Performance assessment of multiple biodiesel blended diesel engine and NOx modeling using ANN. (December 2021)

• Record Type: Journal Article
• Authors: Sayyed, Siraj; Das, Randip Kumar; Kulkarni, Kishor
• Is Part Of: Case studies in thermal engineering. Volume 28(2021)
• Publication Date: 2021-12
• Subjects: Diesel -- Biodiesel -- Characterization -- Engine characteristics -- Modeling -- ANN; Heat engineering -- Case studies -- Periodicals; 621.40205
• DOI: 10.1016/j.csite.2021.101509
• Languages: English
• ISSNs: 2214-157X

Abstract

Abstract: This paper discusses the effect of dual blends of multiple biodiesels on characteristics of DICI engine in conjunction with NOx modeling using ANN. Six sets of dual biodiesel/diesel blends are prepared on a volume basis (10% and 90%) by using four different biodiesels such as Jatropha, Karanja, Mahua, and Neem. The experimentations are performed on a single-cylinder VCR engine with full throttle condition and constant compression ratio (CR) of 17.5:1. The physicochemical and thermal properties are analysed to assess the suitability and compatibility of biodiesel as an alternative fuel. The physicochemical properties are found within the permissible limits as prescribed by ASTM. Mass loss study and functional groups are analysed through TGA and FTIR analysis and maximum weight loss is observed in pyrolytic zone (Zone-II). Reduction in performance characteristics of BTE, AFR η M e c h ., η V o l . varies from −0.77% to −5.60%. Variations are observed in BSEC (8.91%) and EGT (11.95%). Combustion characteristics (CP, MGT, MFB, NHR, and RPR) of biodiesel blends are similar to neat diesel. Overall reduction in CO (34.05%) and increment in CO2 (42.54%) are observed for all the blends at the penalty of HC and NOx. An attempt has been made to predict NOx emission through physicochemical properties. Highlights: Performance, Combustion, and Emission characteristics of DICI engine using mixture of multiple biodiesels. At full throttle condition, BSEC and EGT are similar for all blends. An emission characteristics observed as CO2 (42.54%) and CO (−34.05%). Blend D80+JB10+KB10 shows improved results in comparison with neat diesel. ANN predicts NOx with an accuracy of 0.997 from physicochemical properties.