Spatial and Temporal Stress Distributions in Spark Ignition Engine Pistons at Wide Open Throttle

Spatial and temporal stress distributions in two piston geometries for a sinusoidal piston motion engine with different rigid piston-connecting rod attachments were investigated; one was a homogeneous, all-aluminum piston with a bolted connection located remote to the piston crown, and the second was a composite piston structure in which a steel connecting rod is imbedded into the aluminum piston crown. The piston stress response to the applied thermo-mechanical loads was predicted using LUSAS (London University Stress Analysis System). A dynamic finite element analysis of the homogeneous piston was performed for the duration of one complete combustion cycle and a static analysis modeled the two component piston with loads estimated at the time when stress was highest. Two types of thermo-mechanical fatigue were analyzed: low cycle high stress and high cycle low stress. Nodal stress histories at 2000 rpm are presented for equivalent Huber (stress fields equivalent to the three dimensional stress state), radial, hoop, longitudinal, and shear stresses. Specific locations of the maximum compressive and tensile stresses were identified and show that maximum stress occurs at different times for different locations. At nodal locations where mechanical loads counteract thermal expansion forces, stress levels peak at the time when gas pressure is low. Stresses peak at the time of maximum pressure where thermal and mechanical loads have the same sense. The principle of superposition was used to differentiate the thermal and mechanical stress contributions in the piston and, most notably, in the piston crown region. The maximum amplitude and frequency of the thermal and mechanical-thermal stresses indicated high cycle fatigue failure was not likely. However, high compressive stress developed in material weakened by high temperatures is the most likely cause of failure. The two component piston is formed with a steel cone insert cast into the piston. The static results show high local Huber (von Mises) stresses at the dissimilar material interface and the highest value of hoop stress and, consequently, the main cause of large equivalent stress levels is in the steel cone insert. A modified flexible tripod shaped cone resulted in significant reduction of hoop stress in the region of contact with the aluminum. However, the peak equivalent stress present in the aluminum crown of this model was too high for the material to withstand. The last results presented are for the homogeneous piston modified to lower stress concentrations and with boundary conditions modified to represent enhanced cooling of the piston underside. The peak piston temperatures were significantly reduced and consequently the Huber stress levels were reduced. Analysis of the dynamic stresses indicated a low probability for fatigue failure. These results indicated that a homogeneous aluminum piston could become a feasible concept, provided additional piston cooling mechanisms are installed.