General Plane Motion - Quasi-static equivalent loading

[TimGow]

Hello,
I would like to model the connecting-rod from a crank-slider arrangement, and include the loading that tends to bend the connecting-rod due to acceleration, the dynamic body loads. The connecting-rod is in general plane motion as one end is rotating while the other end is reciprocating. Angular velocity, angular acceleration and linear acceleration can be calculated for particular positions on the connecting-rod, for the crank position of interest.

In FEA I can apply the reaction loads at each of the two bearing pin connections and I have linear acceleration (grav) and centripetal/centrifugal loading possibilities. I am seeking a method that achieves both lateral and axial loads at the pin connections. I could achieve the correct linear acceleration at one location, say the CofG, although elsewhere magnitude and direction would be incorrect. Perhaps there is a combination of linear and radial accelerations that would approximate the general plane motion.

Is there an established scheme for approximating the body loads in a static analysis?
Do you know of a technical paper discussing this topic?

Tim

 

[rb1957]

we can certainly apply body loads in just about any FEA code. I would expect that you can apply any of 6 body loads (3 translational, 3 rotational).

another day in paradise, or is paradise one day closer ?

 

[TimGow]

Hi rb1957,
Thank you for clarifying the availability of 3 translational and 3 rotational body loads. I'm seeking any known scheme that combines these body loads to approximate the acceleration states experienced by a connecting-rod in a crank-slider arrangement. The rudimentary scheme that I have is to lump the con-rod mass at the CofG, calculate the acceleration of the CofG location at the particular crank angle and apply the magnitude and direction of the acceleration to the FE model. This produces bending of the con-rod and transverse loads at the two bearing locations, as desired, but it is not a close representation of the true situation; the acceleration of the element masses away from the C0fG are not properly represented, e.g. accelerations at each end of the con-rod will vary in relative direction and magnitude, except at TDC, BDC.
Tim

 

[rb1957]

you have modelled the motion of the con-rod at it's CG ... translational and rotational ? I'd've thought that would apply the correct inertial forces.

can you break the con-rod into three pieces, now three CGs, and three set of inertial loads ?

I think you're converting the inertial acceleration into a point force at the CG ? This is not applying a body force in the FEM. Applying a body force in your model should load each node with the required inertial force.



another day in paradise, or is paradise one day closer ?