— 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.

These Useful Hints and Issues, while reported for the specific version 8.6.5, are usually also applicable to the ADINA versions 8.6.4, 8.6.3, 8.6.2, 8.6.1 and 8.6.0.


Useful Hints and Issues
Regarding the Use of the ADINA System, v. 8.6.5

ADINA 8.6.5 has been released. Issues No. 52 - 54 found in version 8.6.5 are listed below.

Please contact us if you would like to have additional details or if you noticed other issues not listed here.


Issue No. 54, posted June 24, 2010

If an element group contains both 8-node and 6-node shell elements, and a large strain formulation is selected for the element group, then

  1. ADINA uses a large displacement/small strain formulation for the 8-node elements, and a large displacement/large strain formulation for the 6-node elements, and

  2. The porthole file cannot be loaded into the AUI.

As a workaround, use a large displacement/small strain formulation for the element group. Alternatively, use 9-node shell elements instead of 8-node shell elements.

This bug is fixed in v. 8.7. With the bug fix, the AUI will not generate a data file for ADINA, but will instead issue an error message.


Issue No. 53, posted June 24, 2010

If a negative solution starting time is used and a deformation dependent loading is applied, then the calculated reactions are incorrect in the first step. This might also lead to incorrect reactions in a subsequent dynamic restart run.

As a workaround, do not use a negative solution starting time when you have deformation dependent loading. This bug is fixed in v. 8.6.6.


Issue No. 52, posted June 24, 2010

A model with tied contact cannot converge when the friction delay option is turned on. When tied contact is used, then there is no relative movement between contacting parts; therefore, the friction delay option should not be used.

As a workaround, do not use the friction delay option with tied contact. This bug is fixed in v. 8.6.6.




Useful Hints and Issues
Regarding the Use of the ADINA System, v. 8.6.4

Issue No. 51, posted June 10, 2010

If a node is specified with a nodal damper and is constrained to more than one master nodal degree of freedom, then the program stops in the input phase during the reading of nodal damper data.

There is no workaround. This bug is fixed in v. 8.6.5.


Issue No. 50, posted June 10, 2010

The contact tractions are calculated incorrectly under the following conditions:

  1. 2D contact, and
  2. contact surface is closed, and
  3. "new" contact surfaces, and
  4. quadratic segments are used

Note that the contact forces and all other results are correct; only the contact tractions are incorrect.

This bug is fixed in v. 8.6.5.


Issue No. 49, posted June 10, 2010

When EPSN=0 in the CGROUP command, ADINA calculates the constraint function parameter "epsilon-n" automatically, but this automatic calculation may not give the expected results in all cases. Therefore it is recommended not to use EPSN=0.

In the documentation, the default value of EPSN is listed as 0, but the documentation is incorrect. The default value of EPSN is actually 1x10-12.

We will consider removing the EPSN=0 option in a future release.

This bug is fixed in v. 8.6.5.


Issue No. 48, posted May 24, 2010

If prescribed temperature gradients are applied using a reference temperature gradient for the whole model, then the temperature gradients are taken by the program without being multiplied by a corresponding time function value.

As a workaround, apply temperature gradients without using the reference temperature gradient. This bug is fixed in v. 8.6.5.


Issue No. 47, posted May 24, 2010

In the Transor LDC dialog box, the parameter DOF is wrongly taken as follows:

X-translation written as DOF=0 (should be DOF=1)
Y-translation written as DOF=1 (should be DOF=2)
Z-translation written as DOF=2 (should be DOF=3)
X-rotation written as DOF=3 (should be DOF=4)
Y-rotation written as DOF=4 (should be DOF=5)
Z-rotation written as DOF=5 (should be DOF=6)

As a result, the solution obtained may not be the expected solution, or the solution may not converge.

For a workaround, if there is no Z-rotation to be entered, then use the dialog box and enter a Y-translation to get an X-translation, etc. Otherwise, if a Z-rotation is required, the user should save and then edit the .in file. This bug is fixed in v. 8.6.5.


Issue No. 46, posted May 24, 2010

When using buckling modes as initial imperfections in a second run in which the degrees of freedom are changed due to contact, ADINA crashes.

There is no workaround. This bug is fixed in v. 8.6.5.


Issue No. 45, posted April 6, 2010

In an ADINA TMC analysis, when a heat-flow-only analysis is requested and when the structural material model is plastic-cyclic, the AUI does not plot or list the heat fluxes.

As a workaround, use the linear elastic structural material model. This bug is fixed in v. 8.6.5.


Issue No. 44, posted April 6, 2010

ADINA stops at the beginning of an analysis that has periodic cyclic symmetry with TMC. This only happens if the model has nodes exactly on the cyclic symmetry axis.

There is no workaround. This bug is fixed in v. 8.6.5.


Issue No. 43, posted April 6, 2010

When a mass proportional or centrifugal load is used in an ADINA TMC heat-transfer-only analysis, the program crashes.

As a workaround, remove the mass proportional or centrifugal loads (these are not necessary in the heat-transfer-only analysis). The bug is fixed in v. 8.6.5.


Issue No. 42, posted April 6, 2010

The element death option of the MNO-G spring mass property does not work. ADINA does not set the mass equal to zero when it detects the death of that element. If a time-history response of that element with death option is plotted, the result is obviously wrong after the death time of that element.

As a workaround, use the nonlinear spring element. This bug is fixed in v. 8.6.5.



Useful Hints and Issues
Regarding the Use of the ADINA System, v. 8.6.2

Issue No. 41, posted March 12, 2010

When a geometrically nonlinear concentrated damper connecting two distinct nodes is used, the results are wrong. The error in damping forces either leads to no convergence, or to convergence with wrong damping forces. In many cases, the error in damping forces will be obvious to the user.

As a workaround, use linear dampers or single-node (node to ground) dampers where possible. This bug is fixed in v. 8.6.4.


Issue No. 40, posted March 12, 2010

The stiffness matrix contribution from concentrated dampers or concentrated time-varying masses is not correct in the second sub-step of the Bathe time integration method. That may lead to a sub-optimal convergence rate or no convergence. Since only the stiffness matrix is incorrect, the results are only marginally affected if the solution converges with tight enough tolerances.

This bug does not affect problems using Newmark time integration, or linear analysis with concentrated dampers and/or masses.

Use Newmark time integration as a workaround. This bug is fixed in v. 8.6.4.


Issue No. 39, posted March 12, 2010

When initial stress is involved in an ADINA restart analysis, the program crashes when the number of equations (NEQ) is changed in the second run. As a workaround, make sure that the number of equations is the same before and after the restart analysis. This bug is fixed in v. 8.6.4.


Issue No. 38, posted March 1, 2010

The contact tractions calculated for quadratic 3-D contact segments are in some cases too high. That occurs for the special case of contact segments in partial contact (only some of their nodes are in contact) that are adjacent to contact segments in full contact. The increase in tractions in the affected segments can be insignificant or can reach a factor of 2.0.

This bug only affects contact tractions. Displacements, stresses, contact forces and other solution variables are not affected. Linear (3-node) and bilinear (4-node) 3-D contact segments are also not affected.

There is no workaround. This bug is fixed in v. 8.6.4.


Issue No. 37, posted March 1, 2010

The option of frequency calculation within an interval with potential-based fluid elements, using the Lanczos method, should have been blocked, it is not available. In v. 8.6.4, this option is blocked.


Issue No. 36, posted March 1, 2010

The results are incorrect when the following combination of options is used: explicit analysis, linear elements in which incompatible modes are used, and restart. To work around this bug, either do not use incompatible modes, or make the element group nonlinear. This bug is fixed in v. 8.6.4.


Issue No. 35, posted March 1, 2010

If an element is born in explicit time integration at a time greater than the solution starting time, and a temperature-dependent material model is used (e.g., thermo-plastic), the program crashes. There is no workaround. This bug is fixed in v. 8.6.4.


Issue No. 34, posted March 1, 2010

When the TLA or TLA-S method is used with a preload bolt, the analysis runs only to 98% of the applied load. The workaround is to scale the applied load by 1.020408 (1.0/0.98). This bug is fixed in v. 8.6.4.


Issue No. 33, posted January 22, 2010

The linear damping element (i.e., specified constant C with parameter NONLINEAR=NO in the PROPERTYSET command), does not work with the element birth/death option. The damping does not have any effect on the solution in this case. The workaround is to set parameter NONLINEAR=YES in the PROPERTYSET command or to use NONLINEAR=MNO-G or NONLINEAR=MNO in the EGROUP SPRING command. This bug is fixed in v. 8.6.4.


Issue No. 32, posted January 4, 2010

When requesting Nastran Op2 or other interface output (i.e., Universal, EnSight output) with the 3D-iterative solver in linear direct step by step dynamic analysis, the program stops in the second solution step and prints an error message: "3D-ITE solver: No convergence reached..."

The workaround is to request a nonlinear solution by making one element group nonlinear (e.g., input a very large death time in any element group) when Op2 or other interface output is needed and the 3D-iterative solver is used. This bug is fixed in v. 8.6.4.


Issue No. 31, posted December 23, 2009

If there are nodes on the cyclic axis in cyclic symmetry frequency analysis, some of the frequencies and mode shapes may be wrong. This only happens when the nodes on the cyclic axis are not fixed or constrained. Nodes on the cyclic boundaries (rather than axis) do not trigger this bug.

The effect on the natural frequencies and mode shapes is generally mild, and the wrong results can be detected by plotting the mode shapes in ADINA-PLOT. Elements on the cyclic axis will overlap if this bug is triggered.

There is no workaround. This bug is fixed in v. 8.6.3.


Issue No. 30, posted December 21, 2009

In the solution of a CFD model when using any turbulence model, using version 8.6.1 and 8.6.2, the calculated reactions are not correct. However, the velocities and stresses are correct. The error is most severe when the wall elements are small. This error has slipped in during the improvement of the turbulence models, specifically the SST model, from version 8.5 to v. 8.6. This bug is fixed in v. 8.6.3.


Issue No. 29, posted November 30, 2009

When free-form meshing is used on multiple face-linked bodies, and if a large number of elements is created (e.g., 2 million or more), then the AUI runs very slowly. There is no workaround except reducing the number of generated elements. This bug is fixed in v. 8.6.3.


Issue No. 28, posted November 30, 2009

In a nonlinear mode superposition analysis with contact, some nodal displacements may be overwritten during analysis. The wrong results will frequently be obvious to the user, in the form of very high local stresses around the immediate vicinity of nodes whose displacements were overwritten. There is no workaround. This bug is fixed in v. 8.6.3.


Issue No. 27, posted November 30, 2009

Prescribed initial conditions (displacements and velocities) are not properly accounted for in mode superposition analysis with contact or glueing. This bug is not triggered for initial accelerations and if the prescribed initial displacements/velocities are zero. The bug is not triggered as well for initial conditions generated from a restart file. The wrong results may not be obvious. There is no workaround. This bug is fixed in v. 8.6.3.


Issue No. 26, posted November 30, 2009

The porthole file in ADINA-PLOT cannot be loaded for mode superposition analysis with periodic (cyclic) symmetry. As a workaround, use full cyclic symmetry, or manually define periodic boundary conditions (create constraint equations between master and slave cyclic boundaries). This bug is fixed in v. 8.6.3.


Issue No. 25, posted November 30, 2009

In the 64-bit Windows version, in some dialog boxes (e.g., the Apply Fixity dialog box) where the body label is selected from a combo box control, the AUI gives the error that 'No label is specified for BODY'. To avoid this problem, click the [P] button and pick the body from the graphics window. This bug is fixed in v. 8.6.3.


Issue No. 24, posted November 30, 2009

The program stops with a message that the stiffness matrix is not positive definite if the gluemesh option is used and if potential-based fluid elements are present in the model. There is no workaround. This bug is fixed in v. 8.6.3.


Issue No. 23, posted November 11, 2009

In the presence of face links, free-form meshing may cause a crash when it tries to update (prior to meshing) the subdivisions of the edges bounding a face that is linked to another face (already meshed). The crash occurs when the mesher cannot find the matching edge and attempts to write an error message.

A possible workaround is to make the tolerance (slightly) larger when setting the facelinks. This bug is fixed in v. 8.6.3.


Issue No. 22, posted November 11, 2009

When the 3D-iterative solver is used with rigid target contact and there is no other type of contact (e.g., constraint function) present in the model, the solver stops with a message that the solution could not converge. As a workaround, use either the sparse solver with the rigid target algorithm or the 3D-iterative solver with the constraint function contact algorithm. This bug is fixed in v. 8.6.3.


Issue No. 21, posted November 11, 2009

This bug involves display of mesh plots using the AUI. Suppose you display a mesh plot using a zone name, say zone CG1. Then suppose that you delete the zone. Then the AUI displays the internal error message "ICNIAA - Meshplot zone name not found".

The workaround depends on the operating system. In Linux, you can click OK and then clear the graphics window. In Windows, you need to kill the process by using the Task Manager.

To avoid the bug, do not delete a zone that is currently displayed by a mesh plot. This bug is fixed in v. 8.6.3.


Issue No. 20, posted November 11, 2009

The reading of an input file will be slow for shell models where the shell director vectors are manually defined by the user at all nodes. This slowdown is most severe for problems when the number of nodes exceeds one million. A reasonable workaround is to let ADINA define the shell director vectors for large problems. This bug is fixed in v. 8.6.3.


Issue No. 19, posted October 16, 2009

The results might be incorrect when the following combination of options is used:

  • plastic-orthotropic or plastic-cyclic material models, and
  • large displacement/small strain formulation, and
  • mixed pressure-displacement interpolation formulation.
As a workaround, use a large displacement/large strain formulation (ULJ for plastic-orthotropic material and either ULH or ULJ for plastic-cyclic material). This bug is fixed in v. 8.6.3.


Issue No. 18, posted October 7, 2009

When there is a restart from implicit time integration to explicit time integration, and the analysis involves contact, shell elements and prescribed displacements, solution results after restart are obviously wrong. There is no workaround. This bug is fixed in v. 8.6.3.



Useful Hints and Issues
Regarding the Use of the ADINA System, v. 8.6.1

Issue No. 17, posted September 15, 2009

When an input file is saved after performing the following operations in the AUI, the input file is not correct if there are multiple contact groups and the actions are not on the current contact group.

  • In the Model Tree, right-click on contact pair (or surface) item and click Define...
  • Modify parameters in a contact pair (or surface) and click OK.

To avoid this problem, users should right-click on a specific contact pair (or surface) and click Modify...

This bug is fixed in v. 8.6.2.


Issue No. 16, posted September 14, 2009

The porthole results file could not be opened for post-processing for a heat transfer only analysis of a TMC model involving initial stresses or initial strains. Since these initial mechanical stresses/strains do not affect the results of a heat transfer only analysis, they can be removed from the model as a workaround. This bug is fixed in v. 8.6.2.


Issue No. 15, posted September 8, 2009

New description added on September 14, 2009.

In 2D axisymmetric turbulent flows, if the axis of symmetry is connected with a no-slip wall or no-slip FSI boundary, the user must apply "fixity" to the normal component of velocity along the center-line (which will override any wall condition). This bug is fixed in v. 8.6.2.


Issue No. 14, posted September 1, 2009

Pressure loads on STL body edges and faces are plotted incorrectly in the 32-bit version of AUI (on Linux and Windows). This bug is fixed in v. 8.6.2.


Issue No. 13, posted September 1, 2009

When BOLT_FORCE or BOLT_SHORTENING is selected for element line plot, the 32-bit version of AUI (on Windows and Linux) crashes. This bug is fixed in v. 8.6.2.


Issue No. 12, posted September 1, 2009

When nodes are defined with respect to a non-global coordinate system and the coordinate system is modified, the coordinates of the nodes are not modified correctly. A redraw of the graphics window will show the wrong positions of the nodes. If it is desired that the nodes move according to the change in the reference coordinate system, the parameter MOVE=YES can be specified with the SYSTEM command to avoid this problem. This bug is fixed in v. 8.6.2.


Issue No. 11, posted September 1, 2009

For an ADINA Structures model, when nodal heat flux is applied to the same node more than once, the AUI program creates a bad data (.dat) file. This will cause the ADINA Structures program to crash when reading the data file. The workaround is to combine all nodal heat flux on the node and specify only one nodal heat flux load on that node. This bug is fixed in v. 8.6.2.


Issue No. 10, posted August 27, 2009

In a text-formatted ADINA-F data (.dat) file, if the number of nodes with prescribed loading is more than 99999, the file will be incorrect. As a workaround, use a binary data file instead: in the ADINA User Interface, ADINA CFD program module, click the Data File/Solution icon, and uncheck the "Create Text Data File" box; or using the command line, set FORMATTED=NO in the ADINA-F command of ADINA CFD. This bug is fixed in v. 8.6.2.


Issue No. 9, posted August 21, 2009

In an ADINA restart analysis, when prescribed displacements are replaced by prescribed forces to model unloading and the unloading is prescribed before restart, the iteration history after the restart is different than if there had been no restart. If convergence is reached the results will be almost the same. There is no workaround other than avoiding this restart. This bug is fixed in v. 8.6.2.


Issue No. 8, posted August 10, 2009

Mass proportional (gravity) loading is not applied to elements with a non-zero birth time in a restart analysis. This only happens to elements born after the restart. As a workaround, the analysis can be performed without restarts, if possible. This bug is fixed in v. 8.6.2.


Issue No. 7, posted August 10, 2009

If prescribed displacements are used in ADINA TMC solutions, then the loading might be incorrect, in particular for 32-bit versions. There is no workaround. This bug is fixed in v. 8.6.2.


Issue No. 6, posted August 10, 2009

In explicit time integration, we have seen models in which high frequency vibrations cause unphysical plasticity to occur. As a workaround, only use large strain shell elements in those areas of the model in which large strains are likely to occur, or use implicit time integration. This bug is fixed in v. 8.6.2.


Issue No. 5, posted August 10, 2009

In thermo-mechanical coupling (TMC) analysis, thermal constraint equations (master-slave relationships) may not be properly enforced if there are more thermal than structural degrees of freedom. There is no workaround. This bug is fixed in v. 8.6.2.



Useful Hints and Issues
Regarding the Use of the ADINA System, v. 8.6.0

Issue No. 4, posted July 10, 2009

In cyclic symmetry analysis, when using the 4-node shell element with incompatible modes the stresses are not correct in some cases. They are always correct in the fundamental cyclic part. This issue does not affect displacements, frequencies and mode shapes. There is no workaround. This bug is fixed in v. 8.6.1.


Issue No. 3, posted July 10, 2009

In cyclic symmetry analysis, if shell elements lie completely on the cyclic boundary their mass matrix is double the correct value (only for these shell elements). This only affects the transient dynamic cyclic symmetry analysis. As a workaround, divide the density of these shell elements by 2. This bug is fixed in v. 8.6.1.


Issue No. 2, posted July 10, 2009

ADINA sometimes does not converge when a generalized constraint is connected to nodes with no attached elements. This does not affect generalized constraints automatically created for mesh glueing. There is no workaround. This bug is fixed in v. 8.6.1.


Issue No. 1, posted July 10, 2009

Wrong initial temperatures are applied in ADINA Thermal if an initial temperature condition is defined together with a non-zero default initial temperature. This should be obvious from the results. As a workaround, set the default initial temperature to zero. This bug is fixed in v. 8.6.1.



For Useful Hints and Issues regarding the use of earlier versions of the ADINA System, see here

Version 8.5            Version 8.4