Off-campus Michigan Tech users: To download campus access theses or dissertations, please use the following button to log in with your Michigan Tech ID and password: log in to proxy server

Non-Michigan Tech users: Please talk to your librarian about requesting this thesis or dissertation through interlibrary loan.

Date of Award


Document Type

Master's report

Degree Name

Master of Science in Mechanical Engineering (MS)

College, School or Department Name

Department of Mechanical Engineering-Engineering Mechanics

First Advisor

Jeffrey D. Nabor


A production Kohler 8.5RES residential stand-by generator set (Genset) was selected as the platform for this study due to its availability, simplicity, and price point. The Genset consists of a spark ignited (SI) two cylinder vee style internal combustion engine (ICE) capable of running natural gas or propane fuel with a 8.5 kW generator connected directly to the engines crankshaft. This allows for electrical load to be applied to the generator which in turn loads the engine without the use of a conventional dynamometer. A water cooled fully adjustable electric resistive load bank allows for easy adjustment to the desired load point. The electrical power generated was measured to determine the ICE output power and calculate the fuel energy to electrical energy conversion efficiency. To allow for control of the engine while testing it was modified from its original carbureted form to a port fuel injected (PFI) configuration and the original fixed spark timing system was removed and replaced with a coil ignition system. An electronic throttle body (ETB) was fitted to allow adjustment to the incoming air flow. The cylinder heads were modified to allow for a production direct inject (DI) fuel injector which used to deliver water to the combustion chamber and an in cylinder pressure transducer for analysis of various combustion parameters.

The genset and test cell were instrumented with low speed and high speed dataacquisition (DAQ) systems to monitor and capture data at the chosen operatingconditions. The high speed data captured by the DAQ was used in conjunction with anear real-time combustion analysis program which calculated and logged combustionparameters and allowed for optimization of spark timing at each test point. Low speed data including fuel consumption, air mass flow rat, water consumption, and electrical power generated along with other engine parameters were monitored and logged as well. The ICE was tested at three different compression ratios (CRs) by changing the pistons and then by removing material from the cylinder head to decrease the clearance volume. The CR that came from the engine supplier was the first to be tested, second a CR in the range of 10:1-11:1 was targeted, and the range of the third CR was 14:1-15:1. The exact values of the CRs tested were calculated once the modifications were complete and volume measurements could be made. The first CR tested was 8.5:1 which is what the engine comes with from the supplier, the second 10.75:1 after changing pistons, and the third 14.3:1 after removing material from the cylinder head.

Baseline data was collected at the 8.5:1 CR using the factory the fuel and ignition system to be used for comparison. Once the fuel, spark, and ETB modifications were complete tests were conducted by varying the load from 0 kW to the maximum attainable load at each test condition in 1 kW increments while targeting a relative air-fuel ratio (lambda, λ) of 1.0 and a speed of 3600 rpm. Using the combustion analysis software the gross indicated mean effective pressure (IMEP) was maximized for each test by varying spark timing. Water was injected into the combustion chamber at water to fuel ratios (WFRs) of 0.38, 1.0, and 1.5 by mass. These WFRs were chosen by the sponsor; the lowest possible WFR was to be tested as well as the 1.0 and 1.5 ratios. The lowest value of 0.38 was determined by testing the mass flow rate of the water injectors at decreasing durations. It was found that at WFRs lower than 0.38 the mass of water injected varied due to the injector’s response properties. The start of injection (SOI) for water was swept from 180 degrees before top dead center (°BTDC) to 40 °BTDC on the compression stroke in 20° increments at each load condition tested. Before water injection tests began, each load point was tested and optimized to obtain baselines to be used for comparison against the water injection results for each CR tested.

For each test performed an analysis was conducted to determine the effects of water injection of net fuel conversion efficiency, coefficient of variation (COV) of IMEP, and heat release rate which are discussed in greater detail later in this paper. Fuel conversion efficiency was used to determine if the water increased or decreased the conversion from fuel energy to mechanical work and quantified how it was impacted. The stability of combustion was determined by using the IMEP coefficient of variance which is common practice in ICE analysis to see how the water effected the variance in IMEP from cycle to cycle. Lastly heat release data was used to determine if the burn rate and ignition delay was impacted with the presence of water. From this data trends were identified and conclusions drawn regarding the overall impact water injection has on combustion.