FilterDict® is a powerful module for the simulation of particulate filtration in filter media and filter elements to complete filters (M2M Technical Report, 2020 Nr. 03). Additionally, predefined settings with user input allow the simulation of cross-flow filtration and the simulation of flow in the complete filter with housing.
FilterDict® is a pioneer software tool, widely used in academic and industrial settings, and the only software on the market providing this functionality on standard hardware (16GB 64bit Windows or Linux computers).
Recently, M2M has become active in simulating room air filtration systems and their optimal room placement to increase particle removal efficacy and possibly reduce the risk of viral infection (Short introductory video).
FilterDict® is a registered trademark of Math2Market GmbH
FilterDict® for filter media
The filter media can be any kind of porous media, typically nonvowen media, woven media (metal, textile, or plastic), foam, membrane (e.g. for sea water desalination), or cellulose paper. FilterDict® can also simulate filtration through soil or sand, and others.
Any of these filter media can be analyzed, optimized and/or newly designed (M2M Technical Report, 2020 Nr. 02)
With FilterDict®, filter efficiency computations requiring a single flow field computation can be achieved for realistic nonwoven models (300 x 300 x 600 computational cells) in a few hours. Filter life time simulations may take several days due to the need for re-computing the flow fields.
The prediction of filter efficiency and pressure drop evolution by FilterDict® has been extensively validated in a variety of industrial and scientific research projects. The predictions agree with measurements to within 10-20%.
FilterDict® determines for filter media:
Initial Filter Efficiency:
FilterDict® determines the filter efficiency and most penetrating particle size (MPPS) of a filter medium by tracking particles through a 3D filter structure. The simulation of the particle movement can include interception, inertial impaction, Brownian motion and electrostatic attraction. Various physical models are available to model the contact of a particle with the filter medium (Caught on first touch, Van der Waals forces / Hamaker constant, Sieving) and allow to model different filtration regimes. User-defined functions allow full control of these model parameters.
Initial Pressure Drop:
Using the flow solvers from FlowDict, FilterDict® determines the initial pressure drop over the filter media. Fast flow rates are modeled with the Navier-Stokes, slow flow rates with Stokes equation.
- Filter Life Time and Capacity:
FilterDict® can simulate the clogging of a filter in a Single-Pass or Multi-Pass experiment. Pressure drop evolution over time and the dust holding capacity is determined. The simulation covers the full range from depth to cake filtration.
Filter lifetime simulation in the air filter media structure
Simulation of filter lifetime in the oil filter media structure
Simulation of filter lifetime in a complex multilayer wowen media
Particle sieving through a two-layer sand control screen model
FilterDict® for pleated filters and complete filters with housing
The ideal filter has a large Dust Holding Capacity (DHC) and high filter efficiency while maintaining a low pressure drop. These three properties usually depend on each other, meaning that improving one or two of them will usually undermine the third property.
With the goal of increasing the DHC, FilterDict® simulates the clogging of a filter in a single-pass experiment on the unresolved media (such as single or multiple pleats, pleat elements and diesel particulate filter structures)
For filter elements such as pleated filters, FilterDict® determines filter life-time, pressure drop evolution and fractional efficiencies over time, and DHC.
The filter media of the pleats can be a nonwoven or woven filter media. The pleated filters can be optimized and/or newly designed by choosing flat sheet filter media and specifying the geometry and the number of pleats.
The implemented pass-through model considers the filtration possibility of each type of particle in different types of unresolved porous media.
The simulation covers the entire filtration regime from depth filtration to cake filtration. The simulation with GeoDict® is validated by comparing the results with experiments and adjusting the numerical parameters.
For complete filters with housing, the geometry of the parts of the complete filter in CAD format (*.stl and/or .obj) is imported and converted into a GeoDict® voxel (3d pixel) using ImportGeo-CAD. The user defines inlet and outlet for the flow and the characteristics of the filter parts (flowing fluid, solids, porous parts).
The characteristics of the filter media at micro scale are also defined (e.g., permeability).
The filter media of the pleats inside the complete filter can be a nonwoven or woven filter media.
The lifetime single-pass simulation determines the changes in pressure drop as a function of time, the fractional filtration efficiencies, and the deposited dust per batch.
For honeycomb structures, such as in Diesel Particulate Filters (DPF) und Gasoline Particulate Filters (GPF), FilterDict® is used to design and optimize the filters for lower pressure drop, higher filter efficiency and longer life time. The simulation steps (Azimian et al., 2018) include
- modeling the ceramic filter media (GrainGeo),
- simulating the air flow through the filter media (FlowDict) and the transport and deposition of soot particles (FilterDict®),
- the conversion of deposited particles into a porous media, and
- determining the soot layer's packing density and viscous flow resistivity.
The simulation of gas filtration at the molecular level can be simulated with the AddiDict module.
Examples of FilterDict® applications
The FilterDict® is particularly well suited to:
- Simulate aerosol filtration in a nonwoven filter.
- Simulate soot filtration through a honeycomb structure of a Diesel Particulate Filter (DPF) or Gasoline Particulate Filter (GPF).
- Simulate an oil filtration multi-pass test bench.
- Simulate micronic particle filtration and clogging phenomena in microfluidic devices.
- Particle tracking through a pleat element.
- Simulation of clogging of a high-efficiency particulate air (HEPA) filter pleat.
- Simulation of room air purifiers and their placement in the room.
Additional modules needed?
- The GeoDict® Base Package is needed for basic functionality.
- For filter media, FilterDict® works on 3D (micro-) structure models that can either be a segmented 3D image (microCT-scan, FIB-SEM) imported with ImportGeo-Vol or a 3D structure model created with one of the GeoDict® modules for digital material design, for example
- For pleats, FilterDict® requires a module to generate unresolved porous media, such as PleatGeo for pleat elements,
or GridGeo for DPF honeycomb structure models.
- For simulations on the complete filter with housing, FilterDict® requires ImportGeo-CAD to import the STL or OBJ files of the housing and convert them to voxel data for the simulation.
- FilterDict® needs the solvers of FlowDict to compute flow fields as well as the pressure drops or mean velocities.