Purpose of modelling Rotational Mass in a Structural Model 6

[DimzK]

Hi Everyone,

I have a simple question on modelling of rotational mass in a structural model (with kg x m2 units) .
I have just come across a model with rotational mass inputted in to certian nodes.
I understand the need for nodal mass modelling (at a fundamental level) as this will effect the stiffness matrix during modal analysis and change period etc....

Why would we include rotational mass in a structural model on a specific node?
Is this to represent a turbine or a spinning object?

If so why not use a bending moment to represent this loading rather than rotational mass?

Thanks for you input.
Regards,

 

[human909]

How would you use a bending moment to represent that loading?

Why would we include rotational mass in a structural model on a specific node?
Is this to represent a turbine or a spinning object?

That would be one such use.

A rotating mass in a structural system will behave in extreme odd ways to what a structural engineer would normally expect a mass would behave. That said I would expect that structural translations as too small matter for all but extremely high inertial mass objects. (such as turbines and generators) And even then I'd expect it to be more a mechanical engineers role to provide the reactions based on structural movements. (After all the bearings would need to be designed for off axis loads.)

 

[DimzK]

The spinning object should apply a torque on to the suppoting element I would have thought? which is best represented using a bending moment, where we know the direction of the loading...rotational mass does not seem to indiciate the direction...

 

[Tomfh]

Higher rotational mass = harder to rotate. Eg a tight rope walkers pole has a much higher rotational mass about the tightrope axis than about its own axis.

 

[human909]

Objects with high rotational inertia behave in very counter intuitive ways.

The spinning object should apply a torque on to the suppoting element I would have thought? which is best represented using a bending moment, where we know the direction of the loading...rotational mass does not seem to indiciate the direction...

No. A spinning object does not apply torque on a supporting element on its own. Friction or it slowing down or accelerating could apply a minor torque but that isn't a reason to model rotational inertia.

You would include rotational inertia in a model because extremely high forces can be produced in if you accelerate (translate) a spinning object in a direction that isn't aligned with the axis of spin. EG if a turbine was mounted with the axis point in the x-direction and your building translated in the y-direct then you would get very high forces being produce in the z-direction.

(If this isn't clear, and my explanation isn't especially thorough anyway, then do some digging in physics on rotational inertia and precession.

I'm sure there are some great youtube videos describing the phenomenon. Start with the basics of how gyroscopes work and go from there.

 

[JStephen]

It might help to consider the type of structure and also look into the software definitions.
There is a presumption above that "rotational mass" and "rotating mass" are the same thing, which is not what I would think.
If this is some sort of generating station or something that would actually have large rotating masses, then that might be the case. I wouldn't think you'd model the entire effect on a single node, though.
It could relate to torsional vibration effects of the structure itself, but why "rotational mass" would be separated from "mass", I don't know. Unless, perhaps, it moves laterally with the structure but is not supported by the structure.

 

[XR250]

(If this isn't clear, and my explanation isn't especially thorough anyway, then do some digging in physics on rotational inertia and precession.

I'm sure there are some great youtube videos describing the phenomenon. Start with the basics of how gyroscopes work and go from there.

..or hold a spinning bicycle wheel by the axle and try to rotate it.

 

[Tomfh]

The question concerns “rotational mass,” not a “rotating mass.”

Rotational mass refers to the polar moment of inertia of a mass, which measures a masses’ resistance to angular acceleration about a specific axis. This is distinct from the angular momentum of a rotating mass, such as rotating turbine blades.

 

[Tomfh]

but why "rotational mass" would be separated from "mass"

"Mass" = lateral inertia, "Rotational mass" = rotational inertia. So mass affects how easily a node can move, and rotational mass affects how easily it can spin.

 

[geesaman.d]

I deal with rotating assemblies and rotational inertia comes into play in three ways.
1) startup torque (if mechanical loads are significant at zero- to partial- speed, then the added load of rotational acceleration may exceed driver torque).
2) Estimating the time to reach full synchronous speed, which for most electric motors is a very high current condition and the time drives the sizing of the starters.
2) Rotordynamic analysis that involves torsional resonance calculations. This is basically irrelevant in my world of assemblies that rotate below their first lateral natural frequency and torsional driver/driven are free of repetitive impulses, but I'm sure there are situations where the first torsional can be at risk of excitation.

 

[centondollar]

If your structure is e.g., a building or bridge (not an object requiring rotordynamic terms in the EOM) and the mesh is fine enough, the error due to neglecting rotational inertia (you have that if you use e.g., a conforming Euler beam mass matrix) can often be found to be minor - I refer you to the technical literature on FEA. The formulation with lumped translational masses also significantly reduces time spent in a certain types of dynamic analysis (explicit FEA with lumped masses produces a diagonal mass matrix, the (numerical) inversion of which is trivial), which is at least one reason for why it is commonly given as an option in numerical solvers.