Minimizing NO and BSFC using Optimization feature in GT-Power
OBJECTIVE:
Explore Example- Diesel_4cyl_DIpulse_injratemap and write a detailed report
- Parameters-SOI, Rail Pressure
- Response- NO and bsfc minimum
- Run all designs for a case and find the optimum values
- Comment on your observations
Geometry Model
The engine is a 4 cylinder, 4 stroke, 4 valves, direct injection, 2L diesel engine. Several modeling concepts like Variable Geometry Turbine, Boost Control, intercooler, EGR circuit with cooler and EGR rate controller, injection limiting for smoke control, and exhaust after-treatment device are included in the model.
With End environment, air enters with 1 bar pressure and collects in the airbox which further passes through the compressor, passing through compressor makes the air pressure and temperature high. Further, an intercooler is attached to decrease the temperature of compressed air, increasing the density of the air and hence more air can be forced into the combustion chamber. An EGR is used where one port of one cylinder is used to circulate the EGR which again passes through the EGR cooler to lessen the high temperatures of exhaust gas. An EGR controller is implemented to control and maintain the required EGR according to the requirement which controls the throttle opening, ultimately controlling the pressure drop across the EGR system.
A Variable Geometry Turbine is used, as we see the exhaust connection is connected with the turbine which rotates the compressor with the connected shaft. After the expansion of the turbine the remaining exhaust gas passes through the After treatment, Muffler to reduce the noise and vibrations and then to the environment.
EGR Cooler
Intercooler
After-Treatment
EGR-Control
A Predictive combustion model 'DIPulse' is used in the model. CylHeatTR as Heat Transfer Model.
As we see, it is a square engine with Bore * stroke 86*86 mm and a connecting rod length of 129mm.
A Bosch Injector is used with predefined injection rate shape. Various attributes of nozzles are defined in the below image.
Here, we are tried to minimize the BSFC and NO exhaust which are a tradeoff for each other using Optimization feature.
An Integrated Design Optimizer is used, Multi-Objective Pareto and for case handling, Optimize each case independently is enabled. For Algorithm, a Genetic Algorithm is been used. Other options GT-Power provided are discrete-grid, Brent-method, Simplex, CMA-ES. Discrete-grid is used when there is a single parameter to be optimized. Genetic Algorithm is recommended for all problems having medium to high complexity, which would result from many factors, the presence of one or more constraints, and non-linear characteristics
The Factors we have considered here are Start of Injection (SOI) and Rail Pressure. For SOI, the Lower limit is -10 CAD and upper limit kept is 10 CAD. While for Pressure rail, the lower limit kept is 1200 bar and the upper limit kept is 2000 bar.
Our targets, bsfc and NO are been selected as the responses. i.e. the algorithm would be finding all the optimized cases where bsfc and NO emissions would be as low as possible depending or changing the values for SOI and rail pressure within the specified range.
Simulation Results
A case was run for 1500 rpm with 13bar target BMEP. Below plot is a Pareto plot of NO vs bsfc which shows us a trade-off relation between both the parameters i.e. as we increase the bsfc value, the NO emission value decreases and vice versa. Blue colour points suggest the frontal Pareto i.e. the most optimized case that the algorithm has achieved. A total of 150 iterations were run.
Here minimization of bsfc was the target and we some cases ran on the higher side above 245 g/kW-hr, but as soon as it realizes the algorithm lowers down the bsfc below and around 235 g/kW-hr.
Here minimization of NO was the target and we some cases ran on the higher side above 700ppm, but as soon as it realizes the algorithm lowers down the NO below 500ppm.
For lessening down the NO ppm level, we see a delay shift of the start of injection after TDC. The allowed range was kept from 10 deg btdc to 10 deg atdc.
The range kept for the Rail Pressure was kept from 1200 bar to 2000 bar. We see an automatic reduction in rail pressure to lower down the NO ppm emission levels.
The contour plots for bsfc and NO with respect to Rail Pressure and the Start of Injection. Here we the minimum rail pressure is obtained between the range of 0 to 5 atdc and below 1500 bar P_Rail. But for NO, we get the lowest NO of 302 ppm in the region between 5 to 10 atdc cad.