Die Design for the Millipede Former

Garrad, Ian Thomas (2008). Die Design for the Millipede Former B.Sc Thesis, School of Engineering, The University of Queensland.

       
Attached Files (Some files may be inaccessible until you login with your UQ eSpace credentials)
Name Description MIMEType Size Downloads
Garrad_Ian_Thomas_Thesis.pdf Garrad_Ian_Thomas_Thesis.pdf application/pdf 14.72MB 0
Author Garrad, Ian Thomas
Thesis Title Die Design for the Millipede Former
School, Centre or Institute School of Engineering
Institution The University of Queensland
Publication date 2008
Thesis type B.Sc Thesis
Supervisor Bill Daniel
Total pages 118
Language eng
Subjects 0913 Mechanical Engineering
Formatted abstract
It was the goal of this project to attempt to model the Millipede Former manufacturing process using Finite Element Analysis (FEA). The Millipede Former is an alternative manufacturing process to Cold Roll Forming for strip and sheet metal.

Cold Roll Forming (CRF) involves the passing of sheet or strip metal at room temperature through a series of rollers to shape the metal into a continuous section of (usually) constant
profile. CRF can be an expensive process, and relies heavily on the abilities of the operators, as many analytical models of the process have been attempted, with limited success.
Therefore, Dr. Bill Daniel, Dr. Scott Ding and Dr. Paul Meehan of The University of Queensland have developed an alternative manufacturing method, known as the Millipede Former, which is a continuous stamping process currently in the prototype stages of development. The prototype dies developed are shown in Figure 1.

The models to be developed using FEA were developed using the ABAQUS package, which is highly supportive of the nonlinear effects that will be experienced in the modelling:

• Contact between the dies and the blank (deformable material)
• Plastic deformation of the blank
• Substantial nodal deflections

It was found that ABAQUS was insufficient to model the Millipede Former process. In terms of implicit analysis methods, it was found that approximately forty percent of the simulation
was completed before the ABAQUS/Standard solver was unable to resolve the contact conditions occurring in the model. Many different modelling techniques were employed to attempt to mitigate this problem:

• Changing the blank boundary condition to prevent conflict between the contact interactions and the boundary condition
• Change the single stamping process into a double stamping process
• Lengthening the blank, so as to avoid conflict between the contact interactions and the boundary condition
• Changing the time increment in the failure step
• Softening the contact to attempt to improve the convergence of the model
• Attempt to more accurately model the Millipede Former process so as to avoid the need for the blank boundary condition
• The use of a compliant material next to the blank boundary condition to improve the convergence
• Use contact controls to relax the contact interactions and improve convergence
• Exploit the symmetry of the Millipede Former process so that the blank boundary condition can be moved to a different edge

With the exception of the softened contact, these modelling techniques were not successful in improving the quality of the solution significantly.

Additionally, the ABAQUS/Explicit models developed were not able to be classified as quasistatic, due to large inertial effects present in the model. This is due to computation restrictions, in order to allow the model to attain a certain degree of stability.

Therefore, it was recommended that a custom code be developed in order to successfully model the Millipede Former process, using an implicit Finite Element Analysis method. This
custom code should contain some of the following modifications to improve the practicality and accuracy of the analysis method:

• Replace the contact interactions between the blanks and the dies with boundary conditions.
• Build the FEA models directly from the CAD geometry.
• Improve the flexibility of analytical rigid surface modelling.
• Couple the contact searching algorithm with the timestep selection.
• Permit the writing of steps in a repetitive fashion, to significantly decrease the modelling time required.
• Build boundary conditions that are able to be released if and only if contact is experienced along that boundary condition.
• Build an adaptive meshing algorithm for two-dimensional elements.
Keyword Die Design
Millipede forming

Document type: Thesis
Collection: UQ Theses (non-RHD) - UQ staff and students only
 
Citation counts: Google Scholar Search Google Scholar
Created: Tue, 16 Dec 2014, 09:25:22 EST by Ahmed Taha Siddiqui on behalf of Scholarly Communication and Digitisation Service