 |
|
  |
|
  |
首 页 新闻中心 产品介绍 应用实例 技术服务 工作机会 联系我们 English |
 制造业---应用实例 |
NORWEGIAN ENGINEERS PERFORM DROP TEST SIMULATION TO DETERMINE LIMITS OF PROTECTION NET FOR NEW ELF OIL RIG Engineers at Selantic Industrier A.S. performed a drop test simulation with ALGOR's Accupak/VE Mechanical Event Simulation software to determine the limits of a protection net for a new ELF offshore oil platform currently under development in the North Sea, off of the East Coast of the UK. The nets will be positioned at the top of two legs to prevent objects from falling into the leg trusses and damaging pipes, which transport oil from the ocean floor. Artist's rendering courtesy of Technip-Geoproduction of Paris, France. December 4, 1998, Pittsburgh, Pennsylvania - Advances
in computer-aided engineering technology are enabling engineers
of all disciplines to create more complex, detailed finite element
models that realistically simulate the behavior of interacting
systems. Recently, engineers at Selantic Industrier A.S. in Agotnes,
Norway used Accupak/VE Mechanical Event Simulation (MES) software
for Virtual Prototyping from Pittsburgh-based ALGOR, Inc. to simulate
the interaction between a small steel container and a dropped
object protection net -- a virtual prototype test that previously
could only be performed physically in a laboratory.
The MES ApproachLars Bjoland, technical manager of Selantic Industrier and an ALGOR customer for over seven years, is no stranger to ALGOR's finite element analysis (FEA) software; however, the protection net project was his first attempt at using ALGOR's Accupak/VE MES capabilities."In the beginning of the project, my colleague performed some rough hand calculations to determine the best approach for the net design," Mr. Bjoland said. "We could not make any solid conclusions from his work because it was just too vague. Performing the calculation manually would be impossible." The next logical step was to turn to FEA. Mr. Bjoland originally modeled the net using Superdraw III, ALGOR's precision finite element modeling tool, and replaced the falling object with nodal forces acting directly on the net. Mr. Bjoland was unsatisfied with this for two reasons: calculating the correct loading was complicated and time consuming, and he wanted to simulate the springboard effect that occurs when a falling object deflects off of the net. Modeling the container and applying known physical properties such as its dimensions and mass, the height from which it falls and gravity acceleration enabled Mr. Bjoland to realistically simulate the interaction of the container and the net within a short time frame. To set up the MES, Mr. Bjoland added a container to the existing protection net model, which is made of truss elements with three degrees of freedom. A rope would exhibit resistance only when pulled outward like a cable when the object strikes the net. As the net rebounds, the ropes would push together without resistance; thus, Mr. Bjoland did not need to consider bending moments. Mr. Bjoland placed the 5,000-kg steel container model, measuring 2 by 2 m, approximately 12.6 m above the net. The net had three sides, each 16-m long, and was terminated in each of the three corners with fully constrained boundary conditions. He specified gravity acceleration for the container. Mr. Bjoland designated contact elements between the surfaces of the container and the net to enable complete interaction, including the transfer of inertia from one object to the other. Then Mr. Bjoland specified the duration of the simulation because he wanted to determine the behavior of the system over time. After specifying material properties of steel for the container and of aramid, a synthetic fiber, for the ropes, Mr. Bjoland processed the nonlinear elastic material model with Accupak/VE. Analysis Results and ModificationsAccording to Mr. Bjoland, the initial deflections exceeded Technip's failure criteria of 2.6 m, the distance to critical objects underneath the net. Mr. Bjoland performed several variations of the analysis, dropping the container at the center, at one corner and along the edge of the net, to confirm the results before conferring with his client.
"After the first set of analyses, we were able to determine that
the current net design would fail under the extreme loading from
the impact," Mr. Bjoland said. "Without Accupak/VE, we would have
been required to make a prototype to get the same conclusion."
Physical Prototype TestingMr. Bjoland anticipates conducting small-scale prototype tests to confirm the simulation results. This testing will replace full-scale testing that would have been necessary if he had not used Accupak/VE. This translates into a decrease in the amount of time and cost of materials needed in the course of physical prototype testing."If we had not used Accupak/VE, we would need to do full-scale testing to find the deflection and termination forces," Mr. Bjoland said. "We also would have been required to perform additional physical testing to determine the design modifications that we made after the initial mechanical event simulation." Physical prototype testing will begin in early 1999 with the platform becoming operational in the year 2000. Mr. Bjoland expects to use Accupak/VE for future projects. "Engineers always need to be open to new ways of solving problems to get the best results," Mr. Bjoland said. "If we hadn't been open to a new method of modeling in this case, we would not have been able to show the behavior of the net as quickly or inexpensively."
|
