13 Adquisición de bienes y servicios
1 DESMONTE DE VEGETACIÓN
1.1 Impacto sobre vegetación y 1.2 Impacto sobre calidad de paisaje
Product: ABAQUS/Standard
Submodeling is the technique of analyzing a local part of a model with a refined mesh, based on interpolation of the solution from an initial, global model (usually with a coarser mesh) onto the nodes on the appropriate parts of the boundary of the submodel. This local refinement procedure provides a cost-effective approach to model enhancement. Shell-to-solid submodeling models a region with solid elements, when the global model is made up of shell elements.
The purpose of this example is to demonstrate the shell-to-solid submodeling capability in ABAQUS.
Geometry and model
The joint between a pipe and a plate is analyzed. A pipe of radius 10 mm and thickness 0.75 mm is attached to a plate that is 100 mm long, 50 mm wide, and 1 mm thick. The pipe-plate intersection has a fillet radius of 1 mm. Taking advantage of the symmetry of the problem, only one-half of the assembly is modeled. Both the pipe and the plate are assumed to be made up of aluminum with E =69 ´ 103 MPa and Poisson's ratio º = 0.3. The global model consists of S4R elements with the mesh layout as shown in Figure 1.1.10-1. Since a shell model is used, the fillet radius is not taken into consideration.
PCAV,CVOL
** STEP 2 - SEAL CAVITY (KEEP RIGID SURFACE FIXED)
** NONLINEAR LOAD DEFLECTION CURVE
1.1.10 Shell-to-solid submodeling of a pipe joint
Product: ABAQUS/Standard
Submodeling is the technique of analyzing a local part of a model with a refined mesh, based on interpolation of the solution from an initial, global model (usually with a coarser mesh) onto the nodes on the appropriate parts of the boundary of the submodel. This local refinement procedure provides a cost-effective approach to model enhancement. Shell-to-solid submodeling models a region with solid elements, when the global model is made up of shell elements.
The purpose of this example is to demonstrate the shell-to-solid submodeling capability in ABAQUS.
Geometry and model
The joint between a pipe and a plate is analyzed. A pipe of radius 10 mm and thickness 0.75 mm is attached to a plate that is 100 mm long, 50 mm wide, and 1 mm thick. The pipe-plate intersection has a fillet radius of 1 mm. Taking advantage of the symmetry of the problem, only one-half of the assembly is modeled. Both the pipe and the plate are assumed to be made up of aluminum with E =69 ´ 103 MPa and Poisson's ratio º = 0.3. The global model consists of S4R elements with the mesh layout as shown in Figure 1.1.10-1. Since a shell model is used, the fillet radius is not taken into consideration.
In the submodel the joint and its vicinity are meshed using three-dimensional continuum elements (C3D20R) with four layers through the thickness (see Figure 1.1.10-2). The solid model extends 10 mm along the pipe length and has a radius of 25 mm in the plane of the plate. The submodel accurately models the fillet radius at the joint. Hence, the submodel capability makes it possible to calculate the stress concentration in the fillet. The problem could be expanded by adding a ring of welded material to simulate a welded joint (for this case the submodel would have to be meshed with new element layers representing the welded material at the joint). The example could also be expanded by including plastic material behavior in the submodel while using an elastic global model solution.
Loading
The pipe in the global model is subjected to concentrated loads acting in the 1-direction applied at the nodes at the free end, representing a shear load on the pipe. The total value of all concentrated forces is equal to 10 N.
Kinematic boundary conditions
In the global shell model the plate is clamped along all edges. In the solid submodel kinematic
conditions are interpolated from the global model at two surfaces of the submodel, one lying within the pipe and the other within the plate. The default center zone size, equal to 10% of the maximum shell thickness, is used. Thus, only one layer of driven nodes lies within the center zone, and only these nodes have all three displacement components driven by the global solution. For the remaining driven nodes only the displacement components parallel to the global model midsurface are driven from the global model. Thus, a single row of nodes is transmitting the transverse shear forces from the shell solution to the solid model.
Results and discussion
The loading and boundary conditions are such that the pipe is subjected to bending. The end of the pipe that is attached to the plate leads to deformation of the plate itself (see Figure 1.1.10-3). From a design viewpoint the area of interest is the pipe-plate joint where the pipe is bending the plate. Hence, this area is submodeled to gain better understanding of the deformation and stress state.
Figure 1.1.10-4 shows the contours of the out-of-plane displacement component in the plate and shows the differences between the behavior in the models. As expected, the solid model is in good agreement with the displacement of the shell model around the joint. The stress concentration in the fillet radius is obtained with the solid model. The maximum Mises stress in this region is equal to 80.6 MPa, which is 51% more than the Mises stress in the shell model. Similarly, the maximum principal stress in the fillet region in the solid model is 54% higher than the corresponding stress in the shell model.
Input files
shellsolidpipe_s4_global.inp S4 global model.
shellsolidpipe_c3d20rsub_s4.inp
In the submodel the joint and its vicinity are meshed using three-dimensional continuum elements (C3D20R) with four layers through the thickness (see Figure 1.1.10-2). The solid model extends 10 mm along the pipe length and has a radius of 25 mm in the plane of the plate. The submodel accurately models the fillet radius at the joint. Hence, the submodel capability makes it possible to calculate the stress concentration in the fillet. The problem could be expanded by adding a ring of welded material to simulate a welded joint (for this case the submodel would have to be meshed with new element layers representing the welded material at the joint). The example could also be expanded by including plastic material behavior in the submodel while using an elastic global model solution.
Loading
The pipe in the global model is subjected to concentrated loads acting in the 1-direction applied at the nodes at the free end, representing a shear load on the pipe. The total value of all concentrated forces is equal to 10 N.
Kinematic boundary conditions
In the global shell model the plate is clamped along all edges. In the solid submodel kinematic
conditions are interpolated from the global model at two surfaces of the submodel, one lying within the pipe and the other within the plate. The default center zone size, equal to 10% of the maximum shell thickness, is used. Thus, only one layer of driven nodes lies within the center zone, and only these nodes have all three displacement components driven by the global solution. For the remaining driven nodes only the displacement components parallel to the global model midsurface are driven from the global model. Thus, a single row of nodes is transmitting the transverse shear forces from the shell solution to the solid model.
Results and discussion
The loading and boundary conditions are such that the pipe is subjected to bending. The end of the pipe that is attached to the plate leads to deformation of the plate itself (see Figure 1.1.10-3). From a design viewpoint the area of interest is the pipe-plate joint where the pipe is bending the plate. Hence, this area is submodeled to gain better understanding of the deformation and stress state.
Figure 1.1.10-4 shows the contours of the out-of-plane displacement component in the plate and shows the differences between the behavior in the models. As expected, the solid model is in good agreement with the displacement of the shell model around the joint. The stress concentration in the fillet radius is obtained with the solid model. The maximum Mises stress in this region is equal to 80.6 MPa, which is 51% more than the Mises stress in the shell model. Similarly, the maximum principal stress in the fillet region in the solid model is 54% higher than the corresponding stress in the shell model.
Input files
shellsolidpipe_s4_global.inp S4 global model.
shellsolidpipe_c3d20rsub_s4.inp
C3D20R submodel, which uses the S4 global model.
shellsolidpipe_s4r_global.inp S4R global model.
shellsolidpipe_c3d20rsub_s4r.inp
Key input data for the C3D20R submodel, which uses the S4R global model.
shellsolidpipe_c3d20r_mesh.inp
Remainder of the input data for the C3D20R submodel.
shellsolidpipe_node.inp
Node definitions for the S4R and S4 global models.
shellsolidpipe_element.inp
Element definitions for the S4R and S4 global models.
Figures
Figure 1.1.10-1 Global shell model of pipe-plate structure.
Figure 1.1.10-2 Magnified solid submodel of the pipe-plate joint.
C3D20R submodel, which uses the S4 global model.
shellsolidpipe_s4r_global.inp S4R global model.
shellsolidpipe_c3d20rsub_s4r.inp
Key input data for the C3D20R submodel, which uses the S4R global model.
shellsolidpipe_c3d20r_mesh.inp
Remainder of the input data for the C3D20R submodel.
shellsolidpipe_node.inp
Node definitions for the S4R and S4 global models.
shellsolidpipe_element.inp
Element definitions for the S4R and S4 global models.
Figures
Figure 1.1.10-1 Global shell model of pipe-plate structure.
Figure 1.1.10-2 Magnified solid submodel of the pipe-plate joint.
Figure 1.1.10-3 Solid submodel overlaid on the shell model in the deformed state, using magnification factor of 20.
Figure 1.1.10-4 Comparison of out-of-plane displacement in plate for both models.
Sample listings
Figure 1.1.10-3 Solid submodel overlaid on the shell model in the deformed state, using magnification factor of 20.
Figure 1.1.10-4 Comparison of out-of-plane displacement in plate for both models.
Sample listings
Listing 1.1.10-1
** Basic steel properties
*MATERIAL,NAME=MAT1
** Basic steel properties
*MATERIAL,NAME=MAT1
*NSET, NSET=YSYMM
** step 1,Default
**
*STEP
Total load of 10.0 N in the 1-direction
*STATIC
** step 1,Default
**
*STEP
Total load of 10.0 N in the 1-direction
*STATIC
Listing 1.1.10-2
**Node, element and node set definitions
**are written into the file
**shellsolidpipe_c3d20r_mesh.inp