E. Fornaro, Massimo Cardone, M. Terzo, S. Strano, C. Tordela
In the aeronautical field, aircraft reliability is strictly dependent on propulsion systems. Indeed, a reliable propulsion system ensures the safety of pilots and passengers and the possibility of making comfortable flights. Typically, on aircraft equipped with spark ignition (SI) engines, one of the principal requirements to make them reliable is the correct balancing between the intake air mass and fuel flows. Advances in the implementation of sophisticated control and estimation strategies on SI engines allow realizing engines with improved features in terms of performance, reducing pollution emissions, and fuel consumption. Approaches based on sensor redundancy are applied to improve the reliability in measurements of the manifold air pressure (MAP) and flow (MAF) to avoid issues related to possible faults of sensors vital for the correct functioning of SI engines. Model-based estimation techniques, based on the speed–density and alpha-speed methods for determining the MAF in engine control units, are employed to obtain sensor-less redundancy. The prediction of MAP and MAF, for sensors redundancy purposes, can be made through neural networks, allowing the avoidance of effects due to unmodeled dynamical behaviors. A sensor redundancy approach based on feedforward neural networks (FNNs) is proposed in this work for MAP and MAF prediction of a SI engine. The present work focuses on the possibility of estimating the physical quantities related to SI engines, such as the MAP and the MAF, fundamental for their monitoring using neural networks trained by means of a model-based approach avoiding expensive experimental tests for producing training data. A well-known intake manifold dynamical model (IMDM), parametrized based on the CMD 22 aeronautical engine, is employed for generating synthetic training data in steady-state conditions functional for making the chosen FNNs able to predict both MAP and MAF even in transient behavior. The MAP and MAF are predicted through two virtual sensors based on two independent FNNs, having the same inputs, constituted by the engine speed and the throttle angle. An experimental investigation based on an aircraft endurance test of two hours proposed by the European Aviation Safety Agency (EASA) has been made on a controlled and monitored CMD 22 engine for comparing the experimentally measured MAP and MAF with the predicted ones by the FNNs. The results demonstrate the suitability of the proposed approach for sensor redundancy purposes in SI engines to increase their reliability.
{"title":"Experimentally Validated Neural Networks for Sensors Redundancy\u0000 Purposes in Spark Ignition Engines","authors":"E. Fornaro, Massimo Cardone, M. Terzo, S. Strano, C. Tordela","doi":"10.4271/03-17-02-0012","DOIUrl":"https://doi.org/10.4271/03-17-02-0012","url":null,"abstract":"In the aeronautical field, aircraft reliability is strictly dependent on\u0000 propulsion systems. Indeed, a reliable propulsion system ensures the safety of\u0000 pilots and passengers and the possibility of making comfortable flights.\u0000 Typically, on aircraft equipped with spark ignition (SI) engines, one of the\u0000 principal requirements to make them reliable is the correct balancing between\u0000 the intake air mass and fuel flows. Advances in the implementation of\u0000 sophisticated control and estimation strategies on SI engines allow realizing\u0000 engines with improved features in terms of performance, reducing pollution\u0000 emissions, and fuel consumption. Approaches based on sensor redundancy are\u0000 applied to improve the reliability in measurements of the manifold air pressure\u0000 (MAP) and flow (MAF) to avoid issues related to possible faults of sensors vital\u0000 for the correct functioning of SI engines. Model-based estimation techniques,\u0000 based on the speed–density and alpha-speed methods for determining the MAF in\u0000 engine control units, are employed to obtain sensor-less redundancy. The\u0000 prediction of MAP and MAF, for sensors redundancy purposes, can be made through\u0000 neural networks, allowing the avoidance of effects due to unmodeled dynamical\u0000 behaviors. A sensor redundancy approach based on feedforward neural networks\u0000 (FNNs) is proposed in this work for MAP and MAF prediction of a SI engine. The\u0000 present work focuses on the possibility of estimating the physical quantities\u0000 related to SI engines, such as the MAP and the MAF, fundamental for their\u0000 monitoring using neural networks trained by means of a model-based approach\u0000 avoiding expensive experimental tests for producing training data. A well-known\u0000 intake manifold dynamical model (IMDM), parametrized based on the CMD 22\u0000 aeronautical engine, is employed for generating synthetic training data in\u0000 steady-state conditions functional for making the chosen FNNs able to predict\u0000 both MAP and MAF even in transient behavior.\u0000\u0000 \u0000The MAP and MAF are predicted through two virtual sensors based on two\u0000 independent FNNs, having the same inputs, constituted by the engine speed and\u0000 the throttle angle. An experimental investigation based on an aircraft endurance\u0000 test of two hours proposed by the European Aviation Safety Agency (EASA) has\u0000 been made on a controlled and monitored CMD 22 engine for comparing the\u0000 experimentally measured MAP and MAF with the predicted ones by the FNNs. The\u0000 results demonstrate the suitability of the proposed approach for sensor\u0000 redundancy purposes in SI engines to increase their reliability.","PeriodicalId":47948,"journal":{"name":"SAE International Journal of Engines","volume":null,"pages":null},"PeriodicalIF":1.2,"publicationDate":"2023-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89526010","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. Brusa, Fenil Panalal Shethia, Jacopo Mecagni, N. Cavina
In the last few years, the artificial neural networks have been widely used in the field of engine modeling. Some of the main reasons for this are, their compatibility with the real-time systems, higher accuracy, and flexibility if compared to other data-driven approaches. One of the main difficulties of using this approach is the calibration of the network itself. It is very difficult to find in the literature procedures that guide the user to completely define a network. Typically, the very last steps (like the choice of the number of neurons) must be selected by the user on the base of his sensitivity to the problem. This work proposes an automatic calibration procedure for the artificial neural networks, considering all the main hyper-parameters of the network such as the training algorithms, the activation functions, the number of the neurons, the number of epochs, and the number of hidden layers, for modeling various combustion indexes in a modern internal combustion engine. However, the proposed procedure can be applied to the training of any neural network-based model. The automatic calibration procedure outputs a configuration of the network, giving the optimal combination in terms of hyper-parameters. The decision of the optimal configuration of the neural network is based on a self-developed formula, which gives a rank of all the possible hyper-parameter combinations using some statistical parameters obtained comparing the simulated and the experimental values. In the end, the lowest rank is selected as the optimal one as it represents the combination having the lowest error. Following the definition of this rank, high accuracy on the results has been achieved in terms of the root mean square error index, for example, on the combustion phase model, the error is 0.139°CA under steady-state conditions. On the maximum in-cylinder pressure model, the error is 1.682 bar, while the knock model has an error of 0.457 bar for the same test that covers the whole engine operating field.
{"title":"Advanced, Guided Procedure for the Calibration and Generalization of\u0000 Neural Network-Based Models of Combustion and Knock Indexes","authors":"A. Brusa, Fenil Panalal Shethia, Jacopo Mecagni, N. Cavina","doi":"10.4271/03-17-02-0009","DOIUrl":"https://doi.org/10.4271/03-17-02-0009","url":null,"abstract":"In the last few years, the artificial neural networks have been widely used in\u0000 the field of engine modeling. Some of the main reasons for this are, their\u0000 compatibility with the real-time systems, higher accuracy, and flexibility if\u0000 compared to other data-driven approaches. One of the main difficulties of using\u0000 this approach is the calibration of the network itself. It is very difficult to\u0000 find in the literature procedures that guide the user to completely define a\u0000 network. Typically, the very last steps (like the choice of the number of\u0000 neurons) must be selected by the user on the base of his sensitivity to the\u0000 problem.\u0000\u0000 \u0000This work proposes an automatic calibration procedure for the artificial neural\u0000 networks, considering all the main hyper-parameters of the network such as the\u0000 training algorithms, the activation functions, the number of the neurons, the\u0000 number of epochs, and the number of hidden layers, for modeling various\u0000 combustion indexes in a modern internal combustion engine. However, the proposed\u0000 procedure can be applied to the training of any neural network-based model.\u0000\u0000 \u0000The automatic calibration procedure outputs a configuration of the network,\u0000 giving the optimal combination in terms of hyper-parameters. The decision of the\u0000 optimal configuration of the neural network is based on a self-developed\u0000 formula, which gives a rank of all the possible hyper-parameter combinations\u0000 using some statistical parameters obtained comparing the simulated and the\u0000 experimental values. In the end, the lowest rank is selected as the optimal one\u0000 as it represents the combination having the lowest error. Following the\u0000 definition of this rank, high accuracy on the results has been achieved in terms\u0000 of the root mean square error index, for example, on the combustion phase model,\u0000 the error is 0.139°CA under steady-state conditions. On the maximum in-cylinder\u0000 pressure model, the error is 1.682 bar, while the knock model has an error of\u0000 0.457 bar for the same test that covers the whole engine operating field.","PeriodicalId":47948,"journal":{"name":"SAE International Journal of Engines","volume":null,"pages":null},"PeriodicalIF":1.2,"publicationDate":"2023-08-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86156188","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}
Pre-chamber turbulent jet ignition (TJI) is a method of generating distributed ignition sites through multiple high-speed turbulent jets in order to achieve an enhanced burn rate in the engine cylinder when compared to conventional spark plug ignition. To study the gas-dynamic interactions between the two chambers in a gasoline engine, a three-dimensional numerical model was developed using the commercial CFD code CONVERGE. The geometry and parameters of the engine used were based on a modified turbocharged GM four-cylinder 2.0 L GDI gasoline engine. Pre-chambers with nozzle diameters of 0.75 mm and 1.5 mm were used to investigate the effect of pre-chamber geometry on pre-chamber charging, combustion, and jet formation. The local developments of gas temperature and velocity were captured by adaptive mesh refinement, while the turbulence was resolved with the k-epsilon model of the Reynolds averaged Navier–Stokes (RANS) equations. The combustion process was modeled with the extended coherent flamelet model (ECFM). Data from engine experiments were compared with the computed main chamber pressures and heat release rates, and the results show good consistency with the model calculations. The scavenging and air–fuel equivalence ratio (λ) distribution of the pre-chambers improved with the larger nozzle, while the smaller nozzle generated jets with higher velocity, greater turbulence kinetic energy, and longer penetration length. Moreover, after the primary jet formation, secondary pre-chamber charging, combustion, and secondary jet formation were observed.
{"title":"Gas-Dynamic Interactions between Pre-Chamber and Main Chamber in\u0000 Passive Pre-Chamber Ignition Gasoline Engines","authors":"Tianxiao Yu, Dong Eun Lee, Jay P. Gore, Li Qiao","doi":"10.4271/03-17-01-0008","DOIUrl":"https://doi.org/10.4271/03-17-01-0008","url":null,"abstract":"Pre-chamber turbulent jet ignition (TJI) is a method of generating distributed\u0000 ignition sites through multiple high-speed turbulent jets in order to achieve an\u0000 enhanced burn rate in the engine cylinder when compared to conventional spark\u0000 plug ignition. To study the gas-dynamic interactions between the two chambers in\u0000 a gasoline engine, a three-dimensional numerical model was developed using the\u0000 commercial CFD code CONVERGE. The geometry and parameters of the engine used\u0000 were based on a modified turbocharged GM four-cylinder 2.0 L GDI gasoline\u0000 engine. Pre-chambers with nozzle diameters of 0.75 mm and 1.5 mm were used to\u0000 investigate the effect of pre-chamber geometry on pre-chamber charging,\u0000 combustion, and jet formation. The local developments of gas temperature and\u0000 velocity were captured by adaptive mesh refinement, while the turbulence was\u0000 resolved with the k-epsilon model of the Reynolds averaged Navier–Stokes (RANS)\u0000 equations. The combustion process was modeled with the extended coherent\u0000 flamelet model (ECFM). Data from engine experiments were compared with the\u0000 computed main chamber pressures and heat release rates, and the results show\u0000 good consistency with the model calculations. The scavenging and air–fuel\u0000 equivalence ratio (λ) distribution of the pre-chambers improved with the larger\u0000 nozzle, while the smaller nozzle generated jets with higher velocity, greater\u0000 turbulence kinetic energy, and longer penetration length. Moreover, after the\u0000 primary jet formation, secondary pre-chamber charging, combustion, and secondary\u0000 jet formation were observed.","PeriodicalId":47948,"journal":{"name":"SAE International Journal of Engines","volume":null,"pages":null},"PeriodicalIF":1.2,"publicationDate":"2023-08-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"72989774","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. Brusa, Jacopo Mecagni, Fenil Panalal Shethia, E. Corti
A previously developed piston damage and exhaust gas temperature models are coupled to manage the combustion process and thereby increasing the overall energy conversion efficiency. The proposed model-based control algorithm is developed and validated in a software-in-the-loop simulation environment, and then the controller is deployed in a rapid control prototyping device and tested online at the test bench. In the first part of the article, the exhaust gas temperature model is reversed and converted into a control function, which is then implemented in a piston damage-based spark advance controller. In this way, more aggressive calibrations are actuated to target a certain piston damage speed and exhaust gas temperature at the turbine inlet. A more anticipated spark advance results in a lower exhaust gas temperature, and such decrease is converted into lowering the fuel enrichment with respect to the production calibrations. Moreover, the pollutant emissions associated with production calibrations and the implementation of the developed controller are compared through a GT-Power combustion model. Finally, the complete controller is validated for both the transient and steady-state conditions, reproducing a real vehicle maneuver at the engine test bench. The results demonstrate that the combination of an accurate estimation of the damage induced by knock and the value of the exhaust gas temperature allows to reduce the brake specific fuel consumption by up to 20%. Moreover, the stoichiometric area of the engine operating field is extended by 20%, and the GT-Power simulations show a maximum CO reduction of about 50%.
{"title":"Model-Based Combustion Control to Reduce the Brake Specific Fuel\u0000 Consumption and Pollutant Emissions under Real Driving Maneuvers","authors":"A. Brusa, Jacopo Mecagni, Fenil Panalal Shethia, E. Corti","doi":"10.4271/03-17-01-0007","DOIUrl":"https://doi.org/10.4271/03-17-01-0007","url":null,"abstract":"A previously developed piston damage and exhaust gas temperature models are\u0000 coupled to manage the combustion process and thereby increasing the overall\u0000 energy conversion efficiency. The proposed model-based control algorithm is\u0000 developed and validated in a software-in-the-loop simulation environment, and\u0000 then the controller is deployed in a rapid control prototyping device and tested\u0000 online at the test bench. In the first part of the article, the exhaust gas\u0000 temperature model is reversed and converted into a control function, which is\u0000 then implemented in a piston damage-based spark advance controller. In this way,\u0000 more aggressive calibrations are actuated to target a certain piston damage\u0000 speed and exhaust gas temperature at the turbine inlet. A more anticipated spark\u0000 advance results in a lower exhaust gas temperature, and such decrease is\u0000 converted into lowering the fuel enrichment with respect to the production\u0000 calibrations. Moreover, the pollutant emissions associated with production\u0000 calibrations and the implementation of the developed controller are compared\u0000 through a GT-Power combustion model.\u0000\u0000 \u0000Finally, the complete controller is validated for both the transient and\u0000 steady-state conditions, reproducing a real vehicle maneuver at the engine test\u0000 bench. The results demonstrate that the combination of an accurate estimation of\u0000 the damage induced by knock and the value of the exhaust gas temperature allows\u0000 to reduce the brake specific fuel consumption by up to 20%. Moreover, the\u0000 stoichiometric area of the engine operating field is extended by 20%, and the\u0000 GT-Power simulations show a maximum CO reduction of about 50%.","PeriodicalId":47948,"journal":{"name":"SAE International Journal of Engines","volume":null,"pages":null},"PeriodicalIF":1.2,"publicationDate":"2023-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89901370","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}
Measuring the dynamic parameters of liquid fragments generated in the near-field of atomizing sprays poses a significant challenge due to the random nature of the fragments, the instability of the spray, and the limitations of current measuring technology. Precise determination of these parameters can aid in improving the control of the atomization process, which is necessary for providing suitable spray structures with appropriate flow rates and droplet size distributions for various applications such as those used in heat engines. In piston and gas turbine engines, controlling spray characteristics such as penetration, cone angle, particle size, and droplet size distribution is crucial to improve combustion efficiency and decrease exhaust emissions. This can be accomplished by adjusting the structural and/or operating parameters of the fuel supply system. This article aims to measure the breakup length, spray cone angle, axial velocity, breakup time, and liquid sheet film thickness for a swirl air-blast atomizer used in a gas-steam engine. The measurement was conducted using a shadowgraph imaging system developed specifically for this study, consisting of a high-speed camera, a lens, and a light source. While lasers are commonly used as light sources in the literature, this study utilized a special LED high-speed pulse light generator, which is cheaper, easier to handle, and provides a more uniform background. Images were processed using a MATLAB code developed for this study. Although the breakup zone is naturally random and the breakup location significantly varies with time, the novel method developed in this study helps quantify critical parameters under different operating conditions.
{"title":"A Novel Experiment Approach for Measurement Breakup Length, Cone\u0000 Angle, Sheet Velocity, and Film Thickness in Swirl Air-Blast\u0000 Atomizers","authors":"D. Phung, Thin V. Pham, Phuong X. Pham","doi":"10.4271/03-17-01-0006","DOIUrl":"https://doi.org/10.4271/03-17-01-0006","url":null,"abstract":"Measuring the dynamic parameters of liquid fragments generated in the near-field\u0000 of atomizing sprays poses a significant challenge due to the random nature of\u0000 the fragments, the instability of the spray, and the limitations of current\u0000 measuring technology. Precise determination of these parameters can aid in\u0000 improving the control of the atomization process, which is necessary for\u0000 providing suitable spray structures with appropriate flow rates and droplet size\u0000 distributions for various applications such as those used in heat engines. In\u0000 piston and gas turbine engines, controlling spray characteristics such as\u0000 penetration, cone angle, particle size, and droplet size distribution is crucial\u0000 to improve combustion efficiency and decrease exhaust emissions. This can be\u0000 accomplished by adjusting the structural and/or operating parameters of the fuel\u0000 supply system. This article aims to measure the breakup length, spray cone\u0000 angle, axial velocity, breakup time, and liquid sheet film thickness for a swirl\u0000 air-blast atomizer used in a gas-steam engine. The measurement was conducted\u0000 using a shadowgraph imaging system developed specifically for this study,\u0000 consisting of a high-speed camera, a lens, and a light source. While lasers are\u0000 commonly used as light sources in the literature, this study utilized a special\u0000 LED high-speed pulse light generator, which is cheaper, easier to handle, and\u0000 provides a more uniform background. Images were processed using a MATLAB code\u0000 developed for this study. Although the breakup zone is naturally random and the\u0000 breakup location significantly varies with time, the novel method developed in\u0000 this study helps quantify critical parameters under different operating\u0000 conditions.","PeriodicalId":47948,"journal":{"name":"SAE International Journal of Engines","volume":null,"pages":null},"PeriodicalIF":1.2,"publicationDate":"2023-07-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"72879189","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 urban roads the engine speed and the load vary suddenly and frequently, resulting in increased exhaust emissions. In such operations, the effect of air injection technique to access the transient response of the engine is of great interest. The effectiveness of air injection technique in improving the transient response under speed transient is investigated in detail [1]; however, it is not evaluated for the load transients. Load step demand of the engine is another important event that limits the transient response of the turbocharger. In the present study, response of a heavy-duty turbocharged diesel engine is investigated for different load conditions. Three cases of load transients are considered: constant load, load magnitude variation, and load scheduling. Air injection technique is simulated and after optimization of injection pressure based on orifice diameter, its effect on the transient response is presented. The results reveal that air injection into the intake manifold is an effective technique to improve the turbo lag of a heavy-duty turbocharged diesel engine under the transient conditions of load.
{"title":"Transient Response of Turbocharged Compression Ignition Engine under\u0000 Different Load Conditions","authors":"S. Saad, Asiya Rummana","doi":"10.4271/03-17-01-0005","DOIUrl":"https://doi.org/10.4271/03-17-01-0005","url":null,"abstract":"In urban roads the engine speed and the load vary suddenly and frequently,\u0000 resulting in increased exhaust emissions. In such operations, the effect of air\u0000 injection technique to access the transient response of the engine is of great\u0000 interest. The effectiveness of air injection technique in improving the\u0000 transient response under speed transient is investigated in detail [1]; however, it is not evaluated for the\u0000 load transients. Load step demand of the engine is another important event that\u0000 limits the transient response of the turbocharger. In the present study,\u0000 response of a heavy-duty turbocharged diesel engine is investigated for\u0000 different load conditions. Three cases of load transients are considered:\u0000 constant load, load magnitude variation, and load scheduling. Air injection\u0000 technique is simulated and after optimization of injection pressure based on\u0000 orifice diameter, its effect on the transient response is presented. The results\u0000 reveal that air injection into the intake manifold is an effective technique to\u0000 improve the turbo lag of a heavy-duty turbocharged diesel engine under the\u0000 transient conditions of load.","PeriodicalId":47948,"journal":{"name":"SAE International Journal of Engines","volume":null,"pages":null},"PeriodicalIF":1.2,"publicationDate":"2023-07-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89805364","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}
Naoki Yoneya, K. Amaya, Kengo Kumano, Y. Sukegawa, Yoshifumi Uchise, Hideo Jitsu, Yukio Fujiyama
For achieving high efficiency and low exhaust emissions, engines need to be operated near the limits of stable combustion, such as lean or exhaust gas recirculation (EGR) conditions. Sensing technologies of the combustion state by existing engine components are of high interest. And the utilization of voltage and current signals from ignition coils is discussed in this article. The discharge channel of an ignition spark is strongly affected by flow variation and spark plug surface conditions, and the behavior of discharge channel stretching and restrike event can vary greatly from cycle to cycle. As a result, the effects of flow velocity, temperature, pressure, and electrode surface resistance are compounded in the voltage-current response, making it difficult to accurately detect the combustion state for each cycle by a threshold judgment process using a single feature value. In this article, a method for inductively detecting misfires from voltage and current signals of ignition coils by applying deep learning image recognition is introduced. First, post-ignition for misfire detection is performed on the engine bench during the expansion stroke in an engine cycle, when the cylinder pressure is expected to differ between the combustion cycle and the ignition cycle, and the ignition coil voltage and current are measured. Next, a two-dimensional frequency distribution of voltage and current (discharge histogram) is created as an input image for deep learning, and the AlexNet model, which has been trained with more than one million images, is trained with images of the ignition and combustion cycles as a supervised learning. The accuracy of classification is then verified using a validation dataset. In addition, to making the deep learning model more explainable, the activation score distribution on the discharge histogram was visualized when the trained model judges the images, and the discharge characteristics that provided the basis for deep learning classifications were analyzed.
{"title":"Misfire Detection Technology with Deep Neural Network Based on\u0000 Ignition Coil Signals","authors":"Naoki Yoneya, K. Amaya, Kengo Kumano, Y. Sukegawa, Yoshifumi Uchise, Hideo Jitsu, Yukio Fujiyama","doi":"10.4271/03-17-01-0004","DOIUrl":"https://doi.org/10.4271/03-17-01-0004","url":null,"abstract":"For achieving high efficiency and low exhaust emissions, engines need to be\u0000 operated near the limits of stable combustion, such as lean or exhaust gas\u0000 recirculation (EGR) conditions. Sensing technologies of the combustion state by\u0000 existing engine components are of high interest. And the utilization of voltage\u0000 and current signals from ignition coils is discussed in this article. The\u0000 discharge channel of an ignition spark is strongly affected by flow variation\u0000 and spark plug surface conditions, and the behavior of discharge channel\u0000 stretching and restrike event can vary greatly from cycle to cycle. As a result,\u0000 the effects of flow velocity, temperature, pressure, and electrode surface\u0000 resistance are compounded in the voltage-current response, making it difficult\u0000 to accurately detect the combustion state for each cycle by a threshold judgment\u0000 process using a single feature value.\u0000\u0000 \u0000In this article, a method for inductively detecting misfires from voltage and\u0000 current signals of ignition coils by applying deep learning image recognition is\u0000 introduced. First, post-ignition for misfire detection is performed on the\u0000 engine bench during the expansion stroke in an engine cycle, when the cylinder\u0000 pressure is expected to differ between the combustion cycle and the ignition\u0000 cycle, and the ignition coil voltage and current are measured. Next, a\u0000 two-dimensional frequency distribution of voltage and current (discharge\u0000 histogram) is created as an input image for deep learning, and the AlexNet\u0000 model, which has been trained with more than one million images, is trained with\u0000 images of the ignition and combustion cycles as a supervised learning. The\u0000 accuracy of classification is then verified using a validation dataset. In\u0000 addition, to making the deep learning model more explainable, the activation\u0000 score distribution on the discharge histogram was visualized when the trained\u0000 model judges the images, and the discharge characteristics that provided the\u0000 basis for deep learning classifications were analyzed.","PeriodicalId":47948,"journal":{"name":"SAE International Journal of Engines","volume":null,"pages":null},"PeriodicalIF":1.2,"publicationDate":"2023-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74704854","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}
Joshua W. Stiborek, Ryan J. Tancin, Nathan J. Kempema, J. Szente, Michael J. Loos, C. Goldenstein
Quantifying exhaust gas composition and temperature in vehicles with internal combustion engines (ICEs) is crucial to understanding and reducing emissions during transient engine operation. This is particularly important before the catalytic converter system lights off (i.e., during cold start). Most commercially available gas analyzers and temperature sensors are far too slow to measure these quantities on the timescale of individual cylinder-firing events, thus faster sensors are needed. A two-color mid-infrared (MIR) laser absorption spectroscopy (LAS) sensor for gas temperature and carbon monoxide (CO) mole fraction was developed and applied to address this technology gap. Two quantum cascade lasers (QCLs) were fiber coupled into one single-mode fiber to facilitate optical access in the test vehicle exhaust. The QCLs were time-multiplexed in order to scan across two CO absorption transitions near 2013 and 2060 cm–1 at 15 kHz. This enabled in situ measurements of temperature and CO mole fraction to be acquired at 15 kHz in the engine-out exhaust of a research vehicle (modified production vehicle) with an 8-cylinder gasoline ICE. Three different vehicle tests were characterized with the LAS sensor as follows: (1) cold start with engine idle, (2) warm start with a drive cycle on a chassis dynamometer, and (3) hot start with a drive cycle on a chassis dynamometer. The measurements obtained from the LAS sensor had a time resolution that was three orders of magnitude faster than that of thermocouple and gas analyzer data acquired at the Ford vehicle emissions research laboratory (VERL) in Dearborn, Michigan. This enabled the LAS sensor to resolve high-speed engine dynamics and exhaust gas transients, which the conventional instrumentation could not, thereby providing valuable insight into the evolution of ICE emissions during transient engine operation.
{"title":"A Mid-Infrared Laser Absorption Sensor for Gas Temperature and Carbon\u0000 Monoxide Mole Fraction Measurements at 15 kHz in Engine-Out Gasoline Vehicle\u0000 Exhaust","authors":"Joshua W. Stiborek, Ryan J. Tancin, Nathan J. Kempema, J. Szente, Michael J. Loos, C. Goldenstein","doi":"10.4271/03-17-01-0002","DOIUrl":"https://doi.org/10.4271/03-17-01-0002","url":null,"abstract":"Quantifying exhaust gas composition and temperature in vehicles with internal combustion engines (ICEs) is crucial to understanding and reducing emissions during transient engine operation. This is particularly important before the catalytic converter system lights off (i.e., during cold start). Most commercially available gas analyzers and temperature sensors are far too slow to measure these quantities on the timescale of individual cylinder-firing events, thus faster sensors are needed. A two-color mid-infrared (MIR) laser absorption spectroscopy (LAS) sensor for gas temperature and carbon monoxide (CO) mole fraction was developed and applied to address this technology gap. Two quantum cascade lasers (QCLs) were fiber coupled into one single-mode fiber to facilitate optical access in the test vehicle exhaust. The QCLs were time-multiplexed in order to scan across two CO absorption transitions near 2013 and 2060 cm–1 at 15 kHz. This enabled in situ measurements of temperature and CO mole fraction to be acquired at 15 kHz in the engine-out exhaust of a research vehicle (modified production vehicle) with an 8-cylinder gasoline ICE. Three different vehicle tests were characterized with the LAS sensor as follows: (1) cold start with engine idle, (2) warm start with a drive cycle on a chassis dynamometer, and (3) hot start with a drive cycle on a chassis dynamometer. The measurements obtained from the LAS sensor had a time resolution that was three orders of magnitude faster than that of thermocouple and gas analyzer data acquired at the Ford vehicle emissions research laboratory (VERL) in Dearborn, Michigan. This enabled the LAS sensor to resolve high-speed engine dynamics and exhaust gas transients, which the conventional instrumentation could not, thereby providing valuable insight into the evolution of ICE emissions during transient engine operation.","PeriodicalId":47948,"journal":{"name":"SAE International Journal of Engines","volume":null,"pages":null},"PeriodicalIF":1.2,"publicationDate":"2023-07-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75945286","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}
Dat X. Nguyen, Cavicchi Andrea, K. Nguyen, Vu H. Nguyen, L. Postrioti, P. Pham
The shot-to-shot variations in common rail injection systems are primarily caused by pressure wave oscillations in the rail, pipes, and injector body. These oscillations are influenced by fuel physical properties, injector needle movement, and pressure and suction control valve activations. The pressure waves are generated by pump actuation and injector needle movement, and their frequency and amplitude are determined by fluid properties and flow path geometry. These variations can result in cycle-to-cycle engine fluctuations. In multi-injection and split-injection strategies, the pressure oscillation from the first shot can impact the hydraulic characteristics of subsequent shots, resulting in variations in injection rate and amount. This is particularly significant when using alternative fuels such as biodiesel, which aim to reduce emissions while maintaining fuel atomization quality. This study examines the shot-to-shot variations in a second-generation common rail system using cooking-oil-residue biodiesel. The results demonstrate that biodiesel properties impact pressure wave oscillation, shot-to-shot variation, and total injection rate. The study also finds that dwell time has a significant effect on the hydraulic characteristics of the second shot, with minimal influence up to a certain value. However, beyond a certain dwell time value (e.g., 0.8 ms in this study), the impact of dwell time on the pressure fluctuation generated by the second shot is limited. Conducting further research could help deepen our understanding of the influence of shot-to-shot deviations. This could involve exploring biodiesel spray characteristics using techniques such as shadowgraph imaging and studying the effect of these deviations on flame propagation and emission formation. Examining engine performance could also provide valuable insights into the effectiveness of the split-injection strategy and biodiesel blends. Additionally, characterizing biodiesel spray using the double-shot technique to examine spray penetration, cone angle, and/or spray impingement and combustion characteristics could be useful in linking it with the shot-to-shot variation investigated in this study. Such research can contribute to advancing the state-of-the-art knowledge on this topic.
{"title":"Shot-to-Shot Deviation of a Common Rail Injection System Operating\u0000 with Cooking-Oil-Residue Biodiesel","authors":"Dat X. Nguyen, Cavicchi Andrea, K. Nguyen, Vu H. Nguyen, L. Postrioti, P. Pham","doi":"10.4271/03-16-08-0062","DOIUrl":"https://doi.org/10.4271/03-16-08-0062","url":null,"abstract":"The shot-to-shot variations in common rail injection systems are primarily caused\u0000 by pressure wave oscillations in the rail, pipes, and injector body. These\u0000 oscillations are influenced by fuel physical properties, injector needle\u0000 movement, and pressure and suction control valve activations. The pressure waves\u0000 are generated by pump actuation and injector needle movement, and their\u0000 frequency and amplitude are determined by fluid properties and flow path\u0000 geometry. These variations can result in cycle-to-cycle engine fluctuations. In\u0000 multi-injection and split-injection strategies, the pressure oscillation from\u0000 the first shot can impact the hydraulic characteristics of subsequent shots,\u0000 resulting in variations in injection rate and amount. This is particularly\u0000 significant when using alternative fuels such as biodiesel, which aim to reduce\u0000 emissions while maintaining fuel atomization quality. This study examines the\u0000 shot-to-shot variations in a second-generation common rail system using\u0000 cooking-oil-residue biodiesel. The results demonstrate that biodiesel properties\u0000 impact pressure wave oscillation, shot-to-shot variation, and total injection\u0000 rate. The study also finds that dwell time has a significant effect on the\u0000 hydraulic characteristics of the second shot, with minimal influence up to a\u0000 certain value. However, beyond a certain dwell time value (e.g., 0.8 ms in this\u0000 study), the impact of dwell time on the pressure fluctuation generated by the\u0000 second shot is limited. Conducting further research could help deepen our\u0000 understanding of the influence of shot-to-shot deviations. This could involve\u0000 exploring biodiesel spray characteristics using techniques such as shadowgraph\u0000 imaging and studying the effect of these deviations on flame propagation and\u0000 emission formation. Examining engine performance could also provide valuable\u0000 insights into the effectiveness of the split-injection strategy and biodiesel\u0000 blends. Additionally, characterizing biodiesel spray using the double-shot\u0000 technique to examine spray penetration, cone angle, and/or spray impingement and\u0000 combustion characteristics could be useful in linking it with the shot-to-shot\u0000 variation investigated in this study. Such research can contribute to advancing\u0000 the state-of-the-art knowledge on this topic.","PeriodicalId":47948,"journal":{"name":"SAE International Journal of Engines","volume":null,"pages":null},"PeriodicalIF":1.2,"publicationDate":"2023-06-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"88091575","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}
The reasonable engine cooling system design can give a better cooling of engine, the coolant flow direction and different cooling structure designs have great impact on the cooling performance and fuel consumption of engine. Therefore, to gain a deeper understanding of the impact of different cooling system designs on engine cooling performance, three different split cooling structures and two oil–water heat exchanger (OWHE) layouts are designed for a two-cylinder motorcycle engine. Three-dimensional CFD analysis method is used for analyzing the coolant velocity distributions and one-dimensional systematic analysis method is used for analyzing the system flow rate at those cooling structure designs and OWHE designs. Meanwhile, experimental investigation of different cooling structures and OWHE layouts on fuel consumption is conducted by the bench test of worldwide motorcycle test cycle. Results indicated that the difference of coolant flow velocity distribution for four cooling structures are small and the flow resistance of Case D is lowest at fully opened thermostat condition. The fuel consumption of Case D is 4.78 L/100 km, 1.4% lower than that of Case A with the fuel consumption 4.85 L/100 km. The combined split cooling structure Case D and OWHE layout one is proven as the optimal cooling design with 4% fuel consumption reduction compared with that of original cooling structure Case A. The research results can provide theoretical reference for engine cooling performance evaluation and give valuable data to motorcycle designers for quick evaluation of design and quick solutions of improved design.
{"title":"Numerical Simulation and Experimental Investigation of Different\u0000 Cooling Structures on Cooling Performance and Fuel Consumption of a Two-Cylinder\u0000 Motorcycle Engine","authors":"Libin Tan, Yuejin Yuan, Can Huang","doi":"10.4271/03-16-08-0063","DOIUrl":"https://doi.org/10.4271/03-16-08-0063","url":null,"abstract":"The reasonable engine cooling system design can give a better cooling of engine,\u0000 the coolant flow direction and different cooling structure designs have great\u0000 impact on the cooling performance and fuel consumption of engine. Therefore, to\u0000 gain a deeper understanding of the impact of different cooling system designs on\u0000 engine cooling performance, three different split cooling structures and two\u0000 oil–water heat exchanger (OWHE) layouts are designed for a two-cylinder\u0000 motorcycle engine. Three-dimensional CFD analysis method is used for analyzing\u0000 the coolant velocity distributions and one-dimensional systematic analysis\u0000 method is used for analyzing the system flow rate at those cooling structure\u0000 designs and OWHE designs. Meanwhile, experimental investigation of different\u0000 cooling structures and OWHE layouts on fuel consumption is conducted by the\u0000 bench test of worldwide motorcycle test cycle. Results indicated that the\u0000 difference of coolant flow velocity distribution for four cooling structures are\u0000 small and the flow resistance of Case D is lowest at fully opened thermostat\u0000 condition. The fuel consumption of Case D is 4.78 L/100 km, 1.4% lower than that\u0000 of Case A with the fuel consumption 4.85 L/100 km. The combined split cooling\u0000 structure Case D and OWHE layout one is proven as the optimal cooling design\u0000 with 4% fuel consumption reduction compared with that of original cooling\u0000 structure Case A. The research results can provide theoretical reference for\u0000 engine cooling performance evaluation and give valuable data to motorcycle\u0000 designers for quick evaluation of design and quick solutions of improved\u0000 design.","PeriodicalId":47948,"journal":{"name":"SAE International Journal of Engines","volume":null,"pages":null},"PeriodicalIF":1.2,"publicationDate":"2023-06-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80774465","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：