Polymer property predictions

D. Case IV Integrating the Concepts of Graphical Theory and ANNs for Polymer Property Predictions in QSPR... 

Two very simple types of QSPR have been developed early on in the evolution of polymer property prediction, namely van Krevelen s group contribution methods [122] and Bicerano s system [123], which mainly relies on the use of topological descriptors. Group contributions regard the overall properties of the polymer as the scalar sum of the properties of the chemical groups contained in the molecules making up the polymer. 

There has been therefore a considerable interest in determining the factors which control the Tm-value. One would hence also expect an extended amount of im-value/chemical structure correlations proposed in literature, just as the many correlations proposed for the Tg-value, see 7.2. It is remarkable however, that this is not the case. Van Krevelen [1] seems to be the only one with a clear Tm-value/chemical structure correlation concept. Even Bicerano [2], who is reporting such an extended polymer properties prediction system, does not mention the Tta-value at all. 

At present in world scientific laboratories a new polymers large amount is synthesized from which small part only reaches industrial production stage [47], It is naturally, that this work requires much time and means expenditure. These expenditures can be decreased essentially by new polymers properties prediction techniques development, proceeding from their chemical constitution. 

Polymer property prediction, large-systems modeling (mesoscale simulation), nanostructures, corrosion... 

Case, F.H. Applications of modeUng in polymer-property prediction,/ Computer-Aided Mater. Design, 3,369-378 (1996). 

The rotational isomeric state (RIS) model assumes that conformational angles can take only certain values. It can be used to generate trial conformations, for which energies can be computed using molecular mechanics. This assumption is physically reasonable while allowing statistical averages to be computed easily. This model is used to derive simple analytic equations that predict polymer properties based on a few values, such as the preferred angle... 

The property to be predicted must be considered when choosing the method for simulating a polymer. Properties can be broadly assigned into one of two categories material properties, primarily a function of the nature of the polymer chain itself, or specimen properties, primarily due to the size, shape, and phase... 

The following sections discuss the prediction of a selection of polymer properties. This listing is by no means comprehensive. The sources listed at the end of this chapter provide a much more thorough treatment. 

J. Bicerano, Prediction of Polymer Properties Marcel Dekker, New York (1996). Polymeric Systems, Adv. Chem. Phys. vol 94 (1996). 

CHEOPS (we tested Version 3.0.1) is a program for predicting polymer properties. It consists of two programs The analysis program allows the user to draw the repeat unit structure and will then compute a whole list of properties the synthesis program allows the user to specify a class of polymers and desired properties and will then try the various permutations of the functional groups to find ones that fit the requirements. On a Pentium Pro 200 system, the analysis computations were essentially instantaneous and the synthesis computations could take up to a few minutes. There was no automated way to transfer information between the two programs. 

Artificial Neural Networks as a Semi-Empirical Modeling Tool for Physical Property Predictions in Polymer Science... 

Recently, a new approach called artificial neural networks (ANNs) is assisting engineers and scientists in their assessment of fuzzy information, Polymer scientists often face a situation where the rules governing the particular system are unknown or difficult to use. It also frequently becomes an arduous task to develop functional forms/empirical equations to describe a phenomena. Most of these complexities can be overcome with an ANN approach because of its ability to build an internal model based solely on the exposure in a training environment. Fault tolerance of ANNs has been found to be very advantageous in physical property predictions of polymers. This chapter presents a few such cases where the authors have successfully implemented an ANN-based approach for purpose of empirical modeling. These are not exhaustive by any means. 

These capabilities of ANNs make them a unique tool for a large number of industrial applications. In this chapter, the authors demonstrate, with case studies, the advantages of using this approach to physical property predictions in polymer science. 

Physical property prediction in polymer science has evolved from the original basic group contribution meth-... 

R. Keshavaraj, R. W. Tock, R. S. Narayan, and R. A. Bartsch, Fluid Property Prediction of Siloxanes with the Aid of Artificial Neural Nets, Polymer-Plastics Technology and Engineering, i5(6) 971-982 ( 996). 

Bicerano J (1993) Prediction of polymer properties. Marcel Dekker, Inc, p 117... 

Thus quantitative analysis of elasticity is currently elusive, despite a great deal of work ongoing in this area. The extensive literature available on rubber elasticity by and large has not been adapted to hydrogels, and further work along these lines is necessary before quantitative predictions of swelling degree can be made from independent measurable polymer properties. 

Models for emulsion polymerization reactors vary greatly in their complexity. The level of sophistication needed depends upon the intended use of the model. One could distinguish between two levels of complexity. The first type of model simply involves reactor material and energy balances, and is used to predict the temperature, pressure and monomer concentrations in the reactor. Second level models cannot only predict the above quantities but also polymer properties such as particle size, molecular weight distribution (MWD) and branching frequency. In latex reactor systems, the level one balances are strongly coupled with the particle population balances, thereby making approximate level one models of limited value (1). 

J. Bicerano, Prediction of Polymer Properties, New York, Marcel Dekker, 1993. 

The use of computational methods for the calculation of molecular properties has been a perennial goal of chemists. In recent years, the field of computational chemistry has become a firmly established discipline. Computational chemists have made impressive contributions to almost every aspect of chemistry, ranging from structural organic and inorganic chemistry to the prediction of polymer properties and the design of medicinally important therapeutic agents. While many computer-based methods are robust and widely utilized, the continued development and refinement of software and the underlying theory remains an active area of research.1,2... 

Cartwright, Sztandera and Chu50 have also used the combination of a neural network with a GA to study polymers, using the neural network to infer relationships between the structure of a polymer and polymer properties and the genetic algorithm to predict new promising polymer structures whose properties can be predicted by the network. 

Bicerano J (1996) Prediction of polymer properties Marcel Dekker... 

---