Kieth1976,
Generally, accuracy is required to be compliant with the domestic water meter regulations and there are a several accuracy classes. Piston and disc meters usually being the more accurate as positive displacement meters compared to inferential meters such as the single and multi-jet meters.
SO the question is, why is a better meter accuracy required for a domestic meter?
Accuracy for my design is speculative as it must be until it can be verified empirically.
The best I can say is that understanding the factors that contribute to accuracy both initially and over time, and by comparing the designs, I know that my design potentially is better than either the nutating disc or the rotary piston meters and it may be that that could be traded into better accuracy or better long term life or any other feature that is prioritised. (PD meters are obviously capable of much higher accuracies than water meters e.g. for fiscal metering but the question is how is it to be achieved and at what cost?).
It is not simply a case of looking at the geometry and tolerances, if it were then there would be no significant theoretical difference between disc and piston meters. In reality a range of factors must be considered including accuracy over time.
Accuracy changes over time and use... some meters tend to wear in and others to wear out. Rotary piston meters tend to have better wear characteristics than disc meters.
If we had an ideal meter then the meter would be perfectly accurate over an unlimited flow range.
However, they must have working tolerances which means that some fluid slips through the tolerances unregistered.
The higher the pressure drop, the greater the slip flow.
Nutating disc and rotary piston meters have different mechanisms and the effects of wear are different.
For example, as a disc meter wears the change is progressive but in a rotary piston meter, as it initially wears, there some compensation - the loading on the guide roller increases with higher flow rates for a nutating disc meter but decreases for a rotary piston meter - the guide roller is necessary to control the tolerances and is more necessary at low flows in a piston meter and at high flows in a disc meter. In the piston meter, at high flows centrifugal forces take over but for the disc meter here is no radial freedom of movement due to the hub bearing but there is a tipping effect not so evident in the piston meter.
Wear is also influenced by peak pressure drop during the meter cycle and that is influenced in part by port design.
The ultimate best design depends on tolerances and materials and these have an evolutionary component to them:
The Kent "bomb" meter was one of the first to exploit a new manufacturing method.
Unlike industrial meters, there is no calibration adjustment, the calibration is determined by the meter's swept volume and the standard reduction gear train in the register assembly.
Until that time PD meters were assembled and components
fitted by skilled craftsmen and were thus very labour intensive.
What Kent did was to simply assemble the meters with as manufactured components with no skilled
fitting and test them.
In the Kent bomb meter, any meters that failed the calibration test were were stripped down and all the parts marked with a spot.
If, when the failed meter is disassembled a part already has a spot signifying that it has previously been found in a failed meter, it is given a second spot.
If a part was found with two spots, then the engineers would examine that batch of parts and look for deviations from the mean manufacturing tolerances etc.
In this way they not only ensured they brought together all the optimum components and identified any problem component batches they were also able to progressively optimise the component specifications.
What I mean to say is, there are a lot of factors to consider and also one has to consider the accuracy under test has to represent the performance of the meter over a long period of time.
My meter design exploits different design features to optimise performance by reducing slip flow generally, reducing peak pressure drops and hence also reducing wear and log term wear effects.
An easy question to answer is about dirt handling.
It will be seen that the flow path around the nutating disc meter is more favourable to dirt handling partly because the ports are in the outer chamber surface and dirt is "centrifuged" out of the chamber and partly because the disc edge sweeps the chamber surface with little opportunity to trap solids in the working tolerances.
In a rotary piston meter, on the other hand, the ports are in the top and bottom plates of the chamber and there dirt would tend to flow over the chamber wall surfaces where the piston skirt "rolls" or slides in a manner far more likely to trap the dirt in the tolerances.
Also, the dirt has to move radially inward to reach the chamber ports due to the port location.
Usually the problem is from larger particles which most frequently are present during the initial start up.
The nutating disc design is less vulnerable to such particles which are usually voided quite quickly and only rarely expected to cause a problem.
One solution would be to fit a filter but while a coarse filter is OK to screen out the few larger particles, a finer screen would trap finer particles in greater quantities and require cleaning i.e. labour. Modern rotary piston meters include slots scored in the piston skirt outer surface to allow particles to be swept through from the inlet to outlet by being carried in the slots rather than trapped in the clearance(the nutating disc meter clearance does not significantly change and the edge sweeps the chamber wall, the rotary piston tends to move radially outward at higher flows closing the tolerances which, while a benefit for long term accuracy has a negative in that it increases the vulnerability to dirt particles between the piston outer surface and the chamber wall).
So my design exploits the positive features of both meters to obtain good wear factors, good dirt handling and good long term accuracy.
Undoubtedly, with very precise manufacturing controls even the domestic water meter could have better instantaneous accuracy but at what price? Long term performance? higher reject rates during manufacture? ore dirt handling vulnerability?
The estimated cost to put into production is based on more complex component geometry (though no more or less components than the nutating disc or piston meters) which may require more time and more tool iterations to get right. However, the figures were given to me by a major global water meter manufacturer. It might be therefore that they:
a) multiplied the costs by the retooling not for one factory but several
b) overestimate because they don't want to obsolete their current designs... and they could probably be sure of similar thinking with other manufacturers.
But I would tend to agree that manufacturing costs would be better in Indonesia or Sri Lanka and I would guess that if you were thinking of a single factory probably not nearly so great to tool up either.
JMW