Unsteady flow temperature is one of the most difficult and complex flow parameters to measure. Main problems concern insufficient dynamic properties of applied sensors and an interpretation of recorded signals, composed of static and dynamic temperatures. An attempt is made to solve these two problems in the case of measurements conducted in a pulsating flow of gas in the 0–200 Hz range of frequencies, which corresponds to real conditions found in exhaust pipes of modern diesel engines. As far as sensor dynamics is concerned, an analysis of requirements related to the thermometer was made, showing that there was no possibility of assuring such a high frequency band within existing solutions. Therefore, a method of double-channel correction of sensor dynamics was proposed and experimentally tested. The results correspond well with the calculations made by means of the proposed model of sensor dynamics. In the case of interpretation of the measured temperature signal, a method for distinguishing its two components was proposed. This decomposition considerably helps with a correct interpretation of unsteady flow phenomena in pipes.

Pulsating flow is typical of displacement machines such as piston engines, compressors, or other displacement devices. The cyclic movement of the working element (piston, valve, membrane, etc.) generates pulsations, which, from a dynamic point of view, can be considered as excitation in the system composed of a flowing gas and pipes in which the flow takes place.This study shows that under certain conditions, a significant increase of pulse amplitudes occurs and transient values of main flow parameters are up to a few times higher than that for the equivalent steady flow conditions. This unsteadiness can be harmful or even damaging; however, in some cases it can be profitable (e.g. induction ramming of combustion engines). In order to investigate which conditions are favourable to obtain such effects in pipes, four different systems, distinguished by a different level of choking at the pipe outlet section, were tested.The experimental tests performed in the range of excitation frequencies of 0-200 Hz have shown that systems with low choking at the pipe outlet are more prone to strong dynamic effects, characterized by considerable growth of pulse amplitudes and an intensive reverse flow occurring in the proximity of resonance frequencies.In systems with high choking at the pipe outlet section, those effects were very weak. Characteristic frequencies, for which the resonance takes place, depend on the pipe length and can be determined theoretically with a simple acoustic model or a one-dimensional model 'x-t' describing a pulsating flow in the pipe. The experimental verification of those resonance frequencies can be conducted with the use of transient flow parameters recorded by appropriate fast response sensors.

No abstract