Roussillon (1974): measuring engine air flow using sonic nozzles

In this document

This offprint reproduces in the journal Ingénieurs de l'Automobile (No. 1-74) the summary of a Doctor Engineer thesis defended on 12 June 1972 by Guy Roussillon at the Orsay Fluid Mechanics Laboratory, under the chairmanship of A. Fortier p. 2.

The problem and the method

The author starts from an observation: conventional pressure-reducing devices (venturi, orifice plate) distort the measurement of pulsating air flows, such as those drawn by an engine, due to pressure fluctuations and an uncertain discharge coefficient in unsteady flow p. 2-3. He proposes an original method: interposing a buffer capacity fed by a sonic flow (via a sonic nozzle, here a butterfly valve) which remains insensitive to engine pulsations, allowing a reliable average flow to be measured without disturbing the actual pulsating regime p. 3.

Theory and tools

The text develops the equations for "average steady" flows, introduces the Hodgson number as a criterion for measurement precision (formula H, p. 4) and distinguishes between sinusoidal or interrupted flow cases p. 4-6. It establishes the conditions for the buffer capacity volume (V₀/d³) to limit error p. 6.

Test bench

The installation (Plate 46, fig. 6) combines an AFNOR 50×20 venturi, a Ø 21 mm butterfly valve serving as an adjustable sonic nozzle, and a 400-litre buffer capacity p. 6-9. The tests are conducted on the scavenging pump of a C.L.M. type 603 two-stroke Diesel engine (bore 75 mm, compression ratio 15.78/1) p. 10.

Results

The author concludes that the sonic nozzle allows for a significant reduction in the necessary buffer capacity and constitutes "an industrial precision flowmeter" with an adjustable nozzle, and "of excellent precision" with fixed sonic nozzles p. 15.

The document closes with a bibliography of 48 references p. 18-19 and questions and answers with the author, where he notably quantifies the hourly cost of compressed air compared to fuel on 100 hp Diesel and petrol engines p. 21.

Context

Guy Roussillon appears here as head of aerodynamics studies at the Paris Study Centre of Peugeot automobiles p. 2 — a useful reminder: this work, although originating from the academic world (CNAM, Paris VI University), is part of the industrial concerns of measurement and engine tuning of French manufacturers in the 1970s.

More broadly, the precise measurement of engine air flows was, at that time, a growing issue for electronic injection and the beginnings of the anti-pollution fight — subjects treated in other pieces in the same file (Bosch K-Jetronic, Chapoux/UTAC works), without this document explicitly dealing with them itself.

Archivist's perspective

This text is a rare piece in the Magès collection: it does not bear his signature, but illustrates the scientific environment in which French automotive engineers of the 1970s operated around the issues of carburetion and air intake — one of the major fields of interest documented in this file, as evidenced by the clippings on the Bosch K-Jetronic injection or the handwritten cover page on carburetion and injection.

It seems that Paul Magès, or his documentary circle, deemed it useful to keep this highly specialized methodological study — probably for its relevance in the development of engine test benches, a field directly linked to his own work on the intake and supply of Citroën engines. The presence of a handwritten reference annotation at the top of the first page p. 2 suggests careful filing within a larger thematic folder.

The interest of this piece lies less in its content directly linked to Magès' work than in what it reveals of the level of metrological requirement expected by automotive engineers of the time on subjects that were a priori secondary — air flow measurement — but crucial for engine performance and reliability.