Another study used FEA with applied loads including bolt tightening load, piston pin interfer- ence load, compressive gas load, and tensile inertia load [4]。 On the basis of the stress and strain measurements performed on connecting rod, close agreement was found with loads predicted by inertia theory。 The study indicated that stresses in a con- necting rod due to bending loads are substantial, and that buckling and bending stiffness are import- ant design factors that must be taken into account during the design process。

Balasubramaniam et al。 [1] used the various indi-

vidual loads acting on connecting rod for performing simulation and obtaining stress distribution by superposition。 The loading consisted of inertia load, firing load, press fit of the bearing shell, and bolt forces。 Athavale and Sajanpawar [5] also modelled the inertia load in their finite element (FE) model。 An interface software was developed to apply the acceleration load to elements on the  connecting rod depending upon their location, as acceleration varies in magnitude and direction with location on the connecting rod。 They fixed the ends of the con- necting rod to determine its deflection and stresses。 This, however, may not be representative of the pinned joints that exist in a connecting rod。 The con- necting rod was separately analysed for tensile load due to piston assembly mass (piston inertia), and for compressive load due to gas pressure。 The effect of inertia load due to mass of the connecting rod was also analysed separately。 Pai [6] presented an approach to optimize the shape of the connecting rod subjected to a load cycle consisting of inertia load deduced from gas load as one extreme, and peak inertia load exerted by piston assembly mass as the other extreme and used fatigue life as the optimization constraint。

In this study, a detailed load analysis under service

steel connecting rod,  followed  by quasi-dynamic FEA to capture stress variation over a cycle of oper- ation。 Such stress analysis under realistic operating loads is critical to any durability or  optimization study of a connecting rod and  is  vastly  different from the typical uniaxial testing and static analysis commonly conducted for this component。 This is because, in a typical static analysis, the loads  acting at the two ends of the connecting rod are equal in magnitude and are in static equilibrium。  On  the other hand, in  a  quasi-dynamic  analysis,  the loads at the two ends need not be equal and the connecting rod is in equilibrium at any instant in time, only when the inertia load resulting from angular velocity and acceleration (both translational and angular) are accounted for。 Therefore, although the  quasi- dynamic analyses are repeated at different time points, they are based  on  time-varying dynamic input data。 For this  reason,  the  analysis  is referred to as ‘quasi-dynamic’。 Details of the dynamic load analysis are discussed in the next section。 Optimiz- ation aspects and fatigue behaviour of the connecting rod are investigated in references [7, 8], respectively。 In this article, FE modelling aspects, resulting stress-time histories, variation of stress  ratio, presence of mean and bending stresses, and multi- axiality of stress states in various locations of the connecting rod under service  operating  conditions are discussed。 A comparison is also made between results obtained using static FEA commonly per- formed, and results using quasi-dynamic FEA   repre-

senting more realistic service operating conditions。

2    LOAD ANALYSIS

Engine configuration to which the typical connecting rod investigated belongs is shown in Table 1 and piston pressure versus crank angle diagram used in the analysis is shown in Fig。 1。 With these data and using commercial software such as ADAMS and I-DEAS, angular velocity and angular acceleration of the connecting rod, as well as linear accelerations of the connecting rod crank end center and of the center of gravity (CG) can be obtained。 Variations of these quantities during one engine cycle can    be