# Do Flow Splitters in Coriolis Flowmeters Really Work?

November 10, 2011

By Jesse Yoder, PhD, Flow Research, Inc.

When they were first made, many Coriolis flowmeters consisted of a single tube. Before long it was discovered that Coriolis flowmeters work better with two tubes. A “flow splitter” is used to divide the flow in two, and it flows through two parallel tubes.

Two fathers of Coriolis flowmeters, Jim Smith and Don Cage, recognized in one of their original patents the possibility that the tubes might not split the flow exactly in half. They make the following comment in their January 1, 1985 patent:

The accuracy in making fluid mass flow rate measurements with such a parallel path flow meter will be dependent on both the accuracy with which fluid flow is evenly divided between the two U-shaped flow tubes, and the equivalence of the angular velocities with which the two U-shaped flow tubes are sinusoidally driven. For, if a higher mass flow rate is passing through one U-shaped flow tube than the other, the U-shaped flow tube conducting the higher mass flow rate will generate larger Coriolis force effects.

Later in the paragraph they say that in one example this would have only a one percent effect on uncertainty.

I have seen the splitter in Coriolis meters and I have always wondered what guarantees that the flow is evenly split. Yet I am always told that yes, the flow is always split evenly between the two tubes. Has anyone tested or proved this?

Here are several factors that could potentially cause the split to be uneven:

1. Flow profile. We all know that flow is fastest thru the center of the pipe. But is this perfectly symmetrical? What if the side of one pipe has some buildup near the point it feeds into the splitter. Couldn''t this cause the fastest point to be slightly off center and thus to favor one tube over the other?

2. What if the machining on the splitter is slightly off? Couldn''t this impact the even splitting of the fluid?

3. What if the machining of the two tubes isn''t exactly identical? Couldn’t this cause the mass in one tube to be greater than another?

4. Could the presence of particles or entrained air in the fluid have an impact on the mass in the two tubes if they are unevenly distributed?

Perhaps these issues could be taken care of in the calibration of the meter. But I think this is an issue that is worth addressing for anyone who is interested in the accuracy of Coriolis flowmeters.