Section 2 Model requirements

2.1 General

2.1.1 The basic recommendations and perceptions for building the FE-model are presented in Ch 1, 2.2 Element mesh recommendations. It is obligatory for the final FE-model to fulfill the requirement in Ch 1, 2.4 Element mesh quality criteria.

2.2 Element mesh recommendations

2.2.1 In order to fulfil the mesh quality criteria it is advised that the FE model for the evaluation of Stress Concentration Factors be constructed according to the following recommendations:

The model consists of one complete crank, from the main bearing centreline to the opposite side main bearing centreline.
Element types used in the vicinity of the fillets:
- 10 node tetrahedral elements;
- 8 node hexahedral elements;
- 20 node hexahedral elements.
Mesh properties in fillet radii. The following applies to ±90 degrees in circumferential direction from the crank plane.
Maximum element size a=r/4 through the entire fillet as well as in the circumferential direction. When using 20 node hexahedral elements, the element size in the circumferential direction may be extended up to 5a. In the case of multi-radii fillet, r is the local fillet radius. (If 8 node hexahedral elements are used even smaller element size is required to meet the quality criteria.)
Recommended manner for element size in fillet depth direction:
- First layer thickness equal to element size of a;
- Second layer thickness equal to element size of 2a;
- Third layer thickness equal to element size of 3a.
Minimum of six elements across web thickness.
Generally, the rest of the crank should be suitable for numeric stability of the solver.
Counterweights only have to be modeled when influencing the global stiffness of the crank significantly.
Modelling of oil drillings is not necessary as long as the influence on global stiffness is negligible and the proximity to the fillet is more than 2r, see Figure 1.2.1 Oil bore proximity to fillet.
Drillings and holes for weight reduction have to be modelled.
Sub-modelling may be used as far as the software requirements are fulfilled.

Figure 1.2.1 Oil bore proximity to fillet

2.3 Material

2.3.1 The Rules do not consider material properties such as Young’s Modulus (E) and Poisson’s ratio (ν ). In FE analysis those material parameters are required, as strain is primarily calculated and stress is derived from strain using the Young’s Modulus and Poisson’s ratio. Reliable values for material parameters have to be used, either as quoted in literature or as measured on representative material samples.

2.3.2 For steel the following is advised:

2.4 Element mesh quality criteria

2.4.1 If the actual element mesh does not fulfil any of the following criteria at the examined area for SCF evaluation, then a second calculation with a refined mesh is to be performed.

2.4.2 Principal stresses criterion: The quality of the mesh should be assured by checking the stress component normal to the surface of the fillet radius. Ideally, this stress should be zero. With principal stresses σ₁ , σ₂ and σ₃ , the following criterion is required:

2.4.3 Averaged/unaveraged stresses criterion: The criterion is based on observing the discontinuity of stress results over elements at the fillet for the calculation of SCF:

Unaveraged nodal stress results calculated from each element connected to a node i should differ less than by 5 per cent from the 100 per cent averaged nodal stress results at this node i at the examined location.