Division (a)
United Kingdom
Molar based approach. To calibrate an SSV flow meter the following steps shall be performed:
- (i)
The Reynolds number, Re #, for each reference molar flow rate, shall be calculated using the throat diameter of the venturi, d t [equation (7-145)]. Because the dynamic viscosity, μ, is needed to compute Re# , a specific viscosity model may be used to determine μ for calibration gas (usually air), using good engineering judgment [equation (7-146)]. Alternatively, the Sutherland three-coefficient viscosity model may be used to approximate μ (see Table 7.7):
(7-145) Where:
d t M mixdiameter of the SSV throat [m]
ṅ refmixture molar mass [kg/mol]
reference molar flow rate [mol/s]
and, using the Sutherland three-coefficient viscosity model:
(7-146) Where:
μ μ 0Dynamic viscosity of calibration gas [kg /(m · s)]
SSutherland reference viscosity [kg /(m · s)]
T 0Sutherland constant [K]
T inSutherland reference temperature [K]
Absolute temperature at the venturi inlet [K]
Table 7.7 Sutherland three-coefficient viscosity model parameters Tabulated parameters only for the pure gases, as listed, shall be used. Parameters to calculate viscosities of gas mixtures shall not be combined.
Gasf1 μ 0 T0 S Temp range within ± 2 % error Pressure limit kg /(m · s) K K K kPa Air 1,716 × 10– 5 273 111 170 to 1 900 ≤ 1 800 CO2 1,370 × 10– 5 273 222 190 to 1 700 ≤ 3 600 H2O 1,12 × 10– 5 350 1,064 360 to 1 500 ≤ 10 000 O2 1,919 × 10– 5 273 139 190 to 2 000 ≤ 2 500 N2 1,663 × 10– 5 273 107 100 to 1 500 ≤ 1 600 - (ii)
An equation for C d versus Re# shall be created, using paired values of (Re# , C d). C d is calculated according to equation (7-140), with C f obtained from equation (7-141), or any mathematical expression may be used, including a polynomial or a power series. Equation (7-147) is an example of a commonly used mathematical expression for relating C d and Re# ;
(7-147) - (iii)
A least-squares regression analysis shall be performed to determine the best-fit coefficients to the equation and calculate the equation's regression statistics, the standard estimate error SEE and the coefficient of determination r 2, in accordance with Appendix 3;
- (iv)
If the equation meets the criteria of SEE < 0,5 % n ref max (or ṁ refmax) and r 2 ≥ 0,995, the equation may be used to determine C d for emission tests, as described in point 3.6.3(b);
- (v)
If the SEE and r 2 criteria are not met, good engineering judgment may be used to omit calibration data points to meet the regression statistics. At least seven calibration data points shall be used to meet the criteria;
- (vi)
If omitting points does not resolve outliers, corrective action shall be taken. For example, another mathematical expression for the C d versus Re# equation shall be selected, leaks are to be checked, or the calibration process has to be repeated. If the process shall be repeated, tighter tolerances shall be applied to measurements and more time for flows to stabilize shall be allowed;
- (vii)
Once the equation meets the regression criteria, the equation may be used only to determine flow rates that are within the range of the reference flow rates used to meet the C d versus Re# equation's regression criteria.
Annotations
Footnotes
- f1
-
Tabulated parameters only for the pure gases, as listed, shall be used. Parameters to calculate viscosities of gas mixtures shall not be combined.