Finding mesh cycle stiffness in spur gear pair

[Stiffie]

Hello skilled engineers,

I'm trying to generate a plot of mesh stiffness in a spur gear pair as they rotate.

I've tried to do this using a static analysis by applying a moment T on the pinion while the gear rotates. Im measuring the rotation in both the pinion and gear, and the transmission error would be

TE = perfect rotation - measured rotation

and stiffness

K = T / TE

Am i right? I was nontheless hoping that this would give a periodic square plot showing the stiffness as different number of teeth comes into contact during the rotation.

I'm using surface to surface contact with penalty 0.2 and CPS8R elements. What have i not understood correctly?

 

[DrBwts]

Static analysis if a dynamic system?

 

[Stiffie]

Hello DrBwts,

thank you for your interest.

As far as I understand, neglecting dynamic effects and looking quasi staticly at it should give reasonable results. The overall stiffness in the gear pair changes as different number of teeth come into contact and due to the points of contact moving along the teeth working flank.

 

[DrBwts]

I guess it depends on how you are setting up the steps in the model?

When you say that you are applying a moment T on the pinion while the gear rotates, how exactly are you doing this?

 

[Stiffie]

I have a reference point in each gear center which is connected to the gears with kinematic coupling. A moment is acting on one of the reference points and a displacement is applied to the other one, making the gears acting against eachother. I thought that dividing the reaction moment by the difference in angle i would get the stiffness at each increment. This is working to some extent, but the stiffness is varying as the elements rolls over eachother..

I've tried both linear and quadratic elements. The linear are more stable, but i guess i need a heavily refined mesh in the contact areas?