This case study of Ashok Leyland shows that computational fluid dynamics (CFD) can be used to simulate flow through an exhaust system to determine the exhaust back pressure (EBP) and uniformity index (UI) at a pre-prototype stage to ensure that it meets the design requirements.
Environmental considerations have resulted in increasingly strict norms on vehicle emissions. To meet these demands, after-treatment systems have become a necessity in modern vehicles. The essential constituents of an after-treatment system are catalytic converters and filters. While these components reduce soot and harmful gases from the exhaust, they end up generating huge resistance to the flow of exhaust gases. This increases the EBP in the exhaust system. A higher EBP in the exhaust system requires more work to be done by the engine to expel the exhaust gases. It could also cause more exhaust to be mixed with fresh inlet air into the engine causing issues in the performance of the engine. It is therefore essential to design exhaust systems such that the pressure drop across the system is lower than the recommended limit for the engine.
To maximise the efficiency of after-treatment systems, it is desired that the flow entering them have uniform velocity over the entire area. Non-uniformity of flow leads over-utilisation or under-utilization of regions of some regions of the catalytic converter. CFD can be used to study the velocity distribution and pressure drop during the CAD stage of design so that suitable modifications can be done in the design to achieve desired level of uniformity and pressure drop before prototyping. AcuSolve CFD solver has been used in this study. The flow has been considered as steady state. RANS with SA turbulence model is used. A porous media approach has been used for the catalytic converter.
While noise reduction and after-treatment, both are functions of regulatory importance, the muffler and catalytic converter are also regions of high energy loss. The muffler chambers are regions of separated flow. The substrate of the catalytic converter is a dense honeycombed structure that provides increased surface area for reactions, but also introduce flow restriction and consequent pressure loss. The flow of the exhaust gases through the exhaust pipes is also associated with typical pipe flow losses which are particularly high at bends and constrictions.
To drive flow of exhaust through such a system, sufficient pressure should be developed in the cylinder against which the piston has to push. This pressure is commonly referred to as exhaust back pressure (EBP). A high EBP also results in higher quantity of combustion products to be trapped in the cylinder which mixes with the fresh air for the next cycle affecting engine performance. It is, therefore, necessary to design an exhaust system that meets the performance and regulatory requirements with the EBP contained within certain limits. CFD helps estimate the EBP during design stage itself.
While using a catalytic converter, it is desired that the velocity of flow entering the catalytic converter is uniform over the entire inlet surface. Non-uniformity in velocity particularly results from the diverging cone that bridges the larger catalytic converter diameter to the pipe diameter. It is therefore helpful to estimate the uniformity of flow on the catalytic converter inlet surface. Uniformity index (UI) is a measure of the uniformity of flow.
The exhaust gas is considered to have properties same as air and is modeled as an ideal gas. Acusolve - a general purpose CFD solver of Altair - provides the RANS (Reynolds Averaged Navier Stokes) model to solve for flow and energy and SA to model turbulence in the flow. The mass flow rate and temperature corresponding to the desired test condition are specified as the boundary conditions at the inlet. The outlet boundary is specified to have pressure equal to the atmospheric pressure.
The catalytic converter is modeled using a porous medium approach where the pressure drop is modeled as a sum of an inertial and viscous component, the inertial component proportional to the velocity squared and viscous component proportional to the velocity. An anisotropic pressure drop profile is used wherein the pressure drop in cross wise direction is fixed at a value several times higher than that along the flow direction thereby enforcing the flow to be directed along the flow direction. There is a provision in AcuSolve to input a characteristic pressure drop v/s velocity data from which it calculates the inertial and viscous coefficients.
With the post-processing tool AcuFieldView, the average pressure over any desired surface can be calculated. Streamlines can be generated to observe the nature of flow in the domain and plot the variation of any property with flow. Contours can be plotted to observe the variation of any property over a surface. Uniformity Index (UI) of velocity on a given surface can be calculated by defining the equation for the quantity at any surface. The separated flow in the muffler chambers can be seen in Figure 3.
AcuSolve has models such as RANS, SA and porous media approach to simulate the flow through different components in the exhaust system. The catalytic converter pressure drop data can be directly fed in from which the software calculates the coefficients. The software has various post processing tools to visualise results and calculate necessary values such as pressure drop and UI.
CFD can be used to study the velocity distribution and pressure drop during the CAD stage of design so that suitable modifications can be done in the design to achieve desired level of uniformity and pressure drop before prototyping.