The Denali Flushometer is a unique flush valve in that it delivers a fixed volume of water independent of the position of the restriction on the rolling diaphragm that is randomly installed at various positions during the valve assembly. The innovative flush valve was designed by Johnson Design with the help of Flomerics’ EFD.Lab CAD-integrated computational fluid dynamics (CFD) software, which enabled engineers to solve design problems with software rather than hardware prototypes.

Johnson Design used EFD.Lab to evaluate the valve performance while the restriction was rotated 360 degrees. This helped to ensure that it delivered the right amount of water the valve each position. EFD.Lab considerably reduced the amount of time required to optimize the design by enabling us to evaluate design concepts in software without having to build a prototype.

The Denali valve operates by taking advantage of pressure differentials between inlet and control chambers. The pilot valve, when engaged by the handle, vents the control chamber, lowering its pressure and allowing the diaphragm to stroke up, which in turn begins the flush cycle. Water rushes from the inlet to the outlet of the valve during the flush cycle while the diaphragm is open. A small orifice through the diaphragm allows a small proportion of the water to flow back into the control chamber during the flush cycle. As the control chamber is filled and pressurized, the diaphragm strokes closed.  The flush cycle ends once the pressure in the control chamber reaches upstream line pressure and reseats the diaphragm.

The Denali flush valve uses a rolling diaphragm instead of a piston or a large area diaphragm. Rolling diaphragms offer the advantage of exposing only a very small area of unsupported diaphragm to pressure differentials, resulting in much lower forces compared to large-diameter, flat diaphragms. So the rolling diaphragms combine the best features of piston valves (small size, and long stroke) with the best features of large-area diaphragm valves (low friction and high sensitivity).

 

Denali’s design

But it’s usually not possible to determine in advance where the restriction on the rolling diaphragm will end after the valve is assembled. When the diaphragm is opened, the water rushes past the restriction through the diaphragm to the outlet of the valve with complicated flow patterns. Generally the pressure is much higher below the diaphragm on the side that is opposite to the inlet because the water drastically changes direction and jams up against the opposite wall. This means that if the orifice happens to end up on the inlet side, it will be exposed to relatively low pressures. The control chamber will then fill slowly and the flush volume will be high. As a result, a weakness of earlier diaphragm valve designs is that one valve of the same model may deliver 1.5 gal. while another will deliver 1.7 gal. For example, a competitive flush valve has shown a volume delivery variance of 13.4 percent in our tests.

Through the 1950s, most toilets used 7 gal. of water per flush. By the end of the 1960s, toilets were designed to flush with only 5.5 gal., and in the 1980s, new toilets were using only 3.5 gal. In order to conserve water, the National Energy Policy Act of 1995 mandated that the toilets use 1.6 gal. for each flush. The American Society of Mechanical Engineers/American National Standards Institute (ASME/ANSI) standards for 1.6-gal. toilets were adopted by reference, establishing them as the standards across the U.S. The ASME/ANSI standards require that toilets must flush with an average of 1.6 gal. on five test flushes, with none of the flushes exceeding 2.2 gal.

We decided to achieve a consistent flush volume in the Denali valve by evening out the pressure that the restriction through the diaphragm is exposed to around its entire circumference. This would have been a difficult task using physical prototyping because it would have been necessary to repeatedly assemble and disassemble the valve, each time slightly changing the position of the rolling diaphragm. We would have had no way to visualize the flow inside the valve, so if we found a problem in a certain position we would have to rely on intuition and guesswork to try and solve it. To evaluate each possible solution, we would have to build a new prototype valve. Instead, we decided to simulate the operation of the flush valve with CFD to see if we could even out the pressure differences inside the downstream chamber.

A CFD simulation provides fluid velocity, temperature, and chemical-concentration values throughout the solution domain for systems with complex geometries. As part of the analysis, a designer may change the geometry of the system or the boundary conditions and view the effect on fluid-flow patterns. But conventional CFD software is expensive and requires a lot of knowledge on the part of the users such as various meshing options, physical models and boundary conditions. Up to this point, these limitations have caused CFD software to be used mainly by analytical specialists relatively late in the development process, primarily to evaluate problems with existing prototypes or products.

But as CFD becomes more important to the design of a wide range of different products and processes, the limitations of the current use model have become more obvious. They have been addressed by a new generation of software that is fully embedded in the MCAD environment. This type of software works directly with the native MCAD geometry and requires no translation of data, so that the solid model maintains intelligence such as assembly hierarchy, constraints, and features. The software analyzes the geometry and generates the computational grid in the background, while the user only needs to interact with the familiar MCAD interface.

Developing the Denali valve

Large automotive and aerospace companies have experts with graduate degrees in fluid dynamics that have no difficulty using very complicated CFD software. Since we are much smaller company, we wanted a package that was intuitive enough to be used by a design engineer that does not have a Ph.D. We found two software packages that seemed to fit the bill. We built a model in our lab of a pipe with a restriction. We ran water through it and measured the flow rate and the upstream and downstream pressure. We then simulated the experiment with each of the two engineer-friendly CFD software packages. The first software package took us four hours to simulate the experiment and produced misleading results. With EFD.Lab, we simulated the problem in 30 minutes and our results were within a few percent of the correct answers.

The Denali valve was originally designed in Pro/ENGINEER computer-aided design (CAD) software. We imported the CAD model of the valve into EFD.Lab, which automatically distinguished between the solid and empty spaces in the CAD model and meshed the empty regions to prepare for flow analysis. We measured the flow rate and pressure at the inlet and outlet of a similar existing valve and used those as boundary conditions. The analysis results showed flow rate and pressure at every section of the valve.

The simulation of the initial design showed that, as expected, the pressure was considerably higher on the side of the downstream chamber opposite to the inlet. We modified the geometry of the downstream chamber to reduce the amount of flow directed to high-pressure areas and to increase the flow directed to low-pressure areas. Each time we modified the geometry, we ran the simulation again to determine the impact on the pressure variations inside the downstream chamber. On the sixth iteration, we produced a design that evened out the pressure in the downstream chamber. In the final design, the orifice can be positioned anywhere without affecting flush volume.

Next, we used the existing model of the flush valve to look at the variations in the flow rate over a single flush cycle. The goal was to maintain an even 28 gal. per minute (gpm) flow for as long as possible to clean out the basin then shift smoothly to a 5 gpm flow to reseal the diaphragm. When the diaphragm is fully open, a restriction in the lower end of the downstream chamber limits the flow rate. As the diaphragm closes, an upper restriction takes over and limits the peak flow rate. As the diaphragm continues to close, the upper restriction finally becomes so restrictive that the flow ends. We used CFD simulation to size the restrictions to achieve the desired flow rate.

EFD.Lab helped us optimize the design quickly. The ability to visualize the flow within the valve made it easy to see what was causing pressure variations. We made a few design changes and soon had equalized the pressure around the valve. When we built and tested a prototype valve, it worked exactly as we expected by providing consistent volume delivery, regardless of the position of the rolling diaphragm.

For more information, email: info@flomerics.com