Research On Computer Aided Homopolymer Property Prediction And Ethylene Polymerization Process Design

As the market demand for high performance polymer products increases,China’s polymer market is characterized by excess capacity at the low end and dependence on imports at the high end,and the key to upgrading the product structure of the polymer market is to change the polymer research and development-production model.An effective way to shorten the product development cycle is to customize the structure of polymer products according to their properties and optimize their reaction processes by using process systems engineering methodologies.To this end,this thesis focuses on homopolymer and copolymer with no branching and crosslinking,with the field of multi-level structure and property correlation of polymers,computer-aided polymer design,reaction process design for high performance polymer production,and reaction process optimization for a given product,with the aim of providing a systematic methodology for structure design and process development of high performance polymers.The main research of the thesis is divided into four parts as follows:(1)To address the problem that polymer properties are affected by multiple structural factors but the existing prediction methods predict polymer properties using only structure of repeating units,a method for predicting homopolymer properties considering the effect of number-average molecular weight is proposed,taking the glass transition temperature of key polymer properties as an example.Firstly,to solve the problem of inability to establish a oneto-one correspondence between structure and properties due to the range of property values in polymer handbooks and databases,limiting properties were used to standardize and unify the data,and the limiting glass transition temperature data set was established for 198 homopolymers in 16 classes.After that,the group contribution method was introduced to establish the homopolymer group structure-limit glass transition temperature correlation considering the side chain structure and polymer characteristics,and the R square of the model can reach 0.9925.Finally,the glass transition temperature-numbered average molecular weight correlation equation is used as a bridge to further calculate the temperature values at given number-average molecular weight.The results show that compared with the conventional method,which uses only the glass transition temperature to establish a one-to-one mapping relationship directly,this method uses more reasonable and accurate limiting property to build the correlation model,and the introduction of the number-average molecular weight can further describe the relationship between the glass transition temperature and multiple structural factors.(2)Computer-aided molecular design methods are capable of screening molecular structures based on property requirements,but when applied to polymers,they can provide polymer repeating units based on property requirements,which is not sufficient to provide sufficient information for the following polymer synthesis and production,and since the existing polymer design methods only have simple property constraints that are insufficient to describe the actual scenario requirements,an improved framework for computer-aided homopolymer design methods that takes into account the actual polymer attributes and product structure information requirements is proposed based on the property prediction model from(1).Firstly,by analyzing the homopolymer product attributes,the product properties are transformed into phase state requirements,molecular weight requirements,working temperature requirements and multiple property requirements,and then the homopolymer properties are divided into four levels of sequential calculations to avoid the computational complexity caused by simultaneous calculations of multiple properties and the over-constraint caused by blind requirements of multiple properties.The improved property prediction model is used as the core of the homopolymer design method to design homopolymer repeat unit structures and the required number-average molecular weight range according to the specific needs of the actual scenario,and the candidates can be ranked according to the synthesis difficulty,thus this new method is more practical than the original method and can provide more specific polymer structure information for downstream polymer synthesis and production.(3)Combining with the property prediction model of(1)and process design model,the single kettle Ziegler-Natta ethylene slurry polymerization process was studied,and the reaction process and structural-property relationships are modeled simultaneously,in order to optimize the reactor feed conditions and reactor volume for a given polyethylene property requirement.In the homopolymerization process,the reaction process is modeled based on the polymerization reaction kinetics,and the process modeling uses the method of moments to massively reduce the computational complexity due to the increase of polymerization degree,and the glass transition temperature and melting index are taken as an example in properties modeling to establish the correspondence between the properties and the number-average molecular weight and weight-average molecular weight.In the copolymerization process,the copolymer composition and sequence distribution are introduced to describe the structural specificity of ethylene/1-butene copolymers based on the homopolymerization model,and the copolymer property prediction method is improved to respond to the effects of different average molecular weights,copolymer compositions and sequence compositions on properties.The proposed method bridges the gap from product design to polymer production and provides a systematic product-process design idea for the development of high performance polyethylene products.(4)The integrated polyethylene product-process design approach from(3)was applied to the two-kettle slurry HDPE polymerization cascade process to address the problem of scaling up and increasing efficiency under given product quality in the actual production process.The process simulation was carried out in conjunction with the actual reaction process,and the process parameters and reaction kinetic parameters were calibrated based on literature data,and reasonable assumptions were made to establish a mathematical model for this process.In order to ensure the production of bimodal polyethylene with given requirements,the average molecular weight of the final bimodal polyethylene is calculated while the average molecular weight of the first kettle product is constrained at the same time,and finally the optimization results are re-simulated rigorously to avoid any deviation of the overall process calculation results due to model assumptions.Under the guidance of the objective function of maximum annual profit,the final optimized output was increased by 10.34% compared to the preoptimized output,while the total profit was increased by 16.58%,which proves that the method is meaningful for process optimization of the actual polyethylene production process and provides a solution to the scaling up problem after the new product is put into production.