# A reduced order pseudochannel model accounting for flow maldistribution in automotive catalysis

## TL;DR

This paper introduces a new pseudochannel model that improves predictions of automotive catalytic converters while keeping computational costs low.

## Contribution

A novel pseudochannel model is proposed that better captures flow maldistribution effects compared to traditional single-channel models.

## Key findings

- The pseudochannel model outperforms conventional single-channel models in both transient and steady-state test cases.
- The computational cost of the pseudochannel model is comparable to that of single-channel models.
- Flow maldistribution effects can be effectively incorporated into single-channel models using steady-state data.

## Abstract

Exhaust aftertreatment systems (EATS) play a critical role in reducing emissions and ensuring compliance with stringent emission regulations. Catalytic converters, as part of EATS, involve complex physico-chemical processes. To accurately predict their behavior in realistic geometries, transient 3D models are necessary. However, the computational cost associated with simulations based on such models prevents their application to long-time behaviors as well as in real-time control and diagnostics. While single-channel models (SCMs) are computationally efficient, they struggle to provide accurate predictions during real-time operations with flow maldistribution. In this study, we propose a pseudochannel model derived using steady-state reactive 3D simulations and a nonlinear least squares optimization technique. We show that the performance of this pseudochannel model is superior to a conventional SCM in both transient and steady state test cases. At the same time, the computational cost of the pseudochannel model is equivalent to that of the SCM. These results imply that flow maldistribution effects can be well incorporated in SCMs via a pseudochannel approach that relies on relatively inexpensive steady-state system data.