A virtual engine laboratory application for use in automotive engineering education is proposed to allow the practical teaching of powertrain calibration. The laboratory software is built as a flexible Matlab tool that can easily be transferred for applications in other disciplines and promotes the link between teaching and research.

The use of turbochargers on both gasoline and diesel engines is started to become a common strategy to comply with stringent limits on CO2. The main action towards lowering fuel consumption of powertrains is achieved by reduction of engine size and number of cylinders, annexed to the lower friction. However, this is directly linked to the worsening of deliverable output power under the natural aspirated configuration. Therefore, turbocharging is often adopted to overcome this problem where useful energy contained in the exhaust gasses is used to increase the air density at the intake. The increase in power from a natural aspirated configuration is a direct consequence of higher fuel quantity to be injected. In order to pursue a systematic evaluation of the powertrain system, engine, turbochargers and auxiliary components are included into 1D models. Several conditions can be simulated without the need of an extensive test plan. In 1D software like Ricardo Wave, turbochargers performance are imposed as input. These are previously measured in appropriate turbocharger gas-stand where hot or cold air is blown through the turbine while load on compressor is controlled by adjusting a back pressure valve. Compressor and turbine maps are generated for constant speed lines which are corrected for total temperature. Pressure ratio, mass flow and isentropic efficiency are also monitored as parameters to characterize performance maps of turbomachinery. In gas-stands, steady flow conditions are imposed at compressor and turbine. However, in turbocharged engines, pulsating flows induced by the engine valvetrain disturb continuously turbocharger conditions during the engine cycle. In fact, the effects that the conditions of the engine air-path could have on the turbocharger operations are excluded from the system modelling. In this study, an appropriate engine gas-stand has been developed in order to improve the accuracy on estimating the turbine extraction power in 1D powertrain simulations. In addition, future analyses on turbocharger transient operations could be investigated. The compressor outlet has been disconnected from the 2.2L Diesel engine intake so that the load on turbocharger and engine can be independently controlled. In order to extend the engine capability in delivering mass flow and pressure at the turbine inlet, an external boost rig has been installed with the capability to control pressure, mass flow and temperature at the engine intake. In a first instance, a 1D model of the system including turbomachinery, Diesel engine and boost rig has been developed using the commercial platform Ricardo Wave. In this way, a preliminary DoE study of the entire system has been performed in order to evaluate the effects of parameters and actuators on the turbocharger operations. Additionally, the control of the rig has been tested by confirming the previous DoE study. Approaches to create turbochargers maps are shown. Last section of the paper focuses on turbine pulsations and the interpretation of efficiency calculated in experiments and simulations.

The paper attempts to correlate simulations and measurements of turbine performance under pulsating flows for automotive turbochargers. Under real automotive powertrain conditions, turbochargers are subjected to pulsating flows, due to the motion of the engine’s valves. Experiments on a purpose-built 2.2 L diesel engine gas-stand have allowed the quantification of unsteady pulsating turbine performance. Temperature, pressure and mass flow measurements are fundamental for the characterisation of turbine performance. An adequate sampling frequency of the instruments and acquisition rates are highly important for the quantification of unsteady turbomachine performance. In the absence of fast, responsive sensors for monitoring mass flow and temperature, however, appropriate considerations would have to be taken into account when making estimates of turbine performance under pulsating flows. A 1D model of the engine gas-stand has been developed and validated against experimental data. A hybrid unsteady/quasi-steady turbine model has been adopted to identify unsteadiness at the turbine inlet and outlet. To evaluate isentropic turbine efficiency and reduce the influence of external heat transfer upon measurements, the turbine inlet temperature has been measured experimentally in the vicinity of the turbine rotor in the inlet section, upstream of the turbine tongue. The hybrid unsteady/quasi-steady turbine model considers the presence of unsteady flows in the turbine inlet and outlet, leaving the rest of the turbine to react quasi-steadily. Virtual sensors and thermocouples have been implemented in a 1D model to correlate experimental time-averaged temperature measurements.

No abstract

No abstract

Computational modeling, an important task for design, research and development stages, is evolving fast with the increase of computational capabilities over the last decades. One-dimensional (1D) CFD simulation is commonly used to analyze the flow rates and pressures of an entire fluid system of interconnected parts such as pipes, junctions, valves, and pumps. In contrast, three-dimensional (3D) CFD simulation allows detailed modeling of components such as manifolds, heat exchangers, and combustion cylinders where the flow contains significant 3D effects. Coupling a 1D model with a 3D domain potentially offers the benefits of both simulation strategies in one co-simulation approach. The present study provides a deep understanding of the co-simulation approach by listing all necessary steps need to be followed before and during the coupling of the 1D and 3D simulation software. It analyses the simulation and convergence time requirements based on the 3D model mesh quality and compares this approach with the current 1D–3D uncoupled approach followed in the industry. The outputs of both simulation approaches are then compared with experimental results. The co-simulation time mainly depends on the mesh quality of the 3D domain and the number of inner iterations per time-step which is entirely determined by the nature and complexity of the simulation. The co-simulation time per engine cycle is almost identical to the uncoupled approach. However, it was found that the number of cycles required for convergence in the coupled approach is nearly double than the uncoupled approach. The comparison between the two simulation approaches and the experimental results demonstrated the very 3D nature of the flows, the sensitivity of the uncoupled approach to input conditions and the sensitivity of co-simulation to the averaged boundary conditions transferred from the 1D model back to the 3D domain.

Turbochargers are inherently dynamic devices, comprising internal flow volumes, mechanical inertias and thermal masses. When operating under transient conditions within an engine system, these dynamics need to be better understood. In this paper, a new non-dimensional modelling approach to characterise the turbocharger is proposed. Two new dimensionless quantities are defined with respect to mechanical and thermal transient behaviour, which are used in conjunction with the Strouhal number for flow transients. The modelling approach is applied to a small wastegated turbocharger and validated against experimental results. The model is used to simulate the turbocharger mass flow rate, turbine housing temperature and shaft speed responses to different excitation frequencies for different sizes of turbine. The results highlight the influence of turbocharger size on the dynamic behaviour of the system, which is particularly marked for the turbine housing temperature. At certain frequency ranges, the system behaviour is quasi-steady, allowing modelling through static maps in these operating regions. Outside these ranges, however, transient elements play a more important role. The simulation study shows that the proposed dimensionless parameters can be used to normalise the influence of turbine size on the dynamic response characteristics of the system. The model and corresponding dimensionless parameters can be applied in future simulation studies as well as for turbocharger matching in industry.

In regulated two-stage sequential turbocharging systems, a smaller, high pressure (HP), and a larger, low pressure (LP), turbocharger are sequentially positioned to recover the energy available in the exhaust gases and deliver acceptable level of boost to the intake of an internal combustion engine. Due to the different sizes of the turbochargers, by-pass valves are placed in the system to control operations. Due to the turbocharging system layout, it is clear that the air pressurized by the LP compressor enters non-uniformly the HP compressor. This is caused by the rotating radial compressor and the interconnecting bends which cause swirl and velocity to scatter, respectively. Furthermore, the heat transfer in the two turbochargers may have an effect on the apparent efficiencies. For these reasons, the standard mapping approach for turbochargers is not able to take into account the effect of non-uniform flow and heat transfer. In this paper, a novel approach for mapping the two-stage turbocharging system is proposed and performed into a mono-dimensional simulation code. Although, flow non-uniformity and turbochargers heat transfer effects on the performance of the turbocharging system are not considered, at this present time, the study centralizes on the investigation and the validation of the mapping approach. In fact, a two-stage sequential turbocharging system has been considered for the study and a simulation code to investigate the mapping technique has been implemented.