Test cases--results

JakobCHR.com

Quick Navigation:

Personal:

Go to Home
MS Research
PhD Research
Curriculum Vitae

General:

Linux

Soon to come:

Matlab
On-line Stores
Cycling
Medicine & Health
LaTeX
OOP & C++
Sony PCM-R500 DAT

Next: Order test Up: Semi-analytical test case Previous: Test cases description Contents Index

Test cases--results

In Figure 9.1 and Figure 9.2 we have compared the total pressure loss, $-\Delta p_{\mbox{\protect\scriptsize tot}}$ , calculated by the Matlab and the C implementations. As we can see there is no visible difference between results calculated by the two implementations^9.4. We have also plotted the difference between the two results^9.5, ie $\Delta p_{\mbox{\protect\scriptsize tot,Matlab}} - \Delta p_{\mbox{\protect\scriptsize tot,C}}$ . The result of this action is illustrated in Figure 9.3 and Figure 9.4. We notice that the difference is indeed very small and below 1 Pa in all cases. The main source of error is the local truncation error. This claim is substantiated by the variation of the error with the mass flux in the two cases--the observations can be stated in the following way

In both cases the error is highest for lowest mass flux since we have the most rapid growth of both flow quality and void fraction.
In the low power case (Figure 9.3) the error is very low for $\mbox{$<\!{G}\!>$} = 1100 {\mbox{ kg}}/({\mbox{s}}\cdot{\mbox{m}}^2)$ since we in this case have zero void at the exit of the core. This means that we have no contribution to the local truncation error from the $<\!{\alpha}\!>$ and $<\!{x}\!>$ profiles.
In the high power case (Figure 9.4) the error in the first part decreases with $\mbox{$<\!{G}\!>$}$ due to decreased slopes of $<\!{\alpha}\!>$ and $<\!{x}\!>$ . In the last part of the figure ( $\mbox{$<\!{G}\!>$} \ge 750 {\mbox{ kg}}/({\mbox{s}}\cdot{\mbox{m}}^2)$ ) the error increases due to increased frictional pressure drop which results in steeper slope of p(z) as we can see in Figure 9.5.

In this investigation we have concentrated on checking the pressure. This is mainly caused by the fact that the pressure is the most complicated of the dependent variables to calculate. Furthermore, the pressure variation is strongly influenced by both the void fraction and the flow quality which means that if the pressure is correct so is the void fraction and flow quality. Just to exclude all doubt we have depicted the error on the void fraction for the high power case and $\mbox{$<\!{G}\!>$} = 750 {\mbox{ kg}}/({\mbox{s}}\cdot{\mbox{m}}^2)$ in Figure 9.6. As illustrated the error is below 10^-5.

$\begin{figure} % latex2html id marker 32908\rule{\textwidth}{0.2mm} \rule{0cm}... ...0 = 0.25/2.44 {\mbox{ MW}}/{\mbox{m}}$) as a function of mass flux.}\end{figure}$

$\begin{figure} % latex2html id marker 32953\rule{\textwidth}{0.2mm} \rule{0cm}... ... 2.5/2.44 {\mbox{ MW}}/{\mbox{m}}$) as a function of the mass flux.}\end{figure}$

$\begin{figure} % latex2html id marker 32977\rule{\textwidth}{0.2mm} \rule{0cm}... ... implementations for the low power case as a function of mass flux.}\end{figure}$

$\begin{figure} % latex2html id marker 32997\rule{\textwidth}{0.2mm} \rule{0cm}... ...implementations for the high power case as a function of mass flux.}\end{figure}$

$\begin{figure} % latex2html id marker 33048\rule{\textwidth}{0.2mm} \rule{0cm}... ...{ kg}}/({\mbox{s}}\cdot{\mbox{m}}^2)$\ (\raise0.5ex\hbox{$.....$}).}\end{figure}$

$\begin{figure} % latex2html id marker 33117\rule{\textwidth}{0.2mm} \rule{0cm}... ...e dashed curve (\mbox{$--$}) is a normalized void fraction profile.}\end{figure}$

Next: Order test Up: Semi-analytical test case Previous: Test cases description Contents Index

Top Quality
Developed with
Danish
Brain Power