(380e) Multi-Scale Modelling of Gasoline Particulate Filters – How the Porous Structure of Filter Affects Its Performance

Introduction

The latest emission standards
require the use of particulate filters in both diesel and gasoline powered
cars. Catalytically active coating is applied directly in/onto the porous
filter walls [1] in order to achieve size & cost
reduction of the exhaust aftertreatment system, prevention of heat losses and
improved filter regeneration. Catalytic filters need to pass the requirements
for filtration efficiency and conversion of gaseous pollutants while
maintaining low enough pressure loss. A vital part of the filter design is
therefore porous structure optimization of both substrate and catalytic coating,
as well as washcoat location. In this work, we employ
multi-scale mathematical modeling (Fig. 1) combined with 3D X-ray tomography
(XRT) to study the relationship between the filter microstructure [1, 2] and
entire device performance [3]. 

Experimental

Model gasoline particulate
filters (GPF) supplied by Johnson Matthey consisted of cordierite substrate
coated with either γ-Al₂O₃ or Pd/γ-Al₂O₃.
Three samples (2.5 cm diameter, 11.5 cm length) of each were prepared aiming
for the same washcoat loading and varying distribution:
in-wall coated (CF1), on-wall coated (CF3) and combined (CF2). Sample
performance was tested in pressure loss and CO oxidation experiments in a
heated lab reactor with FTIR analyser and
differential manometer.

Filter structure was
characterized by Hg porosimetry (porosity, pore-size
distribution), SEM and XRT [1]. 3D morphology of the filter wall including the washcoat distribution was reconstructed from XRT images.
The reconstructed medium was segmented (ImageJ) and transformed into simulation
mesh for the CFD solver (OpenFOAM). Flow through free
pores and washcoat was simulated by porousSimpleFoam solver, while an in-house developed solver
was used for component diffusion and reactions [2]. This way, 3D pressure, velocity
and concentration fields were obtained and post-processed to evaluate the
permeability and reaction effectiveness factor, which were used as input
parameters of the full-scale 1D+1D model of catalytic monolith channels [3]
that includes also inlet and outlet effects on pressure loss (Fig. 1).

Results and Discussion

The simulation results suggest that the gas predominantly
flows through remaining free pores in the filter wall and/or cracks in the
coated layer (Fig. 1). Although the transport into the catalytically active washcoat is enabled by diffusion, it was found that the
presence of an on-wall layer is beneficial for CO conversion (CF2 and CF3 in
Fig. 2a). However, a thick compact on-wall layer brings about a significant
increase of pressure loss (Fig. 2b) as the local permeability of the coating is
an order-of-magnitude smaller than that of bare filter wall. The best trade-off
between effective reaction rate and pressure loss was reached in sample CF2, in
which a thin washcoat layer overlies the in-wall washcoat [2]. The multi-scale model predictions match well
the measured data.

Figure
1. Multi-scale model of catalytic monolith filter. Left: 3D
reconstructed porous wall (grey = substrate, yellow = coated catalytic
material) with calculated flow field (streamlines), sample CF1 [2]. Right:
Scheme of the filter channels simulated with full-scale 1D+1D model [3].

Figure
2. a) CO conversion simulated by the reaction-diffusion
CFD model near the light-off temperature (180°C), b) measured pressure
loss vs. results predicted by the full-scale model, utilizing wall
permeabilities predicted from the CFD pore-scale simulations.

Conclusions

The work introduces novel and systematic methodology for
characterization, modelling and performance evaluation of catalytic particulate
filters for automotive exhaust gas aftertreatment, enabling their rational
design. The methodology can be used for both gasoline and diesel particulate
filters coated with a catalytically active material. The best trade-off between
catalytic activity (CO oxidation) and pressure loss was found in the case of
CF2 sample, in which in-wall coating is supplemented with a thin layer of
on-wall coating.

References

1.     Vaclavik,
M. et al, Structure characterisation
of catalytic particulate filters for automotive exhaust gas aftertreatment. Mater. Charact.
134 (2017), 311-318.

2.     Koci,
P. et al, 3D reconstruction and pore-scale
modeling of coated catalytic filters for automotive exhaust gas aftertreatment.
Catal. Today (2018), in press, DOI:
10.1016/j.cattod.2017.12.025

3.     Schejbal,
M. et al, Modelling of diesel filters for particulates removal. Chem. Eng. J. 154 (2009), 219–230.

Acknowledgements

The work has been financially supported by the European
Union’s Horizon 2020 research and innovation program under grant agreement No. 686086
(project PARTIAL-PGMs), Johnson Matthey and specific university research (MSMT
No 20-SVV/2018).