ProductsAbaqus/StandardAbaqus/Explicit Features testedApplications of the temperature, field variable, and pressure stress procedures are tested. The first set of tests verifies that temperature and field variable data are properly transferred from a heat transfer analysis to a structural analysis. The second set of tests verifies the use of these commands in conjunction with composite structural shells. The third set of tests verifies the interpolation of temperatures to the midside nodes in a sequential thermal-stress analysis, when the heat transfer analysis is carried out using first-order elements and the stress analysis is carried out using second-order elements. The fourth set of tests verifies that temperatures are properly interpolated between dissimilar meshes. Heat transfer models and stress analysis models may have dissimilar meshes, and the nodal temperatures for the current model will be interpolated from the nodal temperatures from the heat transfer model. The fifth set of tests verifies that temperatures and pressures are properly defined using data line input for various combinations of these two commands. The fifth set of tests verifies that a solution-dependent variable from a heat transfer analysis is properly transferred as a field variable into a stress analysis. In several of the tests zero-increment results file output is requested. This output is used to define initial values of temperature, field variables, and pressure stress for subsequent structural analyses. Reading temperature and field variable data from results filesElements testedDC1D2 T3D2 Problem descriptionThese tests verify that temperature and field variable values are properly transferred to a structure. The structure being analyzed is a cantilevered truss made up of 10 T3D2 elements. Three different transient heat transfer runs are used to generate three results files containing temperature histories. These files will be read into subsequent stress analyses as either temperature or field variable data. All of the runs begin with the entire truss at some initial temperature; the temperature throughout the truss is then ramped to some new temperature. The three heat transfer runs are as follows:
The subsequent stress analysis runs are as follows:
Results and discussionThe exact solution to the heat transfer problems (xtfvtrt1.inp, xtfvtrt2.inp) consists of a linear temperature history. Temperature is uniform throughout the structure at each point in time. The solution given by Abaqus matches the exact solution. The only quantity of interest in the stress analysis runs is the temperature in the structure. Expected solutions are shown in Figure 1 through Figure 5. Input files
FiguresFigure 1. Temperature and field variables for xtfvtrs1.inp.
Figure 2. Field variable for xtfvtrs2.inp.
Figure 3. Temperature and field variable for xtfvtrs3.inp.
Figure 4. Temperature and field variable for xtfvtrs4.inp and xtfvtrsr.inp.
Figure 5. Temperature and field variables for xtfvtrs5.inp.
Composite shell temperature loadsProblem descriptionIn Abaqus/Standard these tests verify the use of predefined temperature and field variables in conjunction with composite structural shells. Both temperature and field variable results are generated from a single previously run heat transfer shell analysis. The same analysis can be used to generate field variable results since field variables are stored identically to temperatures in an Abaqus results file. A steady-state heat transfer analysis is performed to obtain the temperature distribution through the thickness of the composite layers. The heat transfer problem involves a three-layer composite shell that is subjected to prescribed thermal boundary conditions on its top and bottom surfaces. In the subsequent structural shell models, five section points per layer are used. The temperatures and field variables are assigned to these five points through a linear interpolation of the three values available per layer from the preceding heat transfer analysis. The results of these analyses verify that the temperatures and field variables are assigned properly. In Abaqus/Explicit, instead of a sequential analysis, a transient coupled dynamic temperature-displacement analysis is performed on a three-layer composite shell that is subjected to prescribed thermal boundary conditions on its top and bottom surfaces. A sufficiently large step time is prescribed such that the analysis can reach the steady-state regime. Three temperature points are used for each layer. The temperature distribution obtained is compared to the exact solution. Results and discussionThe heat transfer run matches the exact solution for the temperature distribution through the composite shell layers. In addition, these values are transferred properly in Abaqus/Standard to the structural composite shell as either temperature or a field variable. In Abaqus/Explicit both heat-transfer and stress analyses are solved simultaneously, and the results match the analytical solution and the Abaqus/Standard solution.
There is a linear variation of temperature or field variable between the top and bottom of each layer. Input files
Temperature interpolation to midside nodesProblem descriptionThese tests verify the interpolation of temperatures to the midside nodes of higher-order elements in a sequential thermal-stress analysis, when the heat transfer analysis is performed using first-order elements and the stress analysis is carried out using second-order elements. The results of the heat transfer analyses are read into the stress analyses, and the initial conditions applied to the heat transfer analysis are read into the stress analyses. In both analyses the temperatures at the midside nodes are interpolated from the corner nodes of the element. Temperature interpolation is carried out on an edgewise basis for each element. Thus, the temperature at the midside node of an element is interpolated linearly from the temperatures at the corresponding corner nodes. The midside node temperature interpolation is tested for one-dimensional, two-dimensional, and three-dimensional elements. Only one element is used in the finite element models for both heat transfer analysis and stress analysis. Arbitrary material properties are assumed. Results and discussionThe results of the stress analysis with higher-order elements compare well with those obtained with linear elements. Input filesHeat transfer analyses:
The input files for the stress analyses with linear elements can be generated by suitably replacing the element type in the above files. Temperature interpolation between dissimilar meshesProblem descriptionThese tests verify the interpolation of temperatures between dissimilar meshes. This capability is available only for use with the output database file. The temperatures must be interpolated from the nodes of the heat transfer models to the nodes of the current stress analysis models. For the cases where the only dissimilarity is an element order, the temperatures at the midside nodes should be interpolated from the corner nodes of the element. However, for the purpose of verification we reused some of the models created for the midside cases. The results of the heat transfer (or coupled temperature-displacement) analyses are read into the stress analyses. The temperatures must be interpolated from the nodes of the element in the heat transfer models to the nodes of the current stress analysis models. The interpolation technique is tested for two-dimensional and three-dimensional elements. Results and discussionThe temperature distribution in the stress analysis models compares well with that obtained in the heat transfer (coupled temperature-displacements) models. Input filesCoupled temperature-displacement and stress analyses:
Transferring temperatures between dissimilar meshes with user-specified regionsProblem descriptionThe verification problems in this section test the interpolation of temperatures between dissimilar meshes with user-specified regions. The model consists of two part instances, as shown in Figure 6. A tiny gap exists between the two parts. A low gap heat transfer is applied along the gap so that a temperature jump results between the two adjacent surfaces. In this case temperature mapping using the general interpolation may result in erroneous temperature assignment to nodes on the adjacent surface due to the ambiguous association between target nodes near the interface surface and driving elements near this surface. The dissimilar mesh interpolation capability resolves the ambiguity by explicitly specifying the source regions in the heat transfer analysis from where the temperatures are read and the target regions in the current analysis onto which the temperatures are mapped. Figure 6. Model geometry with the gap amplified for illustration purposes.
Boundary conditions:The assembly is kept at a constant temperature of zero on the left boundary, and it is subjected to a constant surface heat flux of 0.003 on the right boundary and a constant surface heat flux of 1 on the top. The gap has a low gap heat conduction with a coefficient of 0.01. Results and discussionThe nodes on the top boundary of the inner part in the current analysis are shifted up slightly so that they fall inside the outer part. This shift is done intentionally to illustrate a case that would result in incorrect driving element selection during interpolation. The mapped temperature results when the dissimilar mesh interpolation capability is not specified are shown in Figure 7. The test shows that the temperatures when regions are not specified for the interpolation of temperatures between dissimilar meshes are mapped incorrectly for those nodes on the top surface of the inner part. The error occurs because Abaqus searches for a parent element that encloses each node in the current analysis or is closest to each node. For the nodes on the top surface of the inner part, the parent elements are found inside the outer part, resulting in erroneous temperature definitions at the nodes. The mapped temperature results with the dissimilar mesh interpolation capability are shown in Figure 8. The capability fixes the error by explicitly specifying the source and the target regions of the interpolations. In this test case the driving element set from the previous heat transfer analysis is selected to cover the same instance region as that covered by the driven node set in the current analysis; therefore, instance-to-instance mapping is achieved. Figure 7. Temperature mapping without use of the dissimilar mesh interpolation capability.
Figure 8. Temperature mapping with dissimilar mesh interpolation capability specified.
Input files
Reading temperature and pressure data from results filesElements testedCPE4 DC2D4 Problem descriptionThese tests verify that temperatures and pressures are applied properly to a structure when various combinations of temperature and pressure stress values are used in a mass diffusion analysis. Temperature and pressure stress initial conditions are read from the results file of an Abaqus/Standard analysis, and a series of pressure and temperature loadings are applied to the nodes of an element using data line input in the following sequence:
The material properties of the problem are defined such that When both the temperature and pressure gradients are applied to the model, the diffusion is driven by concentration gradients alone. The following must be confirmed by this test:
Results and discussionThe results match the exact analytical solutions for the applied temperature and pressure gradients. Input files
Reading solution-dependent variables from results filesElements testedDC1D2 T3D2 Problem descriptionThese tests verify that the solution-dependent variables from a heat transfer analysis are properly transferred as field variables in the subsequent stress analysis. The structure being analyzed is a cantilevered truss made up of 10 one-dimensional link elements. The solution-dependent state variables written to the results file are the averages of the values extrapolated to the element nodes. A separate results file is then generated, where the solution-dependent state variables value is stored as the second attribute under record key 201. The temperature and field variable values are set by reading the data from the results file of the heat transfer run as follows: xsdvttrt.fil Temperature xsdvttrt1.fil Field variable Results and discussionThe solution-dependent variable is transferred correctly into the stress analysis as a field variable. Input files
Reading scalar nodal output from the output database into field variablesElements tested
Problem descriptionThese tests verify that Abaqus/Standard:
The basic test procedure is as follows: A set of initial two- and three-dimensional heat transfer, mass diffusion, and piezoelectric analyses are run. In these analyses temperatures, normalized concentrations, and electric potentials are written as nodal data to output databases. Different combinations of temperature, normalized concentrations, and electric potential fields are read from these analyses and used to initialize and define temperature and field variables in subsequent stress/displacement analyses. Using the thermal and field expansion capability in Abaqus/Standard, the temperatures and field variables are used to drive the displacement fields by imposing volumetric strains. Results and discussionThe tests verify that temperature, normalized temperature, and electric potential fields are properly read and interpolated from an output database to initialize and define field variables. Input files
|