Week 2 Bevel Gear Challenge

Grid Dependency Test

Mesh refinement level is most common question that we come across while doing finite element analysis. Since coarse mesh size will result in inaccurate result and fine mesh size will require large amount of computer memory and long run time as well.

So simple way is that refine the size of mesh till there is no significant change in result plot like stress generated and this is called Grid Dependency Test. Refine mesh till result does not change significantly and does not depends upon mesh or grid.

They are many mesh refinement methods like reducing mesh size, increasing element polynomial order etc.

Grid Dependency test for the Bevel gear simulation

Parameter for Grid Dependency test for the Bevel gear simulation Analysis 

1. Analysis System: Static Structural 

2. Mesh Type: Tetrahedral Mesh

3. Material: Structural Steel ( Material is constant in all three cases)

    Properties 

    i)  Density - 7850 Kg/m^3

    ii) Young Modulus - 200 Gpa

   iii) Yeild Strength - 250 Mpa

   iv) Poissons Ratio - 0.3

Setup paramater for Analysis of Bevel Gear

1. Contact between gears

   i) Type: Frictional

   ii) Behavior: Symmetrical

   iii) Formulation: Augmented Lagrange

   iv) Interface Treatment: Adjust to Touch

2. Revolute Joint is defined on both gears

3. Joint moment load of 100 Nmm on large Gear

4. Joint rotation of 45 degree on small gear on 3 steps of 15 degree each.

Comparsion table for number of element and nodes for three cases

Cases	6mm	5mm	4mm
Number of Elements	13980	13977	15175
Number of Nodes	26071	26193	28202

Results

Simulation Result for different mesh size is shown below in order 6mm, 5mm and 4mm respectively.

Cases	6mm
Equivalent stress
Total Deformation
Equivalent Elastic Strain

Case	5mm
Equivalent stress
Total Deformation
Equivalent Elastic Strain

case	4mm
Equivalent stress
Total Deformation
Equivalent Elastic Strain

Conclusion:

1. From above result we can conclude that after performing the grid test on bevel gear by varying global mesh size, we achieve much accurate results.
2. From above result table is notice that equivalent maximum stress for Design point 1 to design point 2 is increase by almost most 14 percent and from design point 2 to design point 3 is increase by only 1.5 percent.
3. It is clear that from above table that equivalent maximum stress value is getting accurate and stable as mesh gets fine. And it indicates importance of grid dependency test while doing analysis.