Recovery of Tail Pipe Species Concentrations and Its Effect on Emissions Calculations from Raw Exhaust Gas Streams during Chassis Dynamometer Tests 2016-01-0984

This paper proposes a method to recover species concentrations at the tail pipe exit of heavy-duty vehicles during chassis dynamometer tests, and investigates its effect in the calculation of emissions from their raw exhaust streams. It was found that the method shown in this paper recovered the sharp peaks of the gas species. The effect on calculations was significant, as time-variant raw exhaust flow rate and emissions concentrations data are acquired continuously during a test (at 10 Hz), and their product is integrated during calculations. The response of the analyzer is delayed due to the time taken for transport of the sample gases from the probe tip to the analyzer, and deformed due to mixing and diffusion during this transport. This ‘convolution’ of the concentration data stream introduces an error in the final result, calculated in g/mile.

The convolution of the concentration data is corrected by the following method: the analyzer response (output) to a step change in the concentration of the species at the probe tip (input) is recorded to determine the ‘convolution function’. The inverse of the convolution function is applied to the output to recover the input concentrations at the probe tip. The signal processing is carried out in the frequency domain and appropriate filters and windowing are applied to improve the quality of the reconstructed signal. This deconvolution algorithm is then applied to the analyzer data collected during a test to recover the concentrations at the probe tip, which is thus corrected for time delay and signal deformation. The deconvoluted concentration data, exhaust flow rate data, and dynamometer data are used to calculate the emission of a gas species (g/mile).

Results for tests conducted in the laboratory for CO₂, CO, NO_x, NO, THC, and CH₄ are presented. The results obtained from the ‘asrecorded’ raw exhaust data and reconstructed data are compared with those from bag samples for the same test. It is shown that the emissions calculations using deconvoluted concentration data from the raw exhaust stream match well with the results from the bag samples for the same test in a full-scale dilution emissions measurement system.