Because of too many parameters which affect the lubrication performances, it is hard to gain the comprehensive information, to overcome this difficult, many experi- mental runs need to carry out to evaluate the bearing performance。 In order to reduce the number of experimental runs, the design of experiment method is researched recently。 Francisco [8] used design of experiments to analyze the connecting rod big-end bearing behavior, and the main objective of the present work is to identify the factors dominating the bearing behavior。 Smith [9] optimized the design of a piston-ring pack using design of experiment (DOE) methods。 It is shown that an improved design can be achieved that reduces ring losses by 57% whilst reducing upward oil flow by 39%。 Johansson [10] used the experiment to evaluate   cylinder

http://dx。doi。org/10。1016/j。asoc。2015。01。009

1568-4946/© 2015 Elsevier B。V。 All rights reserved。

2。Theory

2。1。Lubrication model

2。1。1。Governing equations

EHD lubrication analysis plays an important role in the design of dynamically loaded main bearings as it can offer more realis- tic prediction of the bearing performances。 Apart from the bearing deformation, oil film cavitation is also very important to the bearing performance prediction。 The earlier researchers Elrod and Adams [16], Vijayaraghavan and Keith [17] found cavitation phenomenon and established the cavitation model, then cavitation is researched widely。 Boedo and Booker [18,19] investigated the effect of body force deformation and mass-conserving cavitation on the EHD behavior of connecting rod big-end bearings。 Bonneau [20] and Optasanu [21] used a mass-conservative algorithm to simulate the EHD bearing behavior。 The problem is solved analytically using Reynold’s boundary conditions for film rupture。 The governing equation concludes the full film region and cavitation region is rewritten as:

where $x, $z is the pressure flow factors along x, z direction, $s is the shear flow factor, o is the composite rms roughness, &∗ is the filling factor。

2。1。2。Oil film thickness

The effect of elastic displacements of the bearing surface has to be included in EHD model。 The film thickness including this effect is written as:

liner/piston ring contact friction, it is shown that for the introduced DOE based tri- bometer test the interaction of dynamic viscosity, velocity and contact pressure can be studied within one experiment。

Apart from the researches about DOE, many researchers started to use the DOE to combine with intelligent algorithm to analyze the problem。 Ko [11] applied artificial neural network and Taguchi method to preform design in metal form- ing considering workability。 Papadopoulos [12] combined the experimental design with artificial neural networks to determine the chlorinated compounds in fish using matrix solid-phase dispersion, and the experimental results demonstrated that the proposed soft computing strategy is very effective and efficient to achieve satis- factory results。 Benardos [13] Predicted the surface roughness in CNC face milling using neural networks and Taguchi’s design of experiments, the data used for the training and checking of the networks’ performance derived from experiments con- ducted on a CNC milling machine according to the principles of Taguchi DoE method。 Hao [14] analyze the parameter sensitivity on deformation of composite soil-nailed wall using artificial neural networks and orthogonal experiment, 25 sets of tests are designed to analyze the sensitivity of factors affecting the maximum lateral

h(ˇ, z) = c − εx(y) cos ˇ − εz (y) sin ˇ + ı(ˇ∗ , z) (2)

where c represents the original radical clearance at the under- formed state, εx is the crankshaft deformation in x direction, εz is the crankshaft deformation in the z direction, ı(ˇ∗, z) is the radi- cal deformation of the bearing, ˇ∗ is measured from bearing crown,