— For ADINA Users Only —
This is a restricted area meant for current ADINA users only.
If your license does not include current support of ADINA, please kindly exit this page.
Useful Hints and Issues
Regarding the Use of the ADINA System, v. 8.5.4
Issue No. 19, posted July 10, 2009
Deformation dependent centrifugal loading does not work properly with element birth and with element death. The wrong results, either in the form of no centrifugal load or no convergence, should be obvious to the user. There is no workaround. This bug is fixed in v. 8.5.5.
Issue No. 18, posted June 16, 2009
If high-order elements (2-D 9-node or 3-D 27-node elements) are used for CFD incompressible flow problems other than enclosures (surrounded by prescribed or fixed velocities), the "mathematical formulation" should be used, and not the "physical formulation". The "physical formulation" will be available in ADINA v. 8.6.1.
Issue No. 17, posted May 15, 2009
Wrong prescribed displacements are generated when multiple prescribed displacements with the 'unloading option' and non-zero arrival times are applied to the same degree of freedom. Note that displacement unloading is an advanced feature different from reducing the applied displacement to zero; please check the ADINA documentation for details.
If one displacement load is applied, then unloaded, then another
with a higher arrival time is activated, the second prescribed
displacement load will not be properly applied. Applying the second
displacement in a restart analysis is a workaround. This bug is fixed in v. 8.5.5.
Issue No. 16, posted May 4, 2009
In the Windows 32-bit version, when Automatic Memory Allocation is employed in a collapse analysis with deformation-dependent pressure loading applied on plate elements, the program may crash. As a workaround, do not use the Automatic Memory Allocation option when submitting a job. This bug is fixed in v. 8.5.5.
Issue No. 15, posted May 4, 2009
The 3D-iterative solver fails when generalized constraints are applied on mid-side nodes of 3D solid elements. This does not apply to mesh glueing. Two workarounds should be considered: use the default standard constraints instead of generalized constraints, or use the sparse solver instead of the 3D-iterative solver. This bug is fixed in v. 8.5.5.
Issue No. 14, posted May 4, 2009
When using a multilinear plastic material with 3-D solid elements, the first two points in the stress-strain curve should not have the same yield stress. When the first two points in the stress-strain curve have the same yield stress, then this yield stress is used for all accumulated effective plastic strains (in other words, perfect plasticity is incorrectly assumed and this can be seen in the band plot of the effective stress). This error can be avoided by making sure that the second point in the stress-strain curve is not the yield stress. This bug is fixed in v. 8.5.5.
Issue No. 13, posted March 27, 2009
When ATS is used in ADINA-T (and in the thermal part of ADINA-TMC) the reference temperature norm used to assess convergence can increase to unrealistic values during an ATS substep. This can cause subsequent time steps to converge when they should not. The user can detect this problem by looking at the temperature norm in the output (*.out) file, which in ADINA-T is taken as the maximum value of all steps. This bug is fixed in v. 8.6.
Issue No. 12, posted March 27, 2009
When defining more than 9999 truss element data with the ENODES/EDATA commands, the program stops with a Fortran error during the input conversion. This error can be avoided by defining less than 9999 elements in each truss element group. This bug is fixed in v. 8.5.5.
Issue No. 11, posted February 2, 2009
When 2-node and 4-node pipe-beam elements are defined in the same element group, ADINA might stop with the error of non-positive definite matrices. As a workaround, please define different element groups for 2-node and 4-node pipe-beam elements. This bug is fixed in v. 8.5.5.
Issue No. 10, posted February 2, 2009
When using the rubber models with viscoelasticity, if the ratio of the time step to the time constant is smaller than about 1e-15, then the results may be incorrect. When temperature effects are included, the time step in the above is the "reduced" time step (in which temperature effects are included). As a workaround, make sure that the time step is large enough so that the above condition is avoided. This bug is fixed in v. 8.5.5.
Issue No. 9, posted February 2, 2009
When using the Inquire icon on pipe elements, the "element sub-type" message is incorrect. When clicking on a 4-node pipe element, the message is "Element sub-type = 2-node element; beam-type behavior (no ovalization)", and when clicking on a 2-node pipe element, the "Element sub-type" message is missing. This bug is fixed in v. 8.5.5.
Issue No. 8, posted February 2, 2009
If all elements in a 3D element group are variable-number-node elements (meaning they are higher-order elements with one or more missing mid-side nodes), the results may be wrong. In this case, the program will report negative Jacobian for some elements that are clearly not distorted. This bug will only be triggered if all elements in the 3D element group also have the same number of missing mid-side nodes, and hence it is hard to trigger. This bug is fixed in v. 8.5.5.
Issue No. 7, posted February 2, 2009
If an universal results file is requested for the frequency analysis, the first mode shape listed in the .unv file is garbage, e.g., nodal labels are infinite numbers. The workaround is to set the PRINTOUT VOLUME to the "For all nodal results" option from Menu > Control in the AUI.
Note that this will also result in all nodal eigenvectors being written to the .out file. This bug is fixed in v. 8.5.5.
Useful Hints and Issues
Regarding the Use of the ADINA System, v. 8.5.3
Issue No. 6, posted January 8, 2009
If a nonlinear cyclic symmetry frequency restart analysis is performed for a model with contact, friction and mesh glueing, the calculated natural frequencies and mode shapes may be wrong. All contact nodes in this case will be assumed to be sliding. Unfortunately, there is no workaround. This bug is fixed in v. 8.5.4.
Issue No. 5, posted January 8, 2009
If modal stress output is requested in nonlinear cyclic symmetry frequency analysis and not all elements have the same number of nodes, the stresses (and strains) for cyclic parts two and above may be wrong. The frequencies and mode shapes are not affected. This bug is fixed in v. 8.5.4.
Issue No. 4, posted December 12, 2008
In frequency domain analysis with ground motion, the percent modal mass in each coordinate direction is currently calculated by dividing the modal mass in that direction by the average total mass. But in the case when concentrated masses have different values in different directions, it is more accurate to compute the total mass in each direction, and then to compute the percent modal mass by dividing the modal mass in that direction by the total mass in that direction. This bug is fixed in v. 8.5.4.
Issue No. 3, posted December 12, 2008
When using the bilinear plastic material with kinematic hardening in ADINA, negative strain energy density might be obtained. If fracture analysis is not activated, this only affects the reported strain energy (if strain energy calculation is requested). For fracture mechanics analysis, the results will be affected. This bug is fixed in v. 8.5.4.
Issue No. 2, posted December 12, 2008
Cyclic symmetry loads are not plotted properly in AUI when only one cyclic part is displayed. This is a display issue; the loads are shown properly if the complete model is displayed. This bug is fixed in v. 8.5.4.
Issue No. 1, posted December 12, 2008
In ADINA, convergence may be slow in bolt loading with multiple bolt steps and linear element groups. The converged solution is not affected. Performance is not affected if a single bolt step is used, or if all element groups are nonlinear. This bug is fixed in v. 8.5.4.
For the Useful Hints and Issues Regarding the Use of ADINA System 8.4, click here.