Mostly 50 times the diameter gives good results in most applications where flow geometries change. Although different researchers take any value from 10 D to 100 D, depending how flow development length is defined.
Mostly 50 times the diameter gives good results in most applications where flow geometries change. Although different researchers take any value from 10 D to 100 D, depending how flow development length is defined.
This distance is a function of the Re number. There are several semi-empirical relations in the literature. These expressions are expressed in function of adimensional parameters like Re and L/D being D the diameter for pipes and L the distance from the inlet. These relations are obtained applying Pi theorem of buckingham and experiemental measurements. You could obtain L if you know the Re number and the diameter of the pipe. For non-circular sections pipes, you must use the hydraulic diameter.
As stressed by Miguel Coussirat the development length depends on the Reynolds number. For laminar flows it is inversely proportional to the Reynolds numbers. For turbulent flows (Re>2300) the distance needed depends on the definition of "fully developed". For many applications a distance of 20 diameters can be sufficient. A distance of 50 diameters s safe if there is no longitudinal swirl in the flow. Sch a swirl can occur if there are two successive orthogonal bends in the upstream system (or another asymmetry). In case it is very important to have a fully developed flow after a reasonably short distance, you should use a perforated plate. Specially designed perforated plates exist, which are placed just upstream of turbine flow meters or ultrasonic flow meters.
it is not necessary to find in a research paper to read about this. In a first technical approach, this problem is a handbook's solution. You can read about developed flow length in any classical Fluid Mechanic book, e.g. White, Çengels Cimbala, Fox, etc.
Firstly the distance of 5 to 10 times the diameter duct is not sufficient to reach fully developped flow. I think that Avraham Hirschberg proposition is very reasonnable and very practice.. I tested 20 to 30 times and it gives acceptable results
The 9 degrees is the opening angle of a conical expansion in which one tries to avoid flow separation (Blevins, Fluid Dynamics Handbook). This is actually only valid for high Reynolds number flows (turbulent flow). It is not related to the original question: how far downstream of an abrupt expansion (with flow separation) does one recovers a fully developed flow.
In the scientific references, a distance of 20-50 diameter is recommended for fully developing of flow after sudden expansions. For the open channels, the experimental data shows that using a hump decrease the flow separation and energy loss. In this case, hump with a crest height of 5-9% of the approach depth are recommended. for more information, you can refer to this valuable paper: