George Carver Davis

Diluents and Lean Mixture Combustion Modeling for SI Engines with a Quasi-Dimensional Model

Lean mixture combustion might be an important feature in the next generation of SI engines, while diluents have already played a key role in the reductions of emissions and fuel consumption. Lean burning modeling is even more important for engine modeling tools which are sometimes used for new engine development. The effect of flame strain on flame speed is believed to be significant, especially under lean mixture conditions. Current quasi-dimensional engine models usually do not include flame strain effects and tend to predict burn rate which is too high under lean burn conditions. An attempt was made to model flame strain effects in quasi-dimensional SI engine models. The Ford model GESIM was used as the platform. A new strain rate model was developed with the Lewis number effect included. A 2.5L V6 4-valve engine and 4.6L V8 2-valve modular engine were used to validate the modified turbulent entrainment combustion model in GESIM. Results showed that the current GESIM can differ by as much as 10 crank angle degrees compared with test data. The modified GESIM can predict burn duration to within 1--2 CA of experimental data, which is considered very good for engine models.

The intensity of knock in an internal combustion engine: An experimental and modeling study

Experimental data have been obtained that characterize knock occurrence times and knock intensities in a spark ignition engine operating on indolene and 91 primary reference fuel, as spark timing and inlet temperature were varied. Individual, in-cylinder pressure histories measured under knocking conditions were conditioned and averaged to obtain representative pressure traces. These averaged pressure histories were used as input to a reduced and detailed chemical kinetic model. The time derivative of CO concentration and temperature were correlated with the measured knock intensity and percent cycles knocking. The goal was to evaluate the potential of using homogeneous, chemical kinetic models as predictive tools for knock intensity.

COMBUSTION CHAMBER EFFECTS ON BURN RATES IN A HIGH SWIRL SPARK IGNITION ENGINE

Experimental measurements of burn rates have been carried out in a single cylinger homogeneous charge engine. Three different combustion chambers were investigated (75 % and 60 % squish bowl-in-piston chambers and a disk chamber) using a cylinder head with a swirl producing intake port and near central spark location. Data were obtained with each combustion chamber as a function of spark timing, EGR, and load at 1500 RPM. The combustion rate is strongly influenced by chamber shape. The 10-90 % burn durations of the 75 % and 60 % squish chambers are respectively about 40 % and 60 % that of the disk chamber. Chamber configuration had less effect on 0-10 % burn duration. The disk had about 25 % longer 0-10 % burn time than the bowl-in-piston chambers. Modifications to the GESIM model enabled good overall agreement between predictions and experimental data, a rather severe test of the model because the coupling of fluid mechanics, combustion and chamber geometry must be properly modeled. An improved basic understanding of the influence of combustion chamber shape on burn rate has been achieved through the interactive use of experimental data and modeling. The results suggest that differences in turbulence intensity and flame area development due to changes in chamber shape are responsible for the observed burn rate differences.

THE EFFECT OF INTAKE VALVE LIFT ON TURBULENCE INTENSITY AND BURNRATE IN S.I. ENGINES-MODEL VERSUS EXPERIMENT

An Engine Simulation Model was used to study the effect of changing the maximum intake valve lift to control in-cylinder turbulence intensity and burn rate. Experimental measurements of burn rate for two different valve lift profiles were obtained and compared with predictions. The standard K-epsilon turbulence model was found to be inadequate for predicting the proper behavior of turbulence level during compression and expansion. Further investigation showed that the dissipation of turbulence calculated by the standard k-epsilon model was inadequate, thus causing the turbulence levels and burn rates to be approximately independent of the intake valve lift. A new turbulent dissipation model is proposed which uses the eddy angular momentum to scale the dissipation constant. Turbulent intensity predictions from this model resulted in acceptable agreement between the measured and predicted burn rates as the intake valve lift was changed. The effect of throttling the engine using intake valve lift was investigated and predictions made of turbulence intensity, burn rate, combustion efficiency and brake specific fuel consumption (BSFC) as a function of air-fuel ratio and load. Results showed a significant reduction in BSFC at 13 BMEP, 1500 RPM when conventional throttling was compared with intake valve throttling at equal burn rates. In addition, the effect of B/S ratio on turbulence intensity, burn rate and ISFC was investigated and results showed the independent effects of engine geometry and turbulence on burn rate when the engine was stroked holding the cylinder bore constant.

The Effects of Load Control with Port Throttling at Idle- Measurements and Analyses

This paper describes an experimental and analytical study conducted to investigate the effects of load control with port throttling on stability and fuel consumption at idle. With port throttling, the pressure in the intake port increases during the valve-closed period due to flow past the throttle. If the pressure in the port recovers to ambient before the valve overlap period, back flow into the intake system from the cylinder is eliminated. This allows increased valve overlap to be used without increasing the residual mass fraction in the cylinder.

Modeling the effect of swirl on turbulence intensity and burn rate in S. I. engines and comparison with experiment

An Engine Simulation Model was used to study the effect of in-cylinder swirl level on turbulence intensity and burn rate while holding the inducted kinetic energy constant. Experimental measurements of burn rate for three different swirl levels were obtained and compared with model predictions. The turbulence model used previously did not include wall shear effects and showed little enhancement of turbulence due to swirl, causing small changes in predicted burn rate when the swirl level was changed. An improved turbulence model is proposed which includes production of turbulence due to wall shear effects. Turbulence intensity predictions from the improved model resulted in excellent agreement between the measured and predicted burn rates as swirl level was changed. In addition, the model was used to predict the effect of swirl levels on ISFC. Results showed that ISFC changes were overall small for the range of swirl levels considered.

Effects of Intake Port Design and Valve Lift on In-Cylinder Flow and Burnrate

LDA measurements of the flow in a motored engine near TDC of compression have been obtained, along with burnrate data in a firing engine having a near-central spark plug location. Results are reported for two different intake ports and four intake valve lifts varying from 25% to 100% of full lift. Opposite trends of swirl vs valve lift were found for the two ports, and the rms velocity fluctuation was found to be relatively insensitive to changes in valve lift. Regression analysis of the burn duration data was conducted, with swirl ratio and rms as independent variables. The analysis indicated that burn duration decreases with an increase in swirl ratio and/or rms velocity fluctuation. In light of the experimental findings, a new conceptual model is proposed regarding the effect of valve lift on the dissipation of turbulent velocity via changes in the length scale. And combustion-induced shear in the unburned gas resulting from conservation of angular momentum is hypothesized as a possible mechanism for increased burnrate due to swirl for the case of central ignition.

Application of heuristic and machine-learning approach to engine model calibration

Automation of engine model calibration procedures is a very challenging task because (1) the calibration process searches for a goal state in a huge, continuous state space, (2) calibration is often a lengthy and frustrating task because of complicated mutual interference among the target parameters, and (3) the calibration problem is heuristic by nature, and often heuristic knowledge for constraining a search cannot be easily acquired from domain experts. A combined heuristic and machine learning approach has, therefore, been adopted to improve the efficiency of model calibration. We developed an intelligent calibration program called ICALIB. It has been used on a daily basis for engine model applications, and has reduced the time required for model calibrations from many hours to a few minutes on average. In this paper, we describe the heuristic control strategies employed in ICALIB such as a hill-climbing search based on a state distance estimation function, incremental problem solution refinement by using a dynamic tolerance window, and calibration target parameter ordering for guiding the search. In addition, we present the application of a machine learning program called GID3* for automatic acquisition of heuristic rules for ordering target parameters.

Rapid integration of CAE analysis programs using a blackboard approach

This paper presents an approach for rapidly integrating CAE analysis programs into complex engineering methodologies. The blackboard technique has been adopted in the implementation of this approach, especially for the process scheduling, monitoring, and potentially, diagnosis. This approach has now been implemented and tested in prototyping projects on several engineering methodologies involving multiple analysis programs.<<ETX>>