Understanding and controlling fluid flow with design and computational analysis of a fluid diode.

Organization/Class: ME 123 Computational Engineering — Stanford University

Role: Engineer

Duration: Apr 2021 — May 2021

Problem Statement: Develop a fluidic diode with a diodicity of 2.

BACKGROUND & CONSTRAINTS.

Fluidic diodes are passive devices that utilize geometry to manipulate the properties of fluid flow from different directions, in which flow is favored more so in one direction. In its forward direction, fluid flows through the device easily with little resistance. In its reverse direction, resistance is high to reduce flow.

The objective of this assignment was to design and analyze a fluidic diode with a minimum diodicity of 2. The dimensions of the fluidic diode are constrained to an outer rectangular box with a length of 0.2 m, a height of 40 mm, and a width of 20 mm. The inlet and outlet of the fluidic diode have a length of 10 mm, a height of 20 mm, and a width of 20 mm.

When running the simulations in Ansys, the assumptions are the following: the fluid medium is liquid, gravity is excluded, the fluid is flowing through a non-slip wall, the flow inlet has a mass flow rate of 1kg/s, and the flow outlet has a pressure of 0 Pa.

PROCESS.

When designing the initial fluid diode designs, I took inspiration from nature to think about what facilitates and impedes motion in the ocean. Thus, the outer outline of both my initial concepts resemble kelp. The first design A features two slanted cutouts that span the length of the diode and directs the flow to converge or diverge depending on the direction. In addition, there are curves on the edge that either encourage flow or mix streamlines to further add resistance to the fluid diode. This design is intended to split the flow into three streamlines along different types of resistance throughout the fluid diode. The second design B includes multiple shapes that are intended to be obstructions that create resistance in only one direction, shown in Figure 2. The outer outline of this concept also features curves like that of the first design. This design is intended to maintain one flow that occasionally diverges and either continues flow or mixes streamlines depending on the direction.

COMPUTATIONAL RESULTS.

The following are photos of the simulations run on Ansys Discovery Live for the different iterations of the two different fluid diode designs, A and B, from opposite directions each. An explore study was done for initial designs and the grid independence study to have more data points in a quicker time frame. The final designs were completed using a refine study.

ANALYSIS AND CONCLUSION.

For fluid diode A, the initial design had a diodicity of 1.16. These results were not what I was expecting, as the direction in which the flow was supposed to diverge had a lower pressure drop than that of the opposite. I think this may have had to shape and length of the internal cut outs, which caused there to be a greater pressure increase as the flow leaves where it diverged into three. From this design, I refined the edges to have more angled curves both on the outline of the diode and on the inside such that more resistance is created and to encourage mixing of the flow when going from left to right. This refined design had a diodicity of 2.33. The flow direction that encouraged mixing and turning of the fluid had a higher pressure drop, which was intended and expected. To establish confidence in my results, I ran a grid independence study where I ran the same conditions on different size grids from 0.59mm to 0.32mm. The results can be seen in Figure 10. The largest percent difference from the average of the pressure drop values was 14% at a grid size of 0.37mm.

            For fluid diode B, the initial design had a diodicity of 1.57. These results were not what I was expecting as the curved outline of the diode was not placed directly in the direction of the inlet, the flow was not inclined to curve along the edges and create the resistance intended. Thus, the diodicity was low as the fluid flow appeared similar between the two directions. This design was refined by bringing the curved edges closer together such that the path of the inlet flow will directly approach the edges and thus flow along them to either continue the path or curve to mix with the initial path. This refined design had a diodicity of 1.79. With the consecutive obstructions, it was much harder to anticipate how the fluid would interact with them, and thus, when refining this design, it was an iterative process to learn what works best for creating flow resistance. For this fluid diode design, I also ran a grid independence study covering the same fidelities as the previous design. The results can be seen in Figure 17. In this case, the largest percent difference from the average of the pressure drop values was 21% at 0.34mm fidelity.

            From these studies done on my own fluid diode designs, I developed a better understanding of important design considerations when trying to control fluids, such as the angles of any curves, the size and shape of the incoming flow, and the orientation of any obstructions. Working with Ansys Discovery Live on an independent project and doing the grid independence study was also helpful to better understand my results and establish how they might be applied and used.