Research Topic

Analyze and evaluate the effects of biofuels on combustion in IC-Engine

More than 80% of worldwide primary energy is provided by petroleum, natural gas, coal and other fossil fuels. Due to ever increasing world population and heavy industrialization, the demand for petroleum fuel will be persistent in the foreseeable future. However, the resources of petroleum fuels are limited. Biofuel is a source of alternative fuels. It is widely considered that bio fuel combustion is “carbon neutral” because it only releases carbon taken from the atmosphere during plant growth.

 

The main issues faced by the IC engine developers are how to increase the engine efficiency as well as to reduce the pollutant emission. Computational Fluid Dynamics (CFD) is a powerful tool for researchers, as a supplement to experiments, for unterstanding the complex phenomena taking place in IC Engines, such as fuel and air mixed two-phase flow, ignition, combustion, as well as pollutant formation. On one hand, the three dimensional field of the computed variables can be analyzed and used for better understanding, and subsequent engine analysis can be quickly carried out to improve engine design. In this way, the optimization of engine geometry/design can be carried out more easily and economically using numerical simulation than experiment.

 

The combustion process in IC Engines strongly depends on the turbulent and cyclic fluctuation of in-cylinder flow. In comparison with Reynolds Averaged Navier Stokes (RANS), Large Eddy Simulation (LES) offers in principle the possibility to predict cyclic variations in IC-Engines. In this PhD work, the LES for in-cylinder flow and LES-based simulation for ignition and combustion in a real biofuel engine with complex geometry will therefore be carried out, using CFD code Kiva4-MPI coupled with Flamelet-Generated Manifold (FGM) table.

 

 

 

 

Recent work

The original KIVA4-MPI code has only ability to run RANS simulations. In order to perform LES, the Smagorinsky model with parallel computing capability was implemented into the code. The new KIVA4-MPI-LES code is validated by simulating a fully developed turbulent channel flow driven by gravity (in z direction ) in a square duct with periodic boundary conditions (see in Figure 1). This configuration was chosen on one hand to look for the evidence of secondary fluid motion that generally appears near the corners of the square duct as a reflection of the anisotropy of the shear stress tensor. The prediction of this secondary fluid motion would present a severe test case for a LES code. On the other hand, the geometry is very simple to evaluate the predictive capability of LES methodology, where DNS results are readily available.

 

Fig.1

Figure 1. Geometry (x,y,z=5cm X 5cm X 30cm) and computational mesh (x,y,z=40 X 40 X 60) of square duct. The flow is along the z direction

 

Figure 2 (a) shows the mean streamwise velocity profiles along the central line at the middle section of the duct. Figure 2 (b) shows the resolved velocity fluctuations in streamwise at the same location. The x- coordinate is normalized with H=5cm, which is the width as well as height of this duct. DNS data are taken from Gavrilakis [1] and KIVA3V-LES data are taken from Huijnen [2]. The comparison shows a good agreement between DNS and KIVA4-MPI-LES results. The streamwise turbulence in Figure 2 (b) which is the main source of the turbulent kinetic energy shows slightly offset towards right side in case of KIVA4-MPI-LES as compared to DNS. This appearance can be attributed to the fact that the turbulent dissipation close to the wall is over predicted in the LES case.

 

Figure 2. (a) Mean streamwise velocity profile at z = 15 cm; (b) Turbulence profile in streamwise at z = 15 cm; Dashed line: DNS simulation [1], Solid line: KIVA3V-LES with Smagorinsky model [2], hollow square: KIVA4-MPI-LES with Smagorinsky model

 

Fig. 3 shows the mean secondary velocity profile obtained with KIVA4-MPI-LES. It is clearly visible that developed secondary flows near the duct corners are captured nicely by LES. This also shows the viability of Samgorinsky model in resolving flow structures, that would be very useful while carrying out IC-engine simulation to understand unsteady in-cylinder flow dynamics and cyclic variations.

Figure 3. Spanwise vector field averaged over eight octants at z = 15 cm

 

The goal of this Ph.D. work is to analyze and evaluate the effects of biofuels on combustion characteristics in IC-Engines. For this purpose an engine configuration also known as “Darmstadt optical engine” is chosen. Figure 4 shows the schematic of the engine configuration with 4 valves (2-intake and 2-exhaust). The CAD geometry for the Darmstadt optical engine and respective mesh are shown in Figure 5. Large Eddy Simulation is carried out using the Smagorinsky model for the aforesaid engine configuration using KIVA4-MPI for 50 cycles to be able to carry out statistical analysis of in-cylinder flow field. Quantitative and qualitative comparisons between experimental and numerical data will be then performed.

 

Figure 4. Optical Darmstadt engine setup. Dashed rectangle area illustrates the computational domain for the present study

Figure 5. Darmstadt optical engine: CAD geometry of the computational domain (left), the generated mesh (right)

 

At present, a cold flow (without combustion) LES for Darmstadt optical engine is carried out using 12 processors. Pressure boundary conditions are applied using experimental measurements taken at intake and exhaust. So far from the initial observation, 1 cycle requires only 2.5 days. To carry out a good statistical analysis of any engine a sufficiently large number of engine cycle results is required. A novel parallelization approach based on statistical perturbation method which was introduced by Goryntsev [3, 4] will be adopted for this engine configuration, as well. Using this approach, 50 cycles for a good statistical study in-cylinder flow field can be obtained in 25 days with only 60 processors, and in 12.5 days with 120 processors. The simulated flow field was compared with experimental data and StarCD from Baumann [5]. Phase averaged and rms velocity profiles along lines (see Figure 6 and 9) are shown in Figure 7, 10 for crank angle 90° bTDC and in Figure 8, 11 for crank angle 270° bTDC.

Figure 6. Postions of velocity profiles in radial direction

Figure 7. Radial phase averaged and RMS velocity profiles during compression (90°CA bTDC) Red: KIVA4-MPI-LES; Blue: StarCD; Dot: Experiment

Figure 8. Radial phase averaged and RMS velocity profiles during compression (270°CA bTDC) Red: KIVA4-MPI-LES; Blue: StarCD; Dot: Experiment

Figure 9. Postions of velocity profiles in axial direction

Figure 10. Axial phase averaged and RMS velocity profiles during compression (90°CA bTDC) Red: KIVA4-MPI-LES; Blue: StarCD; Dot: Experiment

Figure 11. Axial phase averaged and RMS velocity profiles during compression (270°CA bTDC) Red: KIVA4-MPI-LES; Blue: StarCD; Dot: Experiment

 

The final step is to simulate the combustion process of biofuel in this Engine. The FGM table is going to be generated.

 

[1] S. Gavrilakis, Numerical simulation of low-Reynolds-number turbulent flow through a straight square duct, Journal of Fluid Mechanics, 244:101-129 (1999)

 

[2] V. Huijnen, L. M. T. Somers, R. S. G. Baert and L. P. H. de Goey, Validation of the LES approach in KIVA-3V on a square duct geometry, Int. J. Numer. Meth. Engng, 00:1-12, (2005)

 

[3] D. Goryntsev, A. Sadiki and J. Janicka, Analysis of Misfire Processes in Realistic Direct Injection Spark Ignition Engine using Multi-Cycle Large Eddy Simulation. Proc. Combust. Inst., 34 pp. 2969-2976 (2003)

[4] D. Goryntsev, A. Sadiki, M. Klein and J. Janicka, Analysis of cyclic variations of liquid fuel-air mixing processes in a realistic DISI IC-engine using Large Eddy Simulation, int. J. Heat Fluid Flow, 31(5) pp. 845-849 (2010)

[5] M. Baumann, F. di Mare and J. Janicka, On the validation of large eddy simulation applied to internal combustion engine flows Part II: Numerical analysis. Flow Turbulence and Combustion 92, 299-317 (2014)

 

 

Key Research Area

Computational Fluid Dynamics, Combustion Modeling: Tabulated Chemistry

Supervisors

Prof. Dr.-Ing. Johannes Janicka - Energie und Kraftwerkstechnik

Prof. Dr. rer. nat. Amsini Sadiki - Energie und Kraftwerkstechnik

Contact

Chao He
Dipl.-Ing.

Address:

Dolivostraße 15

D-64293 Darmstadt

Germany

Phone:

+49 6151 16 - 24388

Fax:

+49 6151 16 - 24404

Office:

S4|10-210

Email:

he (at) gsc.tu...

 Print |  Impressum |  Sitemap |  Search |  Contact |  Privacy Policy
zum Seitenanfangzum Seitenanfang