Modelling of turbocharger heat transfer under stationary and transient engine operating conditions (original) (raw)

Abstract

A lumped capacity heat transfer model has been developed and compared to measurements from a turbocharger operating on a 2.2L Diesel engine under steady and transient conditions ranging from 1000-3000rpm and 2-17bar BMEP. The model parameters have been estimated based on similar devices and this study quantifies the errors associated with this approach. Turbine outlet gas temperature prediction was improved with RMSE reduced from 29.5 o C to 13 o C. A sensitivity study showed the parameters of the heat transfer model influence gas temperatures by only ±4 o C but housing temperatures by up to 80 o C. Transient simulations showed how errors in the thermal capacitance also lead to errors. This study shows the importance of undertaking a full thermal characterisation and the need for accurate adiabatic maps in turbocharger simulations.

Figures (10)

An overview w of energy flows in turbochargers is shown in figure 1 (a). In addition to work, heat is also transferred between turbine and compressor. This heat can flow to ambient or along the bearing housing towards the cooling oil and further ambient losses. Depending on the operating conditions, heat may continue to be conducted to the compressor housing or, under high compression ratios, this heat flow can reverse.

An overview w of energy flows in turbochargers is shown in figure 1 (a). In addition to work, heat is also transferred between turbine and compressor. This heat can flow to ambient or along the bearing housing towards the cooling oil and further ambient losses. Depending on the operating conditions, heat may continue to be conducted to the compressor housing or, under high compression ratios, this heat flow can reverse.

The turbocharger thermal model contains a number of parameters that are not easily determined by inspection of the geometry without considerable simplification (5 capacitances, 4 conductances and 7 convective correlations). Although an experimental procedure has been determined to characterise these parameters for an individual turbocharger, the aim of this work is to assess the predictive capability of the model for a new turbocharger, based on the parameters established previously for a similar device. A range of turbocharger sizes has previously been measured and simple correlations have been observed between external geometries and thermal parameters: these have been used to calculate the parameters in this study, this replicating an industrial scenario.

The turbocharger thermal model contains a number of parameters that are not easily determined by inspection of the geometry without considerable simplification (5 capacitances, 4 conductances and 7 convective correlations). Although an experimental procedure has been determined to characterise these parameters for an individual turbocharger, the aim of this work is to assess the predictive capability of the model for a new turbocharger, based on the parameters established previously for a similar device. A range of turbocharger sizes has previously been measured and simple correlations have been observed between external geometries and thermal parameters: these have been used to calculate the parameters in this study, this replicating an industrial scenario.

Figure 3: Engine speed and torque operating points for stably and transient experiments (BMEP: Brake Mean Effective Pressure) 3.1 Experimental Setup and fest Points A variable geometry turbocharger has been instrumented to capture fluid and _ structure temperatures at each of the locations described by the nodes of the model. For all gas temperatures, multiple thermocouples were used to capture distributions across the cross section of the ducts (figure 2a). For the compressor and turbine housings, temperatures were measured at 3 azimuths and 2 depths to provide information about temperature distributions (figure 2b). The turbocharger was installed on a 2.2L Diesel engine for which the usual application is a light commercial vehicle. The engine was operated on an AC transient dynamometer. Mass flow was measured at engine intake using an ABB Sensyflow hot wire flow meter (accuracy <1%); Pressures were measured using Kistler Piezo-Resistive sensors (accuracy 0.5%); gas and metal temperatures were measured using 1.5mm _ k-type thermocouples and turbocharger speed was measured using an eddy current blade count device from Micro-Epsillon.

Figure 3: Engine speed and torque operating points for stably and transient experiments (BMEP: Brake Mean Effective Pressure) 3.1 Experimental Setup and fest Points A variable geometry turbocharger has been instrumented to capture fluid and _ structure temperatures at each of the locations described by the nodes of the model. For all gas temperatures, multiple thermocouples were used to capture distributions across the cross section of the ducts (figure 2a). For the compressor and turbine housings, temperatures were measured at 3 azimuths and 2 depths to provide information about temperature distributions (figure 2b). The turbocharger was installed on a 2.2L Diesel engine for which the usual application is a light commercial vehicle. The engine was operated on an AC transient dynamometer. Mass flow was measured at engine intake using an ABB Sensyflow hot wire flow meter (accuracy <1%); Pressures were measured using Kistler Piezo-Resistive sensors (accuracy 0.5%); gas and metal temperatures were measured using 1.5mm _ k-type thermocouples and turbocharger speed was measured using an eddy current blade count device from Micro-Epsillon.

Figure 4: Turbocharger model boundary conditions for steady and transient simulations EE One-dimensional simulation code OpenWAM™ (12, 13) has been used to create the turbocharger model with the connecting ducts as shown in figure 4. This highlights the compressor and turbine models (that are based on the manufacturer’s maps) connected via a mechanical shaft and the lumped thermal model of the housing. The turbine map was determined from compressor enthalpy rise and is therefore adiabatic with respect to turbine performance, however this includes mechanical losses and was modified by applying the mechanical losses models developed by Serrano et al. (6, 14). The turbine is represented by a series of two nozzles with an intermediate reservoir to account for acoustic effects (15, 16).

Figure 4: Turbocharger model boundary conditions for steady and transient simulations EE One-dimensional simulation code OpenWAM™ (12, 13) has been used to create the turbocharger model with the connecting ducts as shown in figure 4. This highlights the compressor and turbine models (that are based on the manufacturer’s maps) connected via a mechanical shaft and the lumped thermal model of the housing. The turbine map was determined from compressor enthalpy rise and is therefore adiabatic with respect to turbine performance, however this includes mechanical losses and was modified by applying the mechanical losses models developed by Serrano et al. (6, 14). The turbine is represented by a series of two nozzles with an intermediate reservoir to account for acoustic effects (15, 16).

Figure 5: Predicted heat flows in selected turbocharger nodes

Figure 5: Predicted heat flows in selected turbocharger nodes

Table 2: Ranges of sensitivity tested

Table 2: Ranges of sensitivity tested

Figure 6: Temperature changes due to compression/expansion and heat transfer for (a) turbine and (b) compressor A sensitivity study was conducted to quantify the influence of each parameter in Table 2. Based on the results in figure 7, the following observations are made:

Figure 6: Temperature changes due to compression/expansion and heat transfer for (a) turbine and (b) compressor A sensitivity study was conducted to quantify the influence of each parameter in Table 2. Based on the results in figure 7, the following observations are made:

Figure 7: Model sensitivity to model parameters for (a) turbine gas outlet, (b) compressor gas outlet, (c) Node T and (d) Node C (Axis refers to different operating conditions, HT: Heat Transfer)

Figure 7: Model sensitivity to model parameters for (a) turbine gas outlet, (b) compressor gas outlet, (c) Node T and (d) Node C (Axis refers to different operating conditions, HT: Heat Transfer)

Figure 8: Model sensitivity to thermal capacitance for (a) turbine and (b) compressor

Figure 8: Model sensitivity to thermal capacitance for (a) turbine and (b) compressor

Loading...

Loading Preview

Sorry, preview is currently unavailable. You can download the paper by clicking the button above.

References (16)

  1. M. Cormerais, J. F. Hetet, P. Chesse, and A. Maiboom, Heat Transfer Analysis in a Turbocharger Compressor: Modeling and Experiments, SAE Paper Number 2006-01-0023, 2006.
  2. J. R. Serrano, C. Guardiola, V. Dolz, A. Tiseira, and C. Cervelló, Experimental Study of the Turbine Inlet Gas Temperature Influence on Turbocharger Performance, SAE Paper Number 2007-01-1559, 2007.
  3. S. Shaaban, Experimental investigation and extended simulation of turbocharger non-adiabatic performance, PhD, Fachbereich Maschinenbau, Universität Hannover, 2004.
  4. N. Baines, K. D. Wygant, and A. Dris, The analysis of heat transfer in automotive turbochargers, Journal of Engineering for Gas Turbines and Power, vol. 132(4), 2010.
  5. A. Romagnoli and R. Martinez-Botas, Heat transfer analysis in a turbocharger turbine: An experimental and computational evaluation, Applied Thermal Engineering, vol. 38, pp. 58-77, 2012.
  6. J. R. Serrano, P. Olmeda, A. Tiseira, L. M. Garcia-Cuevas, and A. Lefebvre, Theoretical and experimental study of mechanical losses inautomotive turbochargers, Energy, vol. 55, pp. 888-898, 2013.
  7. M. Cormerais, P. Chesse, and J.-F. Hetet, Turbocharger heat transfer modeling under steady and transient conditions, International Journal of Thermodynamics, vol. 12, pp. 193-202, 2009.
  8. P. Olmeda, V. Dolz, F. J. Arnau, and M. A. Reyes-Belmonte, Determination of heat flows inside turbochargers by means of a one dimensional lumped model, Mathematical and Computer Modelling, vol. 57, pp. 1847-1852, 2013.
  9. J. R. Serrano, P. Olmeda, A. Paez, and F. Vidal, An experimental procedure to determine heat transfer properties of turbochargers, Measurement Science & Technology, vol. 21, Mar 2010.
  10. S. Shaaban, J. R. Seume, R. Berndt, H. Pucher, and H. J. Linnhoff, Part Load performance prediction of turbocharged engines, presented at the 8th International Conference on Turbochargers and Turbocharging, May 17, 2006 -May 18, 2006, London, United Kingdom, 2006.
  11. F. Payri, P. Olmeda, F. J. Arnau, A. Dombrovsky, and L. Smith, External heat losses in small turbocharger: model and experiments, Submitted to Energy, 2013.
  12. OpenWAM, 2012. Available: www.Openwam.org
  13. J. Galindo, J. R. Serrano, F. J. Arnau, and P. Piqueras, Description of a Semi- Independent Time Discretization Methodology for a One-Dimensional Gas Dynamics Model, Journal of Engineering for Gas Turbines and Power- Transactions of the Asme, vol. 131, May 2009.
  14. J. R. Serrano, P. Olmeda, A. Tiseira, L. M. García-Cuevas, and A. Lefebvre, Importance of Mechanical Losses Modeling in the Performance Prediction of Radial Turbochargers under Pulsating Flow Conditions, SAE Int. J. Engines, vol. 6, pp. 729-738, 2013.
  15. J. R. Serrano, F. J. Arnau, V. Dolz, A. Tiseira, and C. Cervelló, A model of turbocharger radial turbines appropriate to be used in zero-and one- dimensional gas dynamics codes for internal combustion engines modelling, Energy Conversion and Management, vol. 49, pp. 3729-3745, 2008.
  16. M. A. Reyes-Belmonte, Contribution to the Experimental Characterization and 1-D Modelling of Turbochargers for IC Engines, PhD, Departamento de Máquinas y Motores Térmicos, Universitat Politècnica de València, València, 2013.