In-cylinder pressure measurement for automotive applications
Accurate measurements of in-cylinder pressure are critical when developing automotive engines with improved exhaust emissions and efficiencies. Huge challenges are being faced by the industry to meet new EU emission regulations being phased in over the next few years.
Internal combustion engines provide an environment with extreme operating conditions for reliable sensor operation. Engine development requires high frequency and amplitude measurements – for example, 'knocking' produces in-cylinder pressure fluctuations with a period of the order of 0.05 ms.
Various pressure measurements are made during engine development to optimise fuel efficiency and to reduce toxic emissions. Knowledge of the response of a sensor and its associated equipment, including the processing electronics, connectors, and pipework, is needed throughout the range of frequencies it will be subjected to. Sensors can be supplied by the manufacturer with calibration certificates but these numbers are usually only determined for a gradual pressure change – the sensor performance is unlikely to be evaluated under the environmental conditions encountered in an engine.
The engine developers need to be confident that sensors:
- are reproducible and consistent
- provide accurate measurements of peak pressures
- have, when connected with pipe-work and processing electronics, a flat (or known) response up to frequencies significantly higher than the encountered frequency (which can be up to 30 kHz)
A traceable dynamic calibration facility for these sensors would be highly beneficial to engine developers, enabling them to improve engine performance in terms of both efficiency and emissions.
The project has two major aims:
- to obtain dynamic pressure traceability for high quality transducers using a shock tube facility and associated modelling
- to develop apparatus in which industrial sensors can be calibrated by comparison against these high quality transducers, using dynamic waveforms of representative frequency and amplitude
Dynamic traceability via shock tube
Shock tubes are capable of generating very fast step changes in pressure – and applying such a pressure step to a sensor may provide a good way of determining its dynamic response over a wide frequency range. The magnitude of the generated pressure step is a function of the properties of the gases used and the starting static pressures and temperatures either side of the diaphragm (which is burst to initiate the shock wave).
However, the 1‑D physical model used to calculate the generated pressure values makes certain assumptions about the system, such as instantaneous and complete burst of the diaphragm, which will never be totally valid, resulting in errors in the calculated shock front intensity. A better estimate of the pressure step can be obtained if the velocity of the shock front is known.
Cranfield University own and maintain a hypersonic gun tunnel which can be used to generate shock waves of the required magnitude. NPL is presently having an adaptor manufactured – this will attach to the end of the tube and hold four pressure transducers (lent to NPL by Kistler for this project). Three of these, located along the tube wall, will be used to measure the speed of the shock front and the fourth, located flush with the end face, will be the transducer under test.
Synchronisation of the readings from the three pressure transducers is important in estimating the shock velocity, so bespoke software has been written to be used with the data acquisition system, based on two National Instruments PXI‑5922 digitizers.
The static sensitivity of the transducer under test will also be determined in a separate calibration and this will be compared with the value obtained during the period after the shock front has passed, when the theory suggests that a steady‑state pressure should be present.
Sensor comparison apparatus
The requirements for the pressure generator to be used to compare working transducers with those characterised in the gun tunnel are still being developed. Ideally, it should be capable of generating 10 MPa amplitude dynamic pressures at a frequency of 30 kHz, but this might not be feasible in a laboratory environment – we are carrying out some modelling work to determine the optimum combination of equipment geometry, pressure medium, dynamic content, and waveform amplitude.
Please note that the information will not be divulged to third parties, or used without your permission