PNNL-13501 - Pacific Northwest National Laboratory
PNNL-13501 - Pacific Northwest National Laboratory
PNNL-13501 - Pacific Northwest National Laboratory
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
subsequently transform them back into the MARC input<br />
format. This capability was crucial to the subsequent<br />
example problems that incorporated remeshing and<br />
multiple analysis steps. Figure 3 shows the second stage<br />
for the stretch forming process. A continuous 2-step<br />
analysis was also performed in MARC to confirm that the<br />
results were for all purposes identical.<br />
Test Problem 2<br />
The second problem added mesh refinement to Test<br />
Problem 1. The goal was to stop the forming analysis,<br />
refine a region of the mesh, remap the intermediate results<br />
to the new mesh, and restart the analysis. Using internally<br />
developed gridding software, NWGrid, we developed data<br />
transformation tools to convert between the common data<br />
format and NWGrid. Modifications to NWGrid were<br />
required to support shell elements with multiple layers<br />
and support the reflection of results. Figure 4 illustrates<br />
the additional data transformation routines that are<br />
required to support the remapping of model results<br />
between different meshes. Figure 5 compares the<br />
equivalent plastic strain distributions at the intermediate<br />
forming stage for the original analysis in Test Problem 1<br />
with the refined mesh in Test Problem 2. Figure 5 also<br />
shows the significant capability in NWGrid to reflect the<br />
results to construct a full 360-degree model from the<br />
previous ¼-symmetry model. This is an important<br />
capability when the forming process is symmetric, but<br />
subsequent operations; such as welding, introduce nonsymmetric<br />
thermal or mechanical loads.<br />
Test Problem 3<br />
The intent of the third problem was to combine and<br />
extend the capabilities developed for test problems 1 and<br />
2 to prepare the meshes for a welding analysis shown in<br />
Figure 6. The schematic of the complete process is<br />
shown in Figure 7. The goal was to map the results of a<br />
forming simulation to a new mesh that included the<br />
necessary mesh detail to simulate joining of a beam to the<br />
hydro-formed part. Mesh refinement is also necessary in<br />
this case to capture the thermal gradients of the welding<br />
analysis. Remeshing the formed part to have coincident<br />
node spacing with the attached beam proved to be an<br />
ambitious goal, and this problem was only partially<br />
completed with the allotted funding. Figure 6 shows a<br />
final thickness plot of the hydroforming analysis and the<br />
combined meshes of the hydroforming analysis and the<br />
beam introduced for welding. However, additional work<br />
is required to generate a mesh at the intersection of these<br />
two geometries that was well defined enough to provide<br />
accurate analysis results. This work was still in progress.<br />
182 FY 2000 <strong>Laboratory</strong> Directed Research and Development Annual Report<br />
Summary and Conclusions<br />
The software components developed under this effort now<br />
provide scientists and engineers with the ability to<br />
perform integrated multistep finite element analysis using<br />
a small collection of commercial finite element software<br />
applications. The present work to support the use of<br />
NWGrid further promises to extend this capability to<br />
provide robust remeshing/remapping capabilities.<br />
Critical technical issues have been identified and<br />
addressed, providing a robust solution to integrated<br />
modeling and simulation needs of automotive<br />
manufacturing scientists and engineers. Future<br />
development efforts should expand the collection of data<br />
transformation tools to support additional finite element<br />
application formats. Future research efforts should focus<br />
on increasing the ease with which NWGrid can be used<br />
by scientists and engineers, the incorporation of material<br />
properties into the common data format and data pedigree<br />
representations, and the development of a Web-based<br />
architecture for supporting data transformation tools (i.e.,<br />
Web-based translation services).<br />
Presentations<br />
Johnson KI. September 2000. “Integrated Modeling and<br />
Simulation, Joining of Hydroformed Structures, Glass<br />
Process Modeling.” Presented to Reza Sadeghi and John<br />
Long, MSC Software Corporation.<br />
Thurman DA. February 2000. Collaborative Problem<br />
Solving Environments: What Use Are They? Policy and<br />
Management Science Technical Network, Battelle Seattle<br />
Research Center.<br />
Thurman DA and T Peterson. March 2000. Information<br />
Technology to Support Public Involvement Processes.<br />
Puget Sound Chapter of the American Urban Planning<br />
Association.<br />
Thurman DA. April 2000. Collaborative Problem<br />
Solving Environments for Wastewater Management.<br />
Modeling and Assessment Unit, Water Treatment Group,<br />
Department of Natural Resources, King County,<br />
Washington.<br />
Thurman DA. June 2000. Collaborative Environments<br />
and Virtual Facilities. Members of the Battelle/<strong>PNNL</strong><br />
Mexican Alliance.