There are several methods: Gauss-Jacobi, Chebyshev ...? How to do it in practice in the manufacture of multi-channel ultrasonic flow meter?
I guess this depends on your demand and the given flow pipes. Multi-channel is used for higher precision and for disturbed flow profile. Often Gauss-Jacobi will be used for flow correction, but it depends on the (fully developed) flow profile.
I found an china patent on this, perhaps that might give you some more impression about that theme.
http://www.google.com/patents/CN103074873A?cl=en
Hope that helps, best regards
Marc
Thank you for your reply and for the link to the patent!
In some ways I have dealt with the work of these scientists only in articles.
Yes, I meant the the flow in the pipes. In my research I encountered with the method of Gauss-Jacobi and OWICS (Optimal Weighted Integration for Circular Sections).
Basically I'm interested in the limits and possibilities of these methods, or they can be modified and how to take into account prior to the distribution of velocities.
Based on the weight function of Gauss-Jacobi and some resemblance in form to a power law distribution, I modified Gauss-Jacobi method. In this case, the coefficient k in the formula weight function Gauss-Jacobi in my case 0.2 or 0.3 (depending on the Reynolds number). And the method OWICS k equal to 0.6.
And so the question arises whether it is possible to factor k could vary from 0 to 1 for example?
Hi Vitaliy,
we built a 6 path flow meter for gas, but didn`t implement any of these weight functions. Our aim was to develop a metering system including transducer and electronics design as well as programming the DSP. So this is a platform for further developments. For liquid media we use single path measurement and the integration over ultrasonic path with a fully developed flow profile depending on the reynolds number. Range of k is usually between 0,7...0,9. In disturbed flow (behind valves, pumps, air as medium...) and therefore demand for multipath metering you have to use weight functions like Gauss-Jacobi, OWICS and so on.
Factor 1 can only be achieved at very high flow or very high reynold numbers and no roughness of you pipe, so that the velocity is still the same over the whole radius of your pipe. (High turbulent profile) But that´s only a theoretical value and will not be achieved in practice.
A factor less than 0,5 seems to point on reverse flow sections in your pipe. 0 should be a theoretical value, too.
So your question is interesting for me, too. Hope that there are some more specialists at RG with practical knowledge in that theme.
Best regards,
Marc
- Badges
- Science topic
- Similar topics
- Cell Biology
- Methods