Experiments And Modeling Of Producer Gas Based Reciprocating Engines From the article: A 20 kW reciprocating engine is operated using producer gas derived from a modern open top downdraft re-burn biomass gasifier that has been evaluated by rigorous laboratory performance testing over several hundred hours. The engine is operated at varying compression ratio (CR) from 11.5 to 17.0 and ignition timings from 30 to 6° before Top Centre (TC). The engine – alternator system is characterised for its performance by the simultaneous measurement of gas and airflow rates, gas composition (online), emission levels and power delivered. It is also instrumented to obtain the in-cylinder behaviour in the form of pressure-crank angle (p - θ) diagram to assess the thermodynamic behaviour of the engine. Three-dimensional (3-D) simulation of the flow field in the combustion chamber (involving piston-bowl arrangement) through the cycle up to the start of the combustion is used to obtain inputs on the turbulence intensity ( u' ) and length scale ( lT ) for the modeling of the flame propagation process in a zero dimensional model (0-D) designed to predict the p – θ curve. The flame propagation and heat release processes make use of eddy entrainment and laminar burn-up model. The data on u' extracted from the 3-D flow calculations match reasonably well with experiments till compression stroke but are in contradiction with trends close to TC. This is reasoned to be due to limitation of the k-ε model to capture transient effects due to reverse squish phenomenon. The 0-D model took into account the experimental behavior of the u' in the post-TC period to attempt to match the observed p - θ data over a range of CRs and ignition timing advances. While these predictions match well with the experimental data at advanced ignition timing at both higher and lower CRs, the peak pressure is under-predicted at lower ignition advances; reason are traced to increase in flame area and propagation speed due to reverse squish effect. When these are accounted in the model, the p - θ curves are predicted better. About Us Pellentesque habitant morbi tristique senectus et netus et malesuada fames ac turpis egestas. Vestibulum tortor quam, feugiat vitae, ultricies eget, tempor sit amet, ante. Pellentesque habitant morbi tristique senectus. Address Twitter, Inc. 795 Folsom Ave, Suite 600 San Francisco, CA 94107 P: (123) 456-7890
This is a companion discussion topic for the original entry at http://dev.driveonwood.com/library/experiments-modeling/