Learning Forum

[srujana]

Greetings, 

I am doing static structural analysis of a sector of a compressor impeller, with rotational velocity applied as inertial load. cyclic symmetry is applied, the hub faces are arrested with the following constraints, 

Remote displacement - x = 0; y = 0; z = 0; rotx = 0; roty = 0; rotz = free;

Remote displacement 2 - x = 0; y = 0; z = 0; rotx = 0; roty = 0; rotz = 0; 

Normal stress in rotating axis is obtained to get centrifugal stresses in the impeller. 

Is this method right? when trying to validate with analytical method (approximated as a disc) using hoop stress formula (σ = (ω^2r^2ρ)/ 3), the results do not match.

Analytically, for a radius of 52mm, 100000 rpm, & 2850kg/m^3 desnity, average hoop stress is 280MPa. 

Is there something fundamentally wrong with the method followed. 

Any insight will be helpful, thank you. 

 

 

[peteroznewman]

It is fundamentally wrong to use two remote displacements to implement cyclic symmetry boundary conditions.

Insert a Symmetry object into the Model tree and follow the directions in the first few minutes of this video. https://www.youtube.com/watch?v=EBa7_R6cVxc

[srujana]

Got it, I just ran the same with fixed condition (as shown in vedio). However, the results still dont match the analytical results (attached below)  

Cyclic symmetry: 

 

Boundary conditions:  

Equivalent stress:  

Normal stress in y (centrifugal stress): 

 

I tried an ansys tutorial as well, Chapter 14: Static and Modal Analysis of a Compressor Model with 4 Axial Stages 

The stress values here also do not match with analytical. Is the method of viewing the results correct? Please suggest if there is any reference for centrifugal stress in turbomachinary components that I can go through. Thank you. 

 

[peteroznewman]

If you want to compare with the analytical hoop stress result, you need to use the analytical geometry.

This paper provides examples of how the build a finite element model of the analytical geometry.

[srujana]

Im a bit confused, which stress does the 3D model analysis represent then. 

[peteroznewman]

The 3D model of the impeller has the stress for that geometry. If you take away the impeller blades, that changes the stress. If you flatten the shape to a flat disk, that changes the stress.

[srujana]

Got it, thank you for the support.