Measurement Delays and Modal Analysis for a Heavy Duty Transportable Emissions Testing Laboratory 950218

Concern over atmospheric pollution has led to the development of testing procedures to evaluate the hydrocarbon, carbon dioxide, carbon monoxide and oxides of nitrogen emissions from internal combustion engines. In order to perform emissions testing on vehicles, a chassis dynamometer capable of simulating expected driving conditions must be employed. West Virginia University has developed a Heavy Duty Transportable Emissions Testing Laboratory to perform chassis testing on trucks and buses. Emissions from the vehicle are monitored and recorded over the duration of a testing schedule. Usually the vehicle emissions from the whole test are reported as mass of emissions per unit distance driven. However, there is interest in relating the instantaneous emissions to the immediate conditions at specific points in the test, and in determining the emissions for discrete segments of the test (modal analysis). Unfortunately, there is significant delay between the point in time when the vehicle experiences a certain operating condition, and the point in time at which the emissions related to that operating condition are measured and recorded. This measurement delay results from a combination of the time due to the transport of exhaust gases from the engine to the appropriate gas analyzers and the response time of the analyzers. During emissions testing, the vehicle exhaust is connected via insulated pipe to the intake of a full size dilution tunnel where fresh air is mixed with the exhaust stream to simulate atmospheric conditions. Sample probes withdraw diluted exhaust gas from the tunnel and this gas is then transported to the gas analyzers via heated lines.

Each of these components leads to measurement delays, and the sum of these delays should reflect the shift between expected and measured emissions transients. A mathematical model was developed using analyzer response times and appropriate fluid flow and heat transfer equations to predict the time delay between an engine transient and the resulting emissions response. Actual time delays were determined by cross correlating engine power with each measured emission. The time delays determined using this process on data from heavy duty vehicles are reported. Experimental and theoretical time delays were found to agree. Typical emission delays for a transient bus test for hydrocarbons, carbon dioxide, carbon monoxide and nitric oxides were found to be 6.2, 12.4, 11.4 and 18.7 seconds respectively. Emissions data were compared piecewise with both speed-time and power-time schedules for the Central Business District transit bus test and the West Virginia University 5 Peak Truck Cycle.