OCR Text |
Show recalculation procedure changes the firing rates (and total exhaust enthalpies, Hg) almost marginally. The principal impact is on the specific exhaust enthalpies, hg, which can be seen from Table 1 to be decreasing with increasing firing rate or output in spite of the rising temperatures; this behavior is on account of the different levels of excess air in the different runs. As we now show, the correction to a common excess air level reverses this trend, and the specific exhaust enthalpies rise with output. Specific Exhaust Loss. Figures 3 and 4 illustrate the variation of the specific exhaust loss, in the ratio form, hglhr, for two separate methods of plotting. Figure 3 is a plot as a test of Eq. 9 which can be linearized in a reciprocal form using a subtraction parameter, KI : where: K. = (1 - lIt); for f = 1, KI = o. The procedure used was to plot the LHS term against II. with the value of K. determined as the value giving the highest R2 (error) value in the linear plot. Within reasonable limits of experimental error, Fig. 3 shows that the reciprocal (LHS) term is linearly proportional with negative slope to the output, HI, thus supporting Eq. 9. The values of the slope, intercept, and K, provided parameter values permitting calculation of the three required quantities: the limiting Intrinsic Efficiency, 0.00, the maximum output, II. m, and the parameter correction factor, f The values were, respectively: 0.21, 20.5, and 1.09. These values were then used in the backplot shown in Fig. 4 and in subsequent graphs; Fig. 4 also compares the revised values of hg, corrected to 25% excess air, with the original, experimental values of hg, obtained with variable excess air as described. As this plot shows, the original values decrease with output and the revised values increase with output. Wall Loss. This is illustrated in Fig. 5 with the (empirical) quadratic back-fitted curve included. A linear fit, it may be noted, gives a negative Idle loss, which is physically meaningless; this supported the fitting of a non-linear curve. The procedure to obtained the tits was similarly iterative, based on linearizing Eq. 10 by writing: [(Hw - HwO)IH,] and maximizing the linear tit to HI by adjustment of the Idle value, Hwo. The value 11 |