2008 International ANSYS Conference

Transcription

1 2008 International ANSYS Conference Hybrid Submodeling Analysis Development and Applications Dr. K. S. Raghavan and H S Prasanna Kumar Structures Discipline Chief Infotech Enterprises Limited, Hyderabad, INDIA 2008 ANSYS, Inc. All rights reserved. 1 ANSYS, Inc. Proprietary

2 Conventional Submodeling Based on Saint Venant s principle Uses cut boundary displacements Limitations : Valid for linear analysis only The geometry of the submodel should be same as the corresponding region in the main analysis Any stiffness change will lead to erroneous stresses 2008 ANSYS, Inc. All rights reserved. 2 ANSYS, Inc. Proprietary

3 Alternate approaches Use a combination of displacements and forces from the full analysis. Possible if the submodel is transferring forces in a well defined load path Use only cut boundary forces. This is more versatile. Limitation of usage is that actual displacements and deformations are not of importance and only stresses are of importance 2008 ANSYS, Inc. All rights reserved. 3 ANSYS, Inc. Proprietary

4 Force + Displ. - Illustration AREAS FOR SUBMODEL Uniform Pressure Constrained CURVED PLATE WITH IN-PLANE LOADING 2008 ANSYS, Inc. All rights reserved. 4 ANSYS, Inc. Proprietary

5 Full Model von Mises stress Elastic Analysis Elastic Plastic Analysis 2008 ANSYS, Inc. All rights reserved. 5 ANSYS, Inc. Proprietary

6 Submodel Analysis FSUM values From elastic analysis (RBE3 or MPC) Results of elastic-plastic Analysis. SEQV plots Local peak zones are Unselected. CBD s from Elastic Analysis 2008 ANSYS, Inc. All rights reserved. 6 ANSYS, Inc. Proprietary

7 A small bracket part of a large pack CBF CBD Time for analysis Full model : 30 hours Submodel : 1 hour. Results of elastic-plastic Analysis. Left : Full model Right : Submodel 2008 ANSYS, Inc. All rights reserved. 7 ANSYS, Inc. Proprietary

8 Only forces (CBF) in the submodel Choose the region of interest. On all the cut boundaries apply the resultant forces and moments from full model This is done using either RBE3 or MPC (surface based constraint) In addition apply all surface, nodal and body forces consistent with the full model This set of forces will be in equilibrium Apply just the necessary constraints to prevent rigid body motion (3 in 2D and 6 in 3D) 2008 ANSYS, Inc. All rights reserved. 8 ANSYS, Inc. Proprietary

9 Only forces (CBF) in the submodel Tapered cantilever with groove and tip load Submodel nodes Full model UY Constr. UX, UY Constr. Forces and moments from full analysis applied at pilot nodes Minimum constraints to prevent Rigid body motion 2008 ANSYS, Inc. All rights reserved. 9 ANSYS, Inc. Proprietary

10 Stresses from the submodel Same mesh Near zero reactions at all constrained DOF Finer mesh 2008 ANSYS, Inc. All rights reserved. 10 ANSYS, Inc. Proprietary

11 Elastic plastic analysis Equivalent plastic strain Full Model Submodel Same mesh Submodel Finer mesh 2008 ANSYS, Inc. All rights reserved. 11 ANSYS, Inc. Proprietary

12 Geometry Changes in the submodel In this problem the force BC s are insensitive to the geometry changes within or outside the submodel region This is true as long as the applied loads and load path remains the same This makes it possible to design and redesign small portions of large structures relatively easily This approach will be superior to substructures analysis 2008 ANSYS, Inc. All rights reserved. 12 ANSYS, Inc. Proprietary

13 Example of geometry change. Smaller groove Radius, Submodel with Same BC s Full model 2008 ANSYS, Inc. All rights reserved. 13 ANSYS, Inc. Proprietary

14 One D to two - D 2D Duct wall 1D model (BEAM3) With BC s Approximate section Properties for this section Submodel domain Deformed shape 2008 ANSYS, Inc. All rights reserved. 14 ANSYS, Inc. Proprietary

15 Submodel analysis Forces and moments applied At pilot nodes. Values from Beam analysis Von Mises stress Left : Submodel Bot : Full 2D model 2008 ANSYS, Inc. All rights reserved. 15 ANSYS, Inc. Proprietary

16 2D frame 2D Frame, BEAM3 model Loads : 1.0G acceleration in Y-direction 3.0G acceleration in X-direction Submodel with BC s 2008 ANSYS, Inc. All rights reserved. 16 ANSYS, Inc. Proprietary

17 Submodel and full 2D model results Higher stress computed for the submodel is, at least, partly due to higher effective span of the cross member in the beam model. Full 2D model Submodel 2008 ANSYS, Inc. All rights reserved. 17 ANSYS, Inc. Proprietary

18 Observations Because of the assumption inherent in in beam analysis (mid-surface or mid-section) as applied to frames, there will be some error in the results with submodeling The error, however, will be conservative There is a conceptual difference between this problem and the earlier ones Here the cross member is sharing load with other members Thus the structural behavior will be sensitive to changes within If submodel is used for redesign a little bit of engineering judgment will be called for 2008 ANSYS, Inc. All rights reserved. 18 ANSYS, Inc. Proprietary

19 Summary Submodeling is a powerful analysis tool The method promotes more accurate analysis and also helps enhance productivity It has been shown that by using displacementforce or force-force cut BC s the method can be made even more versatile A few very simple problems have been solved with this approach to demonstrate the concept 2008 ANSYS, Inc. All rights reserved. 19 ANSYS, Inc. Proprietary

20 Summary It can be appreciated that in all the problems results that are not possible using the classic approach have been generated The method can be viewed as productivity enhancement tool and recommended for optimizing relatively small components in large assemblages 2008 ANSYS, Inc. All rights reserved. 20 ANSYS, Inc. Proprietary