A lot has changed in the transportation industry since the introduction of fuel and its use in internal combustion engines in the late 19th century. As environmental concerns have driven more stringent emissions regulations over the years, today’s diesel fuel needs to be cleaner than ever to protect critical engine components. This has led to the wide use of water-fuel separation filtration technologies in engine designs. Fuel filter water separators are typically used to remove contaminants in the form of water droplets and solid particles from the fuel before it enters the engine.
In a fuel-water separator, the separated water falls to the lower bowl of the filter assembly — the water bowl — and is drained at regular intervals. If the water isn’t drained effectively, it can be re-entrained with the fuel causing significant damage to the engine and its performance.
In this blog, we outline a case study of simulation models used to construct design tools that contribute to creating a good first prototype design for water bowls and drainage in water-fuel filters for combustion engines.
While this drainage mechanism in a fuel-water separator seems simple, the design is quite complex primarily due to three factors:
To best and most efficiently understand these factors, high-fidelity numerical simulations can be used. These simulations will lower the reliance on and requirement for experimental testing. They can be used to understand the physics of water drainage from inside a filter’s water bowl. Further, an analytical solution can be formulated to provide a design engineer with a very good first approximation of the most effective physical attributes that will assure proper water collection and drainage.
When water droplets are separated from fuel by a fuel-water separator, they coalesce and grow larger and heavier. Gravity causes the water droplets to settle and collect down as clean fuel is pulled in the opposite direction. The example shown below is our Fuel Filter / Water Separator – Racor GreenMAX™ Series.
The collected water is drained regularly to assure that it does not re-contaminate the cleaned fuel. If the water collected in the water bowl does not drain due to clogging or other malfunction, the separated water will re-contaminate the cleaned fuel and the performance of the fuel-water separator will dramatically diminish.
An increase in water content in the fuel supplied to the engine can:
Parker Hannifin specializes in fuel-water separator solutions for a variety of applications from marine to gasoline engines. Parker’s engineers found that given the varied geometries and materials used in their solutions, a simplistic physics-based tool to understand critical design features that prevent effective drainage of water from the water bowl is beneficial. The purpose of a study using this tool will provide initial design guidelines to a design- engineer even before a prototype is made for testing.
From the base geometry of the water bowl, the internal fluid volume was extracted. Key geometric variables were refined, and the internal volume extraction was automated so that several different geometric variations could be simulated in the design space. The extracted internal volume was discretized for computational fluid dynamics (CFD) using unstructured polyhedral elements with finer resolutions where water curvature is of significance (for example, when geometry gaps are small). The cell aspect ratios and skewness factors were the key quality factors monitored to estimate meshing quality. Multiple levels of refinement were considered, and results were monitored.
The workflow used for the analysis is shown below.
A 3D transient, multiphase (volume-of-fluid) method was used to model the two-phase flow. To introduce water drops in the fuel domain, periodically, drops of specified volume were introduced into the water bowl. The estimate of water sizes came from experimental observations. The contact angle for water-fuel-surface interface was also specified from experimental measurements for all the different surfaces with which the water comes in contact.
In this application, the interface curvature was very critical, therefore a continuum surface stress (CSS) model was chosen to model the surface tension force.
The local pressure drop at key geometry points was monitored to indicate the event of water clogging the water bowl. The simulation was run for a pre-set time and the pressure and movement of water drops were observed to see if the clogging event was registered.
From the CFD analysis, key geometric features that result in clogging were identified. Analysis of these results paved the way to develop physics-based analytical expressions to predict obstruction of water in a water bowl design. Using the guideline of the analytical expression, experimental testing was performed on some known water bowl designs. The comparison between the analytical expression, CFD results and the experimental observation is shown below. The plot shows that the CFD observations of water buildup with respect to critical geometric gap as well as the experimental observations are within ± 5% of the analytical expression derived. This validates our physical understanding derived from the CFD analysis as well as the analytical expression simplified for this problem.
Equipped with this knowledge, Parker engineers created a simplistic design guide. This is highly valuable to a design engineer as it avoids time consuming CFD and/or experimental prototype analysis and provides a direct pathway to constructing a good first prototype. A sample design guide shown below provides a visual of the simplicity of use. The design engineer would input design parameters and the design guide would provide inputs on issues with geometric parameters.
Engine manufacturers rely heavily on creating prototypes and testing concepts to find appropriate design solutions. Engineers can use high-fidelity numerical simulations to provide quick guidelines and resolutions to water clogging issues they might face in their testing. Apart from that, since this tool provides design guidelines as well, the first prototype built with it is often more accurate than without it. This lowers testing requirements and prototype iterations by a significant number thus helping the engineer get to the final design faster and more effectively.
This post was contributed by Sucharitha Rajendran, advanced system design and modeling engineer, Parker Filtration Innovation Center.