Cars contain numerous filters for air, fuel, and oil. Those who develop these filters are confronted with many requirements. The products must withstand intense temperature fluctuations and vibrations, fit into a predefined installation space, and comply with increasingly stringent quality standards, all while development times get shorter.

Calculated flow through the filter element: flow lines mapping fluid velocity. (©Fraunhofer ITWM)

Filter manufacturers usually source the filter material and optimize their product by varying the arrangement of pleats, the housing design, and the positioning of intake and outlet openings. From the various designs, they develop prototypes that are then tested for the desired characteristics — an expensive and labor-intensive process. Computer simulations are already being used, but they often serve solely to calculate the flow through the filter. Cleaning performance or even durability — the length of time before the filter has to be replaced — can be only roughly estimated on an empirical basis.

Researchers at the Fraunhofer Institute for Industrial Mathematics ITWM are now putting an efficient simulation directly into the hands of filter developers. The Filter Element Simulation Toolbox (FiltEST) lets developers simulate the filtering processes themselves, enabling them to make realistic statements about the cleaning performance of a particular design, and predict how much dirt will accumulate over time.

The FiltEST simulation tool can calculate variables that cannot be simulated with standard procedures, as well as analyze how the filter’s characteristics will change over time; considering, for example, that the filter will get clogged with the particles it is capturing. Also, how quickly will the filter’s effectiveness decline? How much of an impact does this have on the decrease in pressure? And what effect does this have on the filtrate?

In order to calculate this pore-by-pore and particle-by-particle, the computational effort would be impossible to manage. To avoid this, the researchers neither simulate every single pore in the filter medium, nor calculate the individual particles that pass through or remain on the filter. Instead, the filter medium is viewed as a homogeneous continuum, and works with particle concentration.

For the filter manufacturer, it is of little relevance which of the many particles remain caught in the filter — they are concerned with how many the filter catches overall. This is precisely what these models calculate: the proportion of particles that passes through the filter medium, and the proportion that stays behind. In addition, the simulation can determine how many particles are washed out again from the filter over time as a result of flushing effects.

This relieves users of a massive amount of work, and saves time and money by allowing them to test in advance if the filters meet ISO standards — without having to build a single prototype.

Visit https://www.fraunhofer.de/en/press.html .