FEA for stress analysis of the Gear teeth 1

[feajob]

Hi,

I would like to model a Gear. This is my first experience with Gear structure. So, I appreciate to know your experience and any helpful link, ...

Thanks,
AAY

 

[johnhors]

Build finely meshed 2D axisymmetric models, using contact between the male and female teeth. Include non-linear geometry affects and material plasticity. the simplification of a 2D axisymmetric model is obviously you lose the helical geometry, but just try building a 3D model with a sufficiently detailed mesh!

 

[israelkk]

Why are using FEA to analyse a gear? There are formulations and decade of accumulated knowledge on gear design. See AGMA, ISO, DIN or JIS standards.

 

[meshparts]

Another way to reduce the number of DOF's ist to use submodeling and/or substructuring.

Since a gear has cyclic symmetry, you can model one teeth as one superelement and then build the hole gear from superelements. That's substructuring.

With submodeling you mesh the hole gear model with a coarse mesh, then solve it. For the contact region you define a finer mesh and interpolate the displacements from the earlier solution and solve again.

Regards,
Alex

 

[crisb]

mihaiupb........does submodelling/substructuring work with the nonlinearities including contact mentioned by johnhors?

 

[feajob]

Thank you for all of your comments.

israelkk, I will use AGMA as a standard classical method for the stress analysis. But, I would like to do a double check with FEA.

By the way, I have to do fatigue analysis on this structure. I guess that AGMA has some recommendations in this regard. Do you know which AGMA manual talks about fatigue issues?

AAY

 

[meshparts]

crisb,

submodeling work just fine with nonlinear material and contact if the analysis is static.

Superelements must have a linear material property. But Plasticity hapens just at the contact so I would model the one teeth in contact as it is and nonlinear with contact and the other teeths as super elements. But in this case it becomes very complicated... So I would go whith submodeling in the end...

Regards,
Alex

 

[israelkk]

feajob

You can start with any machine design book such as Shigley or Norton. They summarize spur gear calculation procedure for bending and pitting fatigue. From there you can go to AGMA.

 

[paulliu]

If you want to calculate the gear deflections, the
FEA or BEM are the correct methods to use. The AGMA,
ISO, DIN only provide the methods to calculate the
stresses for the involute gears. For the non-involute
gears, one will use these numerical methods to calculate
the stresses too.

In my opinion, the most difficult part of the gear FEA is
to obtain the geometry equations for various parts of
the gear surfaces, which is actually not related to
the FEA. Once you are confident in the geometry equations
of the gears, the gear FEA is not much different from
the general contact program FEA.

But be aware of the fact that someone may question
your results. For example, I've obtained very different
FEA results of the gears from the other two FEA analysts
involved in the same projects.