Results and Discussion of Computer Simulations

Test Case with Constant Value ofkt 

The first set of simulations has been performed using a constant value of kt. In this specific case all radicals thus exhibit the exact same probability for termination irrespective of their chain length. Broadening processes that lead to a certain dispersity in the radical chain-length distribution can by definition not influence the kinetics of such a system. These simulations are therefore not meant to test equation 3.8 but rather to show the ease and 

The time during which primary initiator derived radicals are present can be approximated by t= 1/kp [AT], which would have resulted in an apparent (extrapolated) value of [i ]o of 3.710- mol L  

Once the radical concentration as a function of time has been determined, the final step in determining the termination rate coefficient can be taken. This involves the differentiation of this radical concentration versus time, according to equation 3.7 (the analytically determined second derivative of the cubic spline fit of [AT] versus time is used for this purpose). In doing so, figure 3.3 is obtained. As expected, the described calculation method yields values for kt which are identical to the input values used in the calculations. 

The discrepancy between input and output values is, however, somewhat greater than in the previous graphs. For intermediate chain lengths, the differences are smaller than 10 % but increases towards smaller chain lengths, approximately up to 110 to 210 %. It is difficult to point out the exact reason for this increased discrepancy towards smaller chain lengths, but it can most probably be attributed to the use of the cubic spline fit procedure. If the correct fitting function (equation 3.2) would be used to fit the monomer versus time profile over the smallest chain lengths, then the inaccuracy becomes much smaller. Hence, the cause can not be found in inaccurate numbers in the calculated monomer consumption trace. Nevertheless, 

---