Comparisons and choice of pressure loss coefficients, Ζ for ductwork components

Abstract

In compiling a revision to the CIBSE Guideon fluid flow in pipes and ducts, a survey was undertaken of available data on pressure loss factors of fittings. This paper details some of the contradictory information concerning components of ductwork. The primary source used by engineering institutions of several countries is that compiled by the Russian, Idelchik. His information is frequently far too detailed to be of use for day-to-day duct system design and so requires some interpretation and simplification. This the author has endeavoured to achieve whilst also using other sources where possible. Nevertheless this source was found to contain several errors. In the case of some of these errors, the true interpretation could readily be guessed at, but not always. Some other more recent research by others has also been found to contain contradictions. Even the use of a single preferred source does not guarantee uniformity of results. In one instance, two guides and the author provide three different sets of figures, each resulting from interpretations of the same primary source. In most published data there has been an assumption that friction effects of ductwork can always be presented in a non-dimensional manner. The implication is that size does not matter. Recent research shows that this is not so and confirms that as with pipework, size cannot be ignored for values of Ζ for elbows. Both size, and Reynolds Number are shown to have a large effect on the value of Ζ for elbows but neither have any significant effect on Ζ for tees. As to the question of variation in pressure loss due to minor manufacturing details, this has been tested only recently. Tests carried out on seemingly identical components but each obtained from a different manufacturer, showed a remarkable scatter. Although this was largely due to the results from the item supplied by just one manufacturer, this does illustrate the problem of predicting the pressure drop in any real ductwork system. This paper makes no attempt to include all the comparisons which were made. Rather, it highlights a few of the glaring contradictions and possible errors in hitherto accepted data. In trying to assemble data which is reliable and simple to use in day-to-day design work, the author has been presented with several dilemmas. The more contentious of these, and the final decisions taken, are explained.