Many different models exist for internal combustion engines. When designing and optimizing an internal combustion oftentimes key parameters are missing. Commonly no Pressure > Volume diagrams exist. It is the purpose of this dissertation to prove that a simple and accurate model can generate excellent results.
The results of the model were verified using three different engine configurations and found to give accurate results for power and BTE. These theoretical results helped to better understand each engine configuration. The three engine configurations varied in swirl and squish. The engine configurations were all tested using propane as a fuel.
The three engine configurations were based on the same engine with different cylinder heads. The engine was specifically developed for generator usage in the 15>17kW range. The engine was based on a 1.6L four cylinder engine. The engine was configured for low RPM operation, propane as a fuel, minimal friction and ease of data acquisition. The different cylinder heads had varying squish and swirl in order to better understand their relation to BTE. Of the three configurations tested the Super High Swirl engine configuration had a maximum BTE of 37.5%.
This Air > Fuel ratio used was λ=1.66. This engine configuration had 33.9% squish, an average swirl ratio of 3.2 and a compression ratio of 12.7:1. The model was able to capture excellent results between the experimental and theoretical model. The theoretical model was able to capture the additional heat losses from the increased gas velocities associated with squish and swirl, without modifying the heat loss coefficient for each configuration and data set to match experimental results.
EXPERIMENTAL ENGINE DESIGN METHODOLOGY
A number of businesses, both large and small, manufacture LPG powered electrical generator sets (gensets). The gensets are used within the United States and exported for use overseas. They are used to generate electricity in emergencies and for off>grid locations. The internal combustion engine powering the genset is generally a mass >produced automotive engine and not specifically optimized for use in a genset.
This research effort is an investigation into designing a genset engine which could be
fabricated from rebuilt used automobile engines. An example of a rebuilt automobile engine which demonstrates improved efficiency over present production gensets was designed and tested.
The eventual advantage to doing so would be to provide a lower cost more efficient internal combustion engine to power electrical gensets. The principal goal was to obtain the high brake thermal efficiency without utilizing potentially expensive hardware. This research effort converted an automotive 1.6 L engine for use in a genset.
The combustion model used is based on the model developed by Ferguson for the arbitrary heat release of fuel inducted engines (Ferguson,C. 1986, pg. 168>174). The following equations for combustion unless otherwise noted are from Ferguson. The control volume considered for combustion is modeled by equation 2.
The properties of the fluid in the combustion chamber are determined using the subroutines FARG and ECP. These subroutines use state variables of temperature and pressure to determine the properties of the fluid. The energy within the system is split into two sets of constituents, burned and unburned.
Sensitivity Analysis/Variation of Parameters:
A Sensitivity analysis was carried out to determine which parameters had the greatest influence on IMEP. All parameters were individually increased by 10% to determine their influence.The parameters with the greatest influence on IMEP were pressure, temperature, engine surface temperature, burn duration, start of burn, equivalence ratio, and residual fraction.
Squish and swirl velocities when varied by small amounts have little influence in the model. Their increases velocities attribute only to heat losses. The increased mixing from the induced turbulence does not decrease the time losses that would be realized from the shortened burn duration.
The experimental data was collected in the University of Miami Internal Combustion engines lab. Three different cylinder head configurations were used for testing. The first generated 18.4% squish and generated low amounts of swirl. The second head also generated 18.4 % squish but generated high swirl.
The third head generated extremely high swirl and 33.9% squish. All heads were based on the 1.3L head. Using the smaller head allowed for smaller port diameters which allowed for more swirl to be generated from the increased port velocity.
The compression ratios used were 12.2:1 for the low swirl, and high swirl configurations. The super high swirl configuration used a compression ratio of 12.7:1. All configurations used flat top pistons, two valves per cylinder and a hemispherical combustion chamber.
- From the experimental testing it is evident that the squish areas and swirl ratios used in automobile engines designed for gasoline operation are not sufficient for lean burn LPG fueled genset engines.
- Simply increasing compression ratios to 12.2:1 or 12.7:1 on stock gasoline engines designed for stoichiometric operation does not increase flame speed adequately for lean burn LPG genset engines.
- When increasing the swirl ratio and squish area appreciably and running rich mixtures of LPG (λ≤1.25) large heat losses occur.
- The improvements in BTE do not require the used of fuel injection or any other types of costly technology. The use of a pre>chamber between the gas mixer and intake manifold runners can provide sufficient mixing to virtually eliminate cylinder to cylinder fuel distribution problems and still require only 2.8 kPa intake manifold vacuum at 1800 rpm wide>open throttle.
- These improvements do not require using welding to add material to any engine component but do require CNC’d combustion chambers.
- Valve spring pressure used in automobile gasoline engines can be reduced at least 50% at all valve lifts without adverse effects.
- The changes in valve springs reduced FMEP approximately 3 to 4 kPa at 1800 rpm for this application.
- It is possible to rebuild a 1.6L genset engine using nonstandard but similar cost parts producing an increase in BTE to in excess of 37.0%. This represents an improvement in fuel consumption of over 30% compare d to the engine rebuilt utilizing standard parts.
- The computer model of the three configurations of the engine evolution showed the third engine approaching optimization. This can be seen since the heat losses to the combustion chamber for the three engines were reduced from 20.1% (LS) to 16.7% (HS) and to 16.3% (SHS). Further increases in mixture turbulence will not likely result in decreased heat losses to the combustion chamber.
Source: University of Miami
Author: Patrick Seemann