INREL -1 and INREL -2 1

[flyforever85]

Can someone explain me the difference between the 2 different param setting? I was able to make both run with no problem, I added SUPORT1 for INREL=-1 and nothing for INREL=-2 but I can't understand when I should one one or the other. THank you

 

[karachun]

With PARAM,INREL,-1 you manually select nodes that do not move, all deformations will be relative to this node. With PARAM,INREL,-2 solver selects this node using its internal logic.
That's all. Stressed state will be the same in both cases, but deformed view will differ.

 

[flyforever85]

Thank you Karachun. I assumed for a flying object the inrel = -2 option is the most valuable. Could you confirm?

Thank you

 

[karachun]

They both equal and produce same results. If you do not want to select SUPPORT node manualy then use -2.

 

[dmapguru]

Inertia relief studies the motion of free bodies under constant loading resulting in constant acceleration. A static analysis is used as it is assumed that any transient effects due to applied loading have long died out. An example use case might be re-establishing the degraded orbit of a satellite; burning an impulse rocket applies a load to the structure resulting in acceleration. Out in space, the satellite is not fixed to anything so there are no boundary conditions (SPC).

For a non-disjoint structure with no mechanisms, the structure should be free to move in 3 translational and 3 rotational orthogonal directions without storing any strain energy; these are the 6 rigid body modes familiar to those who know normal modes analysis of free-free structures. In the general case, the model may have fewer or greater than 6 of these modes; fewer than 6 modes may occur for a symmetric half model, for example and greater than 6 modes may occur for a model with a mechanism like a canard or tail surface. In order not to restrict the discussion only to the 6 classic rigid body modes, call the set of all modes where no strain energy is generated “singular modes”

INREL=-1
This method allows fewer, but no more than 6 singular modes. It is an elimination method which requires a statically determinate set of DOF for the reference frame. These DOF are defined on the SUPORT entry, and should be chosen where the structure has stiff behaviour to avoid poorly conditioned shapes. It is vitally important that the choice of SUPORT DOF do not over-constrain the structure - diagnostic output (SUPORT EPSILON) is available to assist with this. Studying the displacements of the structure due to small changes resulting in a different mass distribution can be studied with relative ease because the frame of reference does not change.

INREL=-2
This method must have exactly 6 singular modes (no fewer, no more). There is no check on this, so wrong answers occur if this is not respected; no diagnostic output is available to assist in checking for under or over constraint. It is an augmentation method which does not require a reference frame (DOF) to be defined by the user so no SUPORT entry is defined; the frame of reference is computed from the mass weighted average of all DOF where mass is defined; for the general case, this ends up being very close to the CG of the structure whether there is a GRID point there or not. Studying the displacements of the structure due to small changes resulting in a different mass distribution becomes problematic because the frame of reference changes.

In both methods, the displaced shapes of the singular mode shapes are generated from geometry. A consequence of this is the mass and constrained displacements of SPOINTs and any independent DOF of MPC or RBE3 connections that are fixed (SPC) are not handled correctly leading to wrong answers. Neither method is able to study (correctly) a structure with more than 6 singular modes. For the same model and correct choice of SUPORT DOF for INREL=-1, while staying withing the limitations of INREL=-2, the 2 solutions will yield the same stress field while the displaced shape will be different as the frame of reference in the 2 cases is not the same.

DG