Return to search

Reverse analysis of the epoxy kinetic model : A search for kamal model parameters to fit measured data

This report is about the curing of epoxy and how to find a model to fit the measured temperature data. The curing process of the polymer was modeled using Kamal model, which includes several parameters. The goal of the report was to gain an understanding about the Kamal model and also to learn how to approach the problem of finding these parameter values using measured data of the temperature of the epoxy during its exothermic reaction. Using the heat equation we were able to find a system of equation that describes the temperature of the epoxy. To understand the parameters, we changed each parameter drastically which gave us an intuition of the Kamal model. In an attempt to fit the measured data parameters we first changed the parameters by trial and error. Thereafter, an optimization method was implemented that given an initial guess iteratively changed the parameters to approach the measured data. The 'approach' was quantitatively measured by using a loss function that measures the closeness of the simulated and the measured data. By having a large grid of starting guesses, many local minimas were found and the best fitted parameters were documented. The achieved results were inconclusive as the model didn't fit the exothermic peak sufficiently, but the goal of our report to create an approach to this problem was still successful. To further improve our model, all the assumptions in our model should be analyzed and possibly revised, also more datasets would have to be fitted to draw further conclusions.

Identiferoai:union.ndltd.org:UPSALLA1/oai:DiVA.org:uu-445627
Date January 2021
CreatorsAsalya, Oday, Fjällborg, Joar, Hagert, Lucas
Source SetsDiVA Archive at Upsalla University
LanguageEnglish
Detected LanguageEnglish
TypeStudent thesis, info:eu-repo/semantics/bachelorThesis, text
Formatapplication/pdf
Rightsinfo:eu-repo/semantics/openAccess
RelationMATVET-F

Page generated in 0.0176 seconds