Search Results

In an attempt to reduce particulate and NOx emissions from Diesel exhaust, the combined DPF and SCR filter is now frequently chosen as the preferred catalyst. When this device functions effectively it saves valuable packaging space in a passenger vehicle. As part of its development, modelling of its emissions performance is essential. Single channel modelling would seem to be the obvious choice for an SCRF because of its complex internal geometry. This, however, can be computationally demanding if modelling the full monolith. For a normal flow-through catalyst monolith the porous medium approach is an attractive alternative as it accounts for non-uniform inlet conditions without the need to model every channel. This paper attempts to model an SCRF by applying the porous medium approach. The model is essentially 1D but as with all porous medium models, can very easily be applied to 3D cases once developed and validated.

Use of selective catalytic reduction technology is the most popular strategy for removing NOx from lean diesel exhaust. The reductant is essentially ammonia and this has been supplied as a spray of urea droplets, but more recently alternative technology where ammonia gas is released from a storage medium has become a viable alternative. Experiments have been carried out on an engine test rig run to steady state conditions using NOx composed of either 25% or 50% of NO₂, with ammonia gas as the reductant. This was a 1D study where a long 10 degree diffuser provided uniform temperature and velocity profiles to the SCR catalyst brick. Under the transient conditions that occur during drive cycles, the dosing of the ammonia can deviate from the optimum. In this study, the dosage rate of ammonia was held at a fixed value, while the engine load was varied.

Removal of NOx from lean diesel exhaust can be achieved by the use of selective catalytic reduction technology. The supplied reductant is often ammonia, either as urea or as ammonia gas released from a storage medium. Experiments have been carried out on an engine test rig run to steady state conditions using NOx composed mainly of NO, with ammonia gas as the reductant. This was essentially a 1D study because a long 10 degree diffuser was used to provide uniform temperature and velocity profile to the SCR catalyst brick in the test exhaust system. Tuning of the standard reaction, the NO SCR reaction, in a kinetic scheme from the literature and adjustment of the ammonia adsorption kinetics achieved improved agreement between the measurements and CFD simulations. This was carried out for studies at exhaust gas temperatures between 200 and 300°C.

Modeling of SCR in diesel exhaust systems with injection of urea spray is complex and challenging but many models use only the conversion observed at the brick exit as a test of the model. In this study, the case modeled is simplified by injecting ammonia gas in nitrogen in place of urea, but the spatial conversion profiles along the SCR brick length at steady state are investigated. This is a more rigorous way of assessing the ability of the model to simulate observations made on a test exhaust system. The data have been collected by repeated engine tests on eight different brick lengths, all which were shorter than a standard-sized SCR. The tests have been carried out for supplied NH₃ /NOx ratios of a 1.5, excess ammonia, a 1.0, balanced ammonia, and a 0.5, deficient ammonia. Levels of NO, NO₂ and NH₃ have been measured both upstream and downstream of the SCR using a gas analyzer fitted with ammonia scrubbers to give reliable NOx measurements.

Removal of NOx from a light-duty diesel automotive exhaust system can be achieved by SCR reactions using aqueous urea spray as the reductant. Measurements of emissions from such a system are necessary to provide data for CFD model validation. A test exhaust system was designed that featured an expansion can, nozzle and diffuser arrangement to give a controlled flow profile to define an inlet boundary for a CFD model and to approximate to one-dimensional flow. Experiments were carried out on the test exhaust using injection of either ammonia gas in nitrogen or aqueous urea spray. Measurements were made of NO, NO₂ and NH₃ at inlet to and exit from the SCR using a CLD analyzer. The NO and NO₂ profiles within the bricks were found by measuring at the exit from different length bricks. The spray and gas measurements were compared, and insights into the behavior of the droplets upstream and within the bricks were obtained.

Lean burn after treatment systems using NOX traps for reducing emissions from diesel exhausts require periodic regeneration after each storage stage. Optimising these events is a challenging problem and a model capable of simulating these processes would be highly desirable. This study describes an experimental investigation, which has been designed for the purpose of validating a NOX trapping and regenerating model. A commercial computational fluid dynamics (CFD) package is used, to model NOX trapping and regeneration, using the porous medium approach. This approach has proved successful for three way catalysis modelling. To validate the model a one-dimensional NOX trap system has been tested on a turbocharged, EGR cooled, direct injection diesel engine controlled with an engine management system via DSPACE. Fast response emission analysers have been used to provide high resolution data across the after-treatment system for model validation.

This paper investigates the flow maldistribution across the monolith of an axisymmetric catalyst assembly fitted to a pulsating flow test rig. Approximately sinusoidal inlet pulse shapes with relatively low peak/mean ratio were applied to the assembly with different amplitudes and frequencies. The inlet and outlet velocities were measured using Hot Wire Anemometry. Experimental results were compared with a previous study, which used inlet pulse shapes with relatively high peak/mean ratios. It is shown that (i) the flow is more maldistributed with increase in mass flow rate, (ii) the flow is in general more uniformly distributed with increase in pulsation frequency, and (iii) the degree of flow maldistribution is largely influenced by the different inlet velocity pulse shapes. Transient CFD simulations were also performed for the inlet pulse shapes used in both studies and simulations were compared with the experimental data.

This paper describes the coupling of a 1D engine simulation code (Ricardo WAVE) to a 3D CFD code (STAR-CD) to study the flow behaviour inside a Close-Coupled Catalytic converter (CCC). A SI engine was modelled in WAVE and the CCC modelled in STAR-CD. The predictions of the stand-alone WAVE model were validated against engine bed tests before the coupled 1D/3D simulations were performed at 3000 RPM WOT for both motored and firing conditions. The predicted exhaust velocities downstream of the catalyst monolith in the coupled simulations matched fairly well with Laser Doppler Anemometry (LDA) measurements.

Performance improvements of automotive catalytic converters can be achieved by improving the flow distribution of exhaust gases within the substrate. The flow distribution is often assumed to be adequately described by measurements obtained from steady flow rigs. An experimental study was carried out to characterise the flow distribution through the substrate of a close-coupled catalytic converter for both steady and pulsating conditions on a flow rig and on a motored engine. Computational fluid dynamic (CFD) simulations were also performed. On the flow rig, the flow from each port was activated separately discharging air to different regions of the substrate. This resulted in a high degree of flow maldistribution. For steady flow maldistribution increased with Reynolds number. Pulsating the flow resulted in a reduction in flow maldistribution. Different flow distributions were observed on the motored engine when compared to composite maps derived from the rig.

This study examines the effect of pulsating flow on the flow distribution through contoured substrates. Three ceramic contoured substrates of equal volume were assessed. Two of the substrates were cone shaped with different cone angles and one had a dome shaped front face. The flow distribution was measured for a range of flow rates and pulsation frequencies. Computational Fluid Dynamics (CFD) simulations were also performed. It is shown how a contoured substrate can provide improvements in flow uniformity and that they are less sensitive to changes in flow rate and pulsation frequency when compared to the case of a standard substrate. Improvements in the prediction of flow distribution are reported when substrate “entrance effects” are accounted for.

Heat transfer in automotive exhaust catalyst systems with metallic substrates is modeled using a commercial Computational Fluid Dynamics (CFD) code. The substrate channels are modeled by approximating their geometry as both triangular and sinusoidal. The effect of the packing arrangement of adjacent channels is investigated. The effect of the angle of the flow entering ceramic substrate monoliths on the localised heat transfer is also studied and the related implications for catalyst aging and light off deduced.

Conversion efficiency, durability and pressure drop of automotive exhaust catalysts are dependent on the flow distribution within the substrate. This study examines the effect on flow distribution using substrates which feature contoured front faces. Three ceramic contoured substrates of equal volume were assessed. Two of the substrates were cone shaped with different cone angles and one had a dome shaped front face. Pressure drop and flow distribution was measured for a range of flow rates and substrate positions. Computational Fluid Dynamics (CFD) simulations were also performed to provide insight into flow behaviour. It is shown how a contoured substrate can provide improvements in flow uniformity and pressure drop when compared to the case of a standard non-contoured substrate.

An experimental and theoretical investigation has been performed on the flow and pressure loss in axisymmetric catalytic converters and isolated monoliths under steady, isothermal flow conditions. Monolith resistance has been measured with a uniform, low turbulence, incident flow field. It has been found that the pressure loss expression for fully developed laminar flow is a good approximation to observations for x+ greater than 0.2. However, for x+ less than 0.2 the additional pressure loss due to developing flow is no longer negligible and a better approach is to use the correlation proposed by Shah (16). From experimental studies on the axisymmetric catalytic converters non-dimensional power law relationships have been derived relating maldistribution and pressure drop to expansion length, Re, and monolith length. These expressions are shown to generally fit the data well within ±5%.

The purpose of this paper is to characterise barrel swirl behaviour in a production four-valve engine with pentroof chamber. Steady flow analysis showed that the insertion of tubes into the cylinder head's induction tracts increased the tumbling ratio of the in-cylinder flow field at intake valve closure. A comparison of LDA measurements, conducted along the spark plug axis, for tubes and no tubes inserted yielded the following conclusions. The results indicated that the barrel swirl generated was not efficiently breaking down into turbulence but forming two counter-rotating vortices in the horizontal cylinder plane. The turbulence levels and cycle-to-cycle flow variations towards the end of the compression stroke increased with tumbling ratio. The former suggested faster combustion rates if applied to a lean burn engine, however, the latter suggested greater cyclic combustion instability and may limit lean burn capability.

The purpose of this paper is to characterise the flow field through the inlet valves, and tumble to swirl conversion tube of a steady state flow rig using HWA. LDA and CFD techniques. Three cylinder head configurations were developed which were found to produce three levels of barrel swirl motion. The swirl precesses around the central core of the conversion tube at all degrees of swirl. Varying degrees of swirl produced differing axial velocity profiles, and flow reversal occurred in the central core of the conversion tube for the high swirling case. The results obtained for this study indicate that care must be exercised when deducing the barrel swirl ratio for real engines from steady flow rig analysis.

With the increasingly widespread use of catalytic converters for meeting exhaust emission regulations, considerable attention is currently being directed towards improving their performance. Experimental analysis is costly and time consuming. A desirable alternative would be a computational model based on established numerical techniques. To this end a transient three-dimensional model has been developed using a commercial CFD code. It simulates the fluid dynamics, chemical kinetics and heat and mass transfer that takes place in catalysts and their associated assembly. As a result the model can be used to predict important performance parameters such as conversion efficiency, incurred pressure drop and the thermal environment.