaling parameters. The results were then compared to the actual vehicle through the use of the non-dimensional ratio. The ratio compared the size, the approach speed, and the mass of the model to the actual air vehicle with the ultimate goal of determining whether the decelerations experienced by the model when impacting the airbag, would result in damage to key components on the vehicle when decelerated at full-scale. Testing showed that the model airbag was capable of adequately decelerating the UAV, although improvements needed to be made for greater reliability.","PeriodicalId":146533,"journal":{"name":"Volume 13: Safety Engineering, Risk, and Reliability Analysis; Research Posters","volume":"70 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121221427","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：