Using Ansys Fluent Discrete Phase Model Tools to Develop a Standard Erosion Analysis for Turbomachinery

Back to All

Using Ansys Fluent Discrete Phase Model Tools to Develop a Standard Erosion Analysis for Turbomachinery

DATE:

November 7, 2022

BY:

The American Hydro Team

Share

Using Ansys Fluent Discrete Phase Model Tools to Develop a Standard Erosion Analysis for Turbomachinery

NHA thanks our sponsors:

Kleinschmidt

Sponsored Content

American Hydro (AH) was contracted by one of our customers to supply new wicket gates and complete an erosion-focused computational fluid dynamics (CFD) study to analyze the particle behavior in the unit. The purpose of the erosion study was to provide a comparative analysis of the new wicket gate design vs. the original equipment manufacturer (OEM) gate to ensure the structural improvements AH suggested did not introduce excessive erosion damage. During this process, AH experimented with different 3D model assemblies, input criteria, and varying natural factors to yield a model that most closely resembled reality.

Erosion has proven to be an issue in the hydro industry, especially in areas where high concentration of suspended sand/quartz exists in the water. The repeated impact of these particles can lead to costly maintenance such as frequent component rehabilitation or replacement. In some cases, erosion can cause material failure which increases repair costs and safety risks. Using Fluent’s erosion analysis tools, the locations of erosion damage can be predicted and operating points that lead to damage can be avoided.

Over the last several decades many empirical and general erosion models have been developed to determine the actual erosion rate in products such as pipes, valves, nozzles, and more. Turbomachinery can be more complicated considering multiple wear mechanisms are involved that require specific erosion models to accurately predict erosion rate. Depending on the sediment present, erosion can commonly be found along the pressure side trailing edge or the inlet edge of hydro components. However, every unit is built and operated differently and can be exposed to varying degrees of abrasive material and erosion mechanisms. Erosion analysis is at the forefront of development and is not yet a completely understood phenomenon in the hydro industry.

As standard practice, AH modeled the entire turbine at prototype scale to reduce assumptions and increase accuracy of the fluid’s behavior. Besides the wicket gates, this included all primary water-passages including the spiral case, runner, and draft tube for all cases. AH had also modeled the turbine with and without the penstock to observe its effect on particle behavior. The model of the entire unit including penstock can be found in Figure 1.

Figure 1: Model of entire turbine assembly including penstock

Ansys Fluent Discrete Phase Model (DPM) was used to analyze the erosion effects after a fully converged turbulent k-ε model was resolved. The model worked by using the Euler-Lagrange approach to track particles individually where they were acted upon by forces due to drag, buoyancy, inertia, and surface pressure variation. The particle injection was applied to the specified inlet where particle streams are released from each element and scaled accordingly to ensure a uniform distribution.

In an ideal situation, a water sample would have provided AH with qualities of the abrasive medium such as particle sizes and mass fraction. Due to the absence of a sample, AH experimented with different particle sizes to observe their roles in erosion damage. Larger particles showed to have higher collision rates while small particles were more influenced by the flow field. This phenomenon can be seen in Figures 2.1 and 2.2. The number of particles released was based on the aforementioned parameters and mass flow rate.

In addition, gravity was enabled and disabled to compare its effects on particle distribution. It was found that gravity had different effects on particles in the penstock based on their size. While adding the penstock seemed to be more realistic, gravity caused larger particles to settle to the bottom and terminate before reaching the spiral case. This may have been caused by particles colliding with the no-slip boundary condition at the walls which effectively reduced the particle velocity to zero.

Figure 2.1: Tracking of small particles at 5% wicket gate opening

Figure 2.2: Tracking of large particles at 5% wicket gate opening

One of the many observations made during this analysis was that lower gate angles showed to have the most potential erosion damage towards the bottom of the pressure side trailing edge. Meanwhile, higher gate angles showed to have more potential erosion damage along the entire length of the nose. The erosion rate comparison between the OEM gate and new design at small and large wicket gate openings can be found in Figures 3.1 – 4.2. The patterns observed were directly related to the fluid velocity and gravity’s effect on the particles during different operating conditions. Our customer has shared that they commonly find varying degrees of erosion damage at the wicket gate trailing edges across several of their units. This statement was consistent with the results found during lower gate openings.

For this analysis, AH defined the upper fillet as the location of interest since it was the only modification made to the existing gate that had potential to influence the flow field and erosion characteristics. Ultimately, it was found that the upper fillet did not have any influence on the existing erosion pattern in the unit. Throughout this process, AH learned a lot about the different DPM parameters involved in erosion analysis and was able to solidify an efficient procedure for a comparative erosion study. AH found that this procedure can aid with the turbine design process, the selection of unit operating points, and the predictability of erosion in an existing system.