Data Mining And Group Contribution Calculation Of Polyester Modification

In this paper the big data technology was been applied in PET modification research for the first time. More than 500 groups data has been collected from the literature, which concerns about PET modification conditions and results, such as isophthalic acid(IPA), polyethylene glycol(PEG400--6000), 5-sodium sulfo bis-(hydroxyethyl) isophthalate(SIPE),5-sulfoisophthalic acid sodium salt(SIPA), sodium sulfoisophthalate acid dimethyl ester(SIPM), with different adding amount(0--20%), the properties including glass transition temperature(Tg), melting temperature(Tm), crystallization temperature(Tc), intrinsic viscosity[η].The data has been filtered and list into a table. With the mining technology the data was processed by data statistics software IBM SPSS 22.0 to carry on the factor analysis and the cluster analysis.Then the data was special analyzed. The properties(such as Tg,Tm,Tc,[η]) ranges and the corresponding modifier ranges has been obtained.According to the group contribution principle, the properties of PET can be divided into different parts which contributed by different components. In this thesis the properties of the copolyesters modified with IPA or PEG4000 can be calculated from the contribution of different elements. The calculating formula has been deduced for the glass transition temperature and the melting temperature of IPA modified PET and PEG4000 modified PET.From the data mining and the group contribution calculating results above, several different experimental conditions has been chosen randomly. The corresponding copolyesters has been obtained. The properties test results have matched very well with the data mining and group contribution calculating.This thesis results have shown that the data mining technology and group contribution calculating are very valuable for both the theory and the actual modification of PET.