| The requirement of demand of energy in various fields is increasing continuously since,last century.However,industries are using the maximum amount of energy but the transportation sector also needs enormous energy.Diesel and gasoline are the primary fuels for road vehicles which are main sources of harmful emissions in populate areas.The natural gas emerged as a very effective option for IC engines in past 4-5 decades.Moreover,hydrogen fueled engines can be also a substitution for primary fuels.HCNG engines are much better compared to NG and H2 fueled engines in various aspects.The presented work not only researched the key features of HCNG fueled engine under various excess air ratios,various ignition timings,and two different MAPs for three distinct fuels mixture also includes the study of LBV of CO added HCNG blends.The experiments on Weichai WP6-type HCNG SI engine were done under various HCNG fuels blends,excess air ratios,ignition timings,and manifold absolute pressures.The performance variables like effective efficiency,BSFC,and torque output were analyzed.The results reveal that the hike in excess air ratio,MAP and H2 fraction,torque significantly decreases and BSFC lowers when reached to lowest value and further it again increases.The emission traces like NOx,CO,THC,and CH4 all decreases with increase in ignition advanced angle,and hike of MAP.Herein,the turbulent entrainment QDC model has been generated to study the engine process analytically.Well known turbulent entrainment quasi-dimensional combustion model was built for precise matching of pressure profiles.It is found that the ignition lag coefficient(Cig)always rest at value 1.52 while Turbulent intensity coefficient?C2?rest at 1.05,1.05,and 1.22 for NG,20%H2 and 40%H2 fraction respectively for?=1.23 while Taylor length scale coefficient?C3?changes from 1.6-2.88,1.8-4.84,and 0.5-1.56 respectively for the above.Meanwhile,machine learning is also adopted for prediction of HCNG engine's performance parameters and emissions.For this,the ANN model is built and trained to predict the HCNG engine parameters.The minimum and maximum values of R and MSE were evaluated for various engine parameters,like BSFC,torque,NOx,CO,THC,and CH4emissions.The results show that the values of R?min-max?,and MSE?min-max?were found as for BSFC 0.9519-0.9963 and 5.72-64.29,for torque;0.9948-0.9997 and 8.3935-95.8350,for NOx;0.9938-0.9992 and 0.5271-3.4340,for CO;0.8595-0.9960 and 0.0188-1.0154,for THC;0.9644-0.9972 and 0.0714-1.1695,and for CH4;0.7965-0.9973,0.0351-1.1789.Later,the effect of CO addition of HCNG fuel's laminar flame speed was studied with the help of combustion bomb experiments.A special terminology has been designed for this three gas fuel blend which is Hy CONG.Generally,the consequences of CO doping on the laminar flame characteristics of HCNG mixtures was studied under three equivalence ratios.It is found that the peak LBV of Hy CONG mixture are 24.43 cm/s,45.71 cm/s,and 62.79cm/s for 20-80 Hy CONG?at?=0.6?,40-60 Hy CONG?at?=0.8?,and 40-60 Hy CONG?at?=1.0?respectively.Additionally,the numerical study of LBV of Hy CONG blends performed with chemical kinetic study software CHEMKIN.It is observed that the maximum and minimum error in LBV are:for 0-y Hy CONG;7.76%-0.07%at?=0.6,and?=0.8 respectively.For 20-y Hy CONG;11.91%-0.4%at?=0.6,and 1.0 and 40-y Hy CONG;13.71%-0.12%at?=0.6,and 1.0 respectively.The machine learning approach was applied to predict LBV of Hy CONG by ANN modeling.LBV comparison by experiments and ANN modeling have been presented in this work.The result data reveals that the R and MSE are 0.9986?max?-0.9944?min?&2.1305?max?-1.0195?min?for LM and 0.9990?max?-0.9952?min?&2.7965?max?-0.9501?min?for SCG algorithm.The research achievements of this work expanded the theory of combustion characteristics of fuels also the understanding of HCNG engine in engine research community.Additionally,this work reveals the application of machine learning approach in both the HCNG engine research and fuel combustion research community. |
PDF Full Text Request