• Tidak ada hasil yang ditemukan

Independent assessment and benchmarking of no/low cost finite element analysis software for linear and non/linear static structural analysis.

N/A
N/A
Protected

Academic year: 2023

Membagikan "Independent assessment and benchmarking of no/low cost finite element analysis software for linear and non/linear static structural analysis."

Copied!
183
0
0

Teks penuh

This was done by conducting an initial market research study and identifying a range of available no-cost/low-cost FEA packages. It is therefore concluded that there are no-cost/low-cost FEA packages that can be used in place of high-cost commercial packages for linear and non-linear static structural analyzes of isotropic materials.

Introduction

  • Research Problem
  • Research Question
  • Research Objectives
  • Delineations and limitations
  • Dissertation overview

Their work entitled “The finite element method in structural and continuum mechanisms” is considered the first book entirely devoted to FEM (Moaveni, 1999). This research will not serve as an in-depth study of the mathematics and formulation of the numerical methods used in finite element analysis.

Literature Review

The Impact of FEA in the Engineering Industry

The purpose of the report was to answer the following question: "Are today's engineers making decisions based on simulations?" The report's findings were based on survey responses from 826 respondents who are from a wide range of industries worldwide. When looking at the more detailed design phase, more than 75% of respondents stated that simulation is used to refine or select ideas.

The Finite Element Method

The next step is to determine the displacement function to be used for the finite element formulation. The basis of the finite element solution is the principle of virtual displacement, also referred to as the principle of virtual work.

Figure 2-1. Ladder represented as a discretized element-node model (BSB Technocrats, 2016)
Figure 2-1. Ladder represented as a discretized element-node model (BSB Technocrats, 2016)

The Process of Finite Element Analysis

Thus, the analyst will need to specify the refinement based on an understanding of the physical problem as well as engineering experience (Mac Donald, 2011). The locations and types of boundary conditions are based on underestimation of the physical problem.

Table 2-1. Overview of typical structural finite elements
Table 2-1. Overview of typical structural finite elements

Errors and Uncertainty Present During the FEA Process

The only way to mitigate this error is to increase the complexity of the model. Furthermore, if a mesh is too coarse, it may not properly capture the true response of the component.

Significance of Verification and Validation

To disambiguate these two terms, verification is concerned with the numerical accuracy of the computational model. The verification and validation process must be done simultaneously with the generation of the model.

Different Licensing Approaches for FEA Packages

  • Commercial Proprietary Software Licences
  • Open Source Software Licences
  • Freeware

Code verification should be performed before generating the finite element model, and computational verification should be performed to ensure that the correct numerical method is used to obtain results (American Society for Mechanical Engineers, 2006). Open source software packages are generally licensed under a version of the GNU General Public License (GNU Operating System, 2007).

Independent Studies Conducted on the Performance of No-Cost/Low-Cost FEA

Thakore goes into detail about the user interface and functionality of the popular open source FEA package, Code_Aster with Salome Meca. It was clear that there is an active community of researchers and analysts using open source FEA packages.

No-Cost/Low-Cost FEA Package Investigation

  • Selected software Packages
  • Capabilities
  • Operating System
  • Internal Modules Present in Selected Packages
  • Compatible File Types
  • User Support and Documentation
  • Software Pricing

Built-in pre-processor which allows import of CAD geometry as well as geometry definition. Contains a built-in post processor which allows visualization of the results in the form of scalar plots.

Table  2-4  shows FEA capabilities of the chosen software packages.  The scope of this  research only concerns itself with linear and non-linear static structural FEA, however Table  2-4  also includes multi-body contact models, dynamic structural and  com
Table 2-4 shows FEA capabilities of the chosen software packages. The scope of this research only concerns itself with linear and non-linear static structural FEA, however Table 2-4 also includes multi-body contact models, dynamic structural and com

Premium Commercial FEA Package Investigation

  • Selected Software Packages
  • Capabilities
  • Operating Systems
  • Internal Modules
  • Compatible File Types
  • User Support and Documentation

It can generate files that can be exported to text editing software and other graphics processing software (ANSYS inc, 2013). The software package can be set to automatically select solvers and does this based on problem type, analysis options, contact types etc.

Table 2-10 denotes the capabilities of each proprietary software investigated.
Table 2-10 denotes the capabilities of each proprietary software investigated.

Browser-Based FEA simulation platforms

  • SimScale Browser-based FEA simulation platform

Post-processing: SimScale's post-processing interface is used once the results are generated. It is very basic and it is suggested that external post-processing software can be used for a better representation of the solution if necessary (SimScale GmbH, 2015).

Benchmarks and Standards for Performance of FEA Packages

All of this can be used by the analyst to create a model with the appropriate mesh to determine the accuracy of the FEA package. Most of the major FEA software developers are members of NAFEMS and all seem to subscribe to validating their packages through NAFEMS Standard Benchmarks or some other industry accepted benchmark publication.

Chapter Discussion

The no-cost/low-cost packages that will be further investigated in the performance benchmarking phase of the research are: Code_Aster/Salome Meca, Mecway and Z88 Aurora. Solidworks Simulation is still relatively new and the reputation of the software is not yet at the level of ANSYS or Siemens NX.

Chapter Conclusion

Methodology

Research Methodology

  • Identify and Investigate No/Low Cost FEA Packages Available
  • Identification of Independent Benchmarking Studies relating to No-
  • Subjection of No-Cost/Low-cost Packages to Selection Criteria
  • Selection of One or More Available Premium Commercial Package
  • Review and Selection of Benchmarking Test Cases
  • Solution of Test Cases Using FEA Software Packages
  • Processing and Interpretation of Results

Once the analysis capability criteria were met, the price of the package and licensing approach was verified. The package's internal preprocessor was used for the creation of the finite element model.

Figure 3-1. Package selection procedure
Figure 3-1. Package selection procedure

Limitations

Conclusion

Chosen Benchmark Test Cases

Linear Static Structural Analysis Test Cases

  • Grid Point Load on an Articulated Truss
  • Thin Shell Wall in Pure Bending
  • Axisymmetric Pressure Vessel
  • Internal Pressure on Thick-Walled Spherical Container
  • Flat Bar with Stress Concentration

This test case was chosen because of the following: The problem was taken from a reliable publication on FEA standards, so the results for this test case were reviewed. The decision to include this test case in the research was made because of the following: The test case originated from the fourth edition of a reliable engineering text, which implies that the validity of the results has been established. This test case is taken from an industry-accepted publication "The NAFEMS Standard Benchmarks" (NAFEMS, 1990).

The decision to include this test case in the research was made because of the following: The test case originates from an industry-accepted standards publication, therefore, the results have been reviewed and validated. The decision to include this test case was based on the following: The test case originates from an industry accepted standards publication and therefore, the results have been reviewed and validated. This test case was developed based on empirical models found in Mechanics of Materials by (Beer et al., 2012).

This test case is a 3-D linear static analysis of a flat bar with stress concentrations under a tensile load.

Table 4-1 below shows the information associated with this test case.
Table 4-1 below shows the information associated with this test case.

Non-Linear Static Structural Analysis test cases

  • Large deflection of a Z-Shaped Cantilever Beam under End Load
  • Plastic Deformation of Tensile Test Specimen

In order to successfully analyze this test case, the 2-D shell geometry of the cantilever must be imported into the preprocessor. Finally, the post-processing module of the package must be able to represent the target solution on a suitable plot. This part is shorter than the minimum range of the extensometer (a device used to measure strain).

It can be seen that the majority of the result sets fall within one standard deviation of the median. These fell within one deviation of the selected median curve for most of the extension. In order to successfully analyze this test case, the 3-D solid geometric representation of the sample must be imported into the preprocessor.

Finally, the package's post-processing module must be able to display the target solution on an appropriate plot.

Table 4-12 below shows the target solution as specified in the benchmark problem.
Table 4-12 below shows the target solution as specified in the benchmark problem.

Model Generation Procedure and Presentation of Results

Model Generation

  • Mecway
  • Z88 Aurora
  • SimScale
  • Code_Aster/Salome Meca
  • ANSYS
  • Siemens NX

The application of the shear constraints and the compressive load can be seen in Figure B.32 and Figure B.33 respectively. Use of a displacement limitation and the pressure load can be seen in Figure B.34 and Figure B.35 respectively. The application of the shear constraint and the compressive load can be seen in Figure B.42 and Figure B.43 respectively.

The application of the frictionless supports and internal compression load can be seen in Figure B.94 and Figure B.95. The application of the frictionless supports and compression joint can be seen in Figure B.96 and Figure B.97. The application of the nodal fasteners and force can be seen in Figure B.105 and Figure B.106 respectively.

The use of a fully fixed constraint and force can be seen in Figure B.116 and Figure B.117.

Figure 5-1. Error dialog box displayed when attempting to create a 2-D mesh.
Figure 5-1. Error dialog box displayed when attempting to create a 2-D mesh.

Presentation of Simulation Results

  • Test Case 1: Grid Point Load on an Articulated Truss
  • Test Case 2: Thin Shell Wall in Pure Bending
  • Test Case 3: Axisymmetric Pressure Vessel
  • Test Case 4: Internal Pressure on Thick-Walled Spherical Container
  • Test Case 5: Flat Bar with Stress Concentration
  • Test Case 6: Large deflection of a Z-Shaped Cantilever Beam
  • Test Case 7: Plastic Deformation of Tensile Test Specimen

The specific output results of each package are graphically represented below in Figure 5-6 and Figure 5-7. The specific output results of each package are graphically represented here in Figures 5-9 through Figures 5-14. The specific results from each package are graphically represented here in Figure 5-15 and Figure 5-16.

This section presents the results of each FEA package for Test Case 6: Large deflection of a Z-shaped cantilever beam under ultimate load. Packages that could not perform the loading specified in the test case data, ie.

Figure 5-2. Graph of target output “X-directional displacement at node C” for each package
Figure 5-2. Graph of target output “X-directional displacement at node C” for each package

Discussion of Results

  • Mecway
  • Z88 Aurora
  • SimScale
  • Code_Aster/Salome Meca
  • ANSYS
  • Siemens NX
  • Overall Discussion of Results

A coarse tetrahedral mesh was used and the solver was able to calculate the required stress output. Code_Aster was able to calculate the required directional displacement results and store additional individual results. The model associated with Test Case 3: Axisymmetric pressure vessel was created using the suggested approach.

ANSYS Mechanical was able to calculate the required stress results as well as displacement magnitudes. The solver was able to calculate the required directional displacement results and present the tabulated results in increments. ANSYS and NX were able to generate results showing acceptable accuracy for each test case.

All packages were able to analyze test case 6: large deflection of a Z-shaped cantilever beam under end load.

Conclusion

Findings

Linear elastic cases were routine and the results showed good accuracy compared to the target solutions of the test cases and the premium packages used. Based on the experience of using Mecway and the standard of results produced, it is safe to say that Mecway can be used in place of trusted commercial software for linear and nonlinear isotropic structural analysis. Mecway can be seamlessly integrated into the operational capabilities of an established company looking to replace its expensive commercial package for linear and non-linear static structural analysis using isotropic materials.

The results that were generated showed good accuracy with respect to the target solutions of the test case as well as the premium packages. All things considered, it would be difficult, at this stage of development, for Z88 to be used in place of a reliable commercial package for linear and nonlinear static structural FEA. Model generation capabilities and accuracy of results relative to target solutions and premium packages have led to the conclusion that Code_Aster/Salome Meca can be used in place of a reliable commercial package for linear and non-linear static structural FEA.

Overall, it has been shown that there are actually no-cost/low-cost options for users who wish to avoid the high cost of premium commercial FEA packages for the analysis of linear and nonlinear static structural problems with isotropic materials.

Recommendations for Further Research

In verifying and validating finite element analysis with field tests, universally what is the expected percentage that FEA should meet.

Figure D.1. Displacement in X-direction.
Figure D.1. Displacement in X-direction.

Gambar

Figure  2-4  represents the basic overview of the FEA process, adapted from (Mac Donald,  2011)
Figure 2-5 shows a general procedure for the interpretation of structural FEA results adapted  from (Mac Donald, 2011)
Table 2-6. Internal modules present in selected no-cost/low-cost FEA packages
Figure 4-1. Truss system geometry (Société Française des Mécaniciens, 1990)
+7

Referensi

Dokumen terkait

Therefore, the objectives of the research are to modify a linear programming model for the optimization of cost in lamb feed formulation, to determine the optimal amount of feed