Font Size: a A A

Supercritical Two-step Method Numerical Simulation Of Biodiesel

Posted on:2010-04-30Degree:MasterType:Thesis
Country:ChinaCandidate:P ZhuFull Text:PDF
GTID:2191330332978105Subject:Chemical Process Equipment
Abstract/Summary:PDF Full Text Request
Biodiesel, as the perfect substitutite of fossil diesel, is one kind of typical "green Energy sources". As for its good combustibility, renewable resource, environmental friendly, biodiesel has become the focus for enterprises and researchers.The process of heat transfer,rate of hydrolysis in sub-critical water and rate of esterification in supercritical methanol of producing biodiesel in two-step supercritical method are studied by numerical simulation.The CFD software-FLUENT is used for modeling and solution during numerical simulation of heat transfer.Artificial neural networks(ANN) are used for nonlinear mapping during numerical simulation of hydrolysis rate and esterification rate. Based on the stimulant results of intermission reaction device,a kind of continuous reaction device that similar to the shell tubular heat exchanger is designed.The mathematics model of heat transfer is established by FLUENT, including vapor,pipe and reactant,and distribution of temperature field changed with time is obtained.The simulant results of continuous reaction device indicate that asymmetry of distribution of temperature field is improved effectively. The length of pipe is the key link of continuous reaction device design,which mainly includes reactant pass through the pipe length during temperature increment and the constant temperature reaction stage,and both are decided by the flow rate of reantant in pipe.Therefore the relationships between different flow rate of reantant and reactant pass through the pipe length during temperature.It has been shown by research that ANN that are designed properly can mapping the complicated nonlinear relation with arbitrary precision. Based on this,ANN models are used to realize the nonlinear mapping between input parameter and output parameter of hydrolysis and esterification reaction.Namely,simulate the relationship between the hydrolysis rate and volume ratio of water to fatty oil, reaction temperature, reaction time, the relationship between the esterification rate and volume ratio of methanol to fatty acid, reaction temperature, reaction time.Selecting the type of ANN model.After analysing the characteristic of different ANN,BP ANN and Elman ANN are applied to simulate hydrolysis reaction and esterification reaction.Then the system error and train time are contrasted,optimal NN structures are obtained, which are 3-8-1 and 3-14-1..But BP NN is better than Elman NN at the aspect of generalization, whose correlation coefficient is 0.998, so hydrolysis and esterification reaction are simulated by BP NN.Choosing and Processing sample datas.Uniform design is put forward to choose the sample,by which not only the number of samples needs is small but also the sample datas are representative.In order to improve the performance of ANN,sample datas are processed by the tool box of MATLAB software,which ensure conherence of sample datas.
Keywords/Search Tags:super-critical, biodiesel, numerical simulation, FLUENT software, temperature field, Artificial neural networ
PDF Full Text Request
Related items