Learning Forum

[DrShadow]

Hello,

I am modeling forced oscillations problem (blade is being driven by actuator) and it looks to me the natural way to do this is to run harmonic response analysis. My blade model has 3 points of support, rotation is allowed around 2 points and third support is connected to actuator. The blade is supposed to have zero degrees of freedom. I only modeling the blade with no actuator, my intention is to model actuation through remote force boundary conditions. But, if I am using remote force BC, frequency response does not match modes of 3-points supported model, especially for the first mode. Frequency response actually matches modes of model supported at 2 points, which means the blade gets extra degree of freedom. I believe to have zero degrees of freedom and to have harmonic excitation at the same location are mutually exclusive requirements, so is the problem stated incorrectly or am I missing something?

I looked at ANSYS manual and I can see that force behavior can be specified as rigid, deformable or coupled but I cannot figure it out if coupled behavior is the right one. Frequency response with coupled remote force formulation still does not look right to me. Any help is appreciated.

Thank you

 

[peteroznewman]

Hello DrShadow,

I understand a zero degree of freedom connection to a blade would have one point with three translations defined, a second point with two translations defined that are orthogonal to the line connecting points one and two , and a third point, off the line along the first two points, that defines just one translation, the one that controls rotation about the first two points.  Is that what you have?

When you say the third support is connected to an actuator, do you mean there is a linear motion of that actuator and the blade is connected in all six degrees of freedom, or that it is pinned to the actuator with freedom to rotate?

It would help to have an image of the blade and three support points. 

Regards,

Peter

 

[DrShadow]

Hello Peter,

axis of rotation of the blade model goes through supports A-B, no translational motion is allowed. Support C is linked to actuator, and linear motion of actuator (in Z-direction) translates to rotation of the blade around axis A-B. To me actual geometry is not even important, what's important the blade has zero degrees of freedom. And if I model actuator with remote force boundary conditions (at point C) the blade gets one degree of freedom: rotation around axis A-B. At least harmonic response looks like it, as there is no response to harmonic excitation at the frequencies where first mode should be. Thank you.

Regards,

Alex

 

[peteroznewman]

Hello Alex,

Thank you for the image, that helps me think about the model.

I understand now that if you support the blade just on the A-B axis, it has one degree of freedom, and that rigid body mode may disturb the harmonic response from a harmonic force input at C.

Is the forced oscillation model intended to simulate a static blade hanging in earth gravity, or do you want to simulate a spinning blade that has a constant angular velocity?

A Transient Structural model would yield some information about the response to harmonic force inputs at C, but that model will take a long time to solve and generate a lot of data that has to be post-processed to get good summary information.

I have not done a Rotordynamics kind of analysis, which could be much more efficient at getting some information, so I hope some other members will post their comments.

Regards,

Peter

[DrShadow]

Hello Peter,

thank you for finding time to help me. We are studying vibrating blade and flow interaction, blade is not spinning (yet).

I am thinking about remote displacement boundary conditions, are these boundary conditions different from remote forcing? I can calculate frequency response per unit force (from reaction force), and amplitude is actually exactly the same as response to remote forcing, but at least I have 180 degrees phase change at first mode.

Thank you again for your help. I will try transient.

Regards,

Alex

[peteroznewman]

Hello Alex,

Yes, you could use a remote displacement and input the Z-displacement time-history for the point at C in a Transient Structural model. This will be easier to implement as you probably know the displacements, which are enforced, rather than the forces.

Regards,

Peter

[Ashish Khemka]

Hi Alex,

 

Just  a comment:

If the response is nonlinear, the vibrations will be nonlinear and thus not be a sinusoidal response we associate with 'frequency' as in harmonic analysis.  To state this in a different manner, a nonlinear structure may not give a similar response as a linear structure 'at resonance', so it all depends on what you are trying to do with this extracted information - if you wish to have the nonlinear transient frequency content match a linear harmonic response analysis, this may not match at all, depending on the degree of nonlinearity in the system.

 

Regards,

Ashish Khemka