Access Type

Open Access Dissertation

Date of Award

January 2013

Degree Type

Dissertation

Degree Name

Ph.D.

Department

Mechanical Engineering

First Advisor

Marcis Jansons

Abstract

A fundamental understanding of advanced compression ignition combustion is requisite to meet the simultaneous challenges of stringent fuel efficiency and emission standards. Single zone simulation shows that nitrogen oxide (NOx) production occurs in a high temperature region and soot production in a high equivalence ratio region within a specific temperature window. Combustion temperature, therefore, is a crucial variable that determines soot and NOx emissions under various combustion modes, and it is thus very important to have the capability to quantify this parameter in-cylinder. Optical diagnostic techniques such as -- Rayleigh scattering, filtered Raman scattering (Doppler), Raman scattering, coherent anti-stokes Raman spectroscopy (CARS), vibrational thermally-assisted fluorescence, two-line atomic fluorescence, two-line molecular fluorescence, chemiluminescence emission, absorption thermometry and soot two-color thermometry -- commonly provide the sole means of non-intrusively investigating flame temperature. However, significant challenges of diesel combustion, such as temporal and spatial flame heterogeneity, interference from particle-scattering, background combustion illumination, and the limitations imposed by laser repetition rates hinder the utilization of most of these diagnostics. The final technique, soot two-color thermometry, is less affected by the above factors, and therefore, has been widely utilized in diesel engine studies. This work presents an approach to implement the classic soot classic two-color thermometry technique on a high-speed digital color camera to realize crank-angle-resolved, spatial distribution of in-cylinder combustion temperature. A comparison is made between high-speed two-color thermometry measurements and reacting flow simulations in an engine application fueled with ULSD. The results show the simulations over-estimate soot temperature by 10%-20% in most crank angle intervals. To improve the two-color thermometry technique, a soot emissivity model is developed by incorporating soot optical properties and CARS and laser-induced incandescence (LII) data obtained from a rich (Φ=2.1), C_2 H_4 / Air premixed flat calibration flame. With the flame-calibrated soot emissivity model, the in-cylinder soot temperature error between simulation and experiment decreases to ±5% implying an improvement of the soot two-color thermometry technique. The limitations of line-of-sight two-color thermometry are investigated by comparing the in-cylinder soot optical thickness KL with simultaneous soot laser-induced incandescence (LII) measurements in an optical engine. The results exhibit significant spatial differences, implying temperature gradient along the line of sight.

Included in

Engineering Commons

Share

COinS