A single vehicle (3.8L V6) underwent FTP75, HWFE, and US06 emission tests over a 6-month period. A bag mini-diluter and exhaust flowmeter sample system was installed in series with a CVS, so that mass emission results from an individual test could be directly compared between the bag mini-diluter and CVS. At one-month intervals, the bag mini-diluter and exhaust flowmeter sampling system was replaced with new units, while the vehicle and CVS remained unchanged. Assuming that the vehicle and CVS produce constant results, the variability of emissions over the test period are highly correlated with the variability of the bag mini-diluters and exhaust flowmeters.
The average CO2 mass comparison between the bag mini-diluter and CVS shows the separate sample systems match within 0.5% for an individual test. The established baseline determined from the CVS has a standard deviation of about 2%, which we believe is predominantly due to vehicle variability. A vehicle exhaust emissions simulator with accuracy of 0.5% is used to provide precise mass emissions for bag mini-diluter and CVS comparisons.
Correlations are also presented for hydrocarbons, carbon monoxide, and nitric oxides, with similar results. Other critical variables in BMD correlation, including water correction factors, and exhaust flowmeter volumes are presented.