View Record

TitleMonte Carlo simulations of methanol absorption and clustering in polyvinylchloride and high density polyethylene
AuthorGanesh, Shresta
SubjectMethanol--Absorption and adsorption
SubjectMonte Carlo method
SubjectPolyvinyl chloride
SubjectHigh-density polyethylene
SubjectCluster analysis
SubjectPolymers
Date2017-09-12T11:38:44Z
Date2017-09-12T11:38:44Z
Date2017
TypeThesis
Format106 p
AbstractSubmitted in fulfilment of the Degree of Master of Engineering in Chemical Engineering, Durban University of Technology, Durban, South Africa, 2017.
AbstractA pertinent issue facing the materials industry is that of the lifespan of materials when exposed to certain solvents/ environments. Limitations to the applications of materials require further research into understanding their failure mechanisms and how such problems can be addressed in terms of re-engineering such materials to be more durable. PVC and HDPE are two common polymers used extensively in industrial applications. Gibbs ensemble Monte Carlo simulations were used to simulate systems of PVC and HDPE with methanol, respectively. The temperatures used in this work were 25, 30 and 40 oC and all systems were at atmospheric pressure. Laboratory tests were used to complement simulations for analysis purposes. Results showed an increase in temperature resulted in a decrease in the overall clustering in both polymers although PVC displayed a greater decrease than HDPE. Linear clustering dominated over other forms of clustering with increases in temperature, with dimers being the most prevalent topology type. The results of this study suggest that the presence of chlorine atoms in PVC may not directly affect clustering of absorbed methanol, and their effect may instead be indirect by means of altering the accessible free volume within the polymer. Swelling was also investigated in the simulated systems and it was found that PVC displayed a greater degree of swelling than HDPE despite its lower rate of clustering. The effect of cluster radius on the cluster analysis was also considered.
AbstractM
Identifier676190
Identifierhttp://hdl.handle.net/10321/2542