Analysis Settings

Tekla Structural Designer
2021
Tekla Structural Designer

Analysis Settings

The Analysis Settings page and its subpages allow you to control the analysis settings applied to:

  • the current project - when accessed from Settings on the Analyze ribbon tab.
  • new projects - when accessed from Settings on the Home ribbon tab.

1st order non-linear settings

Button, command, or option Description

Maximum number of iterations

The number of iterations to perform. Default = 100.

Tolerance

At the end of each iteration, values are checked against the previous iteration results. If the difference is less than the tolerance determined here, the result is converged, and the analysis is complete. Default = 0.0001.

Relative

Default: ON.

Relaxation Factors

Relaxation factors control an amount of nominal compression tension-only braces can undergo while remaining active during analysis iteration. This improves convergence for rare problematic models/ loading situations in which otherwise most or all braces may experience compression and become inactive, causing instability and preventing solution. When converged tension-only braces will still have only either zero or tension force. The process is entirely automatic by default and it is anticipated the factors will not require manual editing for most circumstances.

  • Use relaxation factors for tension only elements: Default = cleared
  • Minimum relaxation factor: Default = 0.1
  • Maximum relaxation factor: Default = 0.5

If the model still fails to converge when using relaxation factors, expanding the allowable range (by changing the above minimum and maximum values) may produce convergence.

Note:

Relaxation factors are not activated by default, as it is anticipated they will not be required for the majority of models. In addition, since the solution process is necessarily more complex, their use can increase the analysis time to a degree.

2nd order non-linear settings

Button, command, or option Description

Maximum number of iterations

The number of iterations to perform. Default = 100.

Tolerance

At the end of each iteration, values are checked against the previous iteration results. If the difference is less than the tolerance determined here, the result is converged, and the analysis is complete. Default = 0.0001.

Relative

Default: ON.

Relaxation Factors

Relaxation factors control an amount of nominal compression tension-only braces can undergo while remaining active during analysis iteration. This improves convergence for rare problematic models/ loading situations in which otherwise most or all braces may experience compression and become inactive, causing instability and preventing solution. When converged tension-only braces will still have only either zero or tension force. The process is entirely automatic by default and it is anticipated the factors will not require manual editing for most circumstances.

  • Use relaxation factors for tension only elements: Default = cleared
  • Minimum relaxation factor: Default = 0.1
  • Maximum relaxation factor: Default = 0.5

If the model still fails to converge when using relaxation factors, expanding the allowable range (by changing the above minimum and maximum values) may produce convergence.

Note:

Relaxation factors are not activated by default, as it is anticipated they will not be required for the majority of models. In addition, since the solution process is necessarily more complex, their use can increase the analysis time to a degree.

1st order modal settings

Button, command, or option Description

Extraction Method

Allows you to select the appropriate extraction method for your model. The options are:
  • Automatic (default option): Initially uses the Subspace option to find the lowest modes. If the criteria (either mass or number of modes) is not fulfilled, FEAST is then automatically used to find higher modes until the stopping criteria is fulfilled.
  • Jacobi: An iterative transformation method used to calculate all eigenvalues and eigenvectors. Suitable for small models, but not for medium to large models.
  • Subspace: An iterative simultaneous vector method to calculate the smallest eigenvalues and the corresponding eigenvectors. Suitable for quickly finding the lowest frequencies in medium to large models.
  • FEAST: uses the FEAST algorithm to effectively calculate all the eigenvalues within a specific range. Suitable for any size structure. For more information, see: http://www.ecs.umass.edu/~polizzi/feast/.

Mass Model

Allows you to select how the mass model of the analysis is displayed. The options are:
  • Consistent
  • Lumped
Mass reporting

For users that wish to determine a center of mass for each floor, the Simple mass option is available to simplify the calculation. A single value of mass for each node is listed in the Tabular Data for both the "Active Masses by Node" & "Total Masses by Node" view types.

The simple mass is calculated by averaging the mass values in the active directions.

Note: It is recommended that the simple mass option is used only when the lumped mass model is used.

Stopping Criteria

Stopping criteria prevent analysis continuing forever. If either of the criteria (Maximum number of modes or Stopping Frequency) are met, the analysis will not look for any more modes.

Modes

Allows you to select one of the following options:
  • Automatic number of modes: If the option is selected, you must specify the mass participation required in each direction. You can optionally specify an initial number of modes that should be close to the actual number required in order to speed up the analysis process.
  • Total number of modes: If Automatic number of modes is cleared, you can specify the total number of modes required. The default value is 10.

Jacobi

Allows you to adjust the following settings:
  • Maximum number of sweeps: A sweep is a transformation of every off-diagonal in the global matrices. The option allows you to set the maximum number of sweeps allowed.
  • Sweep tolerance: At the end of each sweep, values are checked against the previous sweep results. If the difference is less than the tolerance specified here, the result is converged, and the analysis is complete.

Subspace

Allows you to adjust the following settings:
  • Maximum number of iterations: The number of iterations to perform.
  • Tolerance: At the end of each iteration, values are checked against the previous iteration results. If the difference is less than the tolerance specified here, the result is converged, and the analysis is complete.

FEAST

Allows you to adjust the following settings:
  • Initial search range: Specifies the initial range of values FEAST will search for eigenvalues in.
  • Overestimation multiplier: Specifies the initial guess for the subspace dimension within each range; an overestimate of the predicted number of modes in the range.
  • Maximum modes in range: The maximum number of modes in the range. If more modes are found in a range, the range is split into several smaller ranges.
  • Minimum search range: When a range is smaller than the value specified here, it will no longer be split, even if the maximum number of modes is greater than that allowed.

2nd order buckling settings

Button, command, or option Description

Maximum number of iterations

The number of iterations to perform. Default = 1000.

Tolerance

At the end of each iteration, values are checked against the previous iteration results. If the difference is less than the tolerance determined here, the result is converged, and the analysis is complete. Default = 0.00001.

Maximum number of sweeps

A sweep is a transformation of every off-diagonal in the global matrices. The option allows you to set the maximum number of sweeps allowed. Default = 50.

Sweep tolerance

At the end of each sweep, values are checked against the previous sweeps results. If the difference is less than the tolerance determined here the result is converged, and the analysis is complete. Default = 1.0E-12.

Total number of modes

Allows you to specify the total number of modes required. Default = 10.

Show negative buckling factors

Default = OFF.

Extraction Method

Allows you to select the appropriate extraction method for your model. The options are:
  • Jacobi: An iterative transformation method used to calculate all eigenvalues and eigenvectors. Suitable for small models, but not for medium to large models.
  • Subspace: An iterative simultaneous vector method to calculate the smallest eigenvalues and the corresponding eigenvectors. Suitable for quickly finding the lowest frequencies in medium to large models.
  • Automatic (default option): Tekla Structural Designer determines the most appropriate extraction method for the structure.

1st order seismic settings

Button, command, or option Description

Extraction Method

Allows you to select the appropriate extraction method for your model. The options are:
  • Automatic (default option): Initially uses the Subspace option to find the lowest modes. If the criteria (either mass, or number of modes) is fulfilled, FEAST is then automatically used to find higher modes until the stopping criteria is fulfilled.
  • Jacobi: An iterative transformation method used to calculate all eigenvalues and eigenvectors. Suitable for small models, but not for medium to large models.
  • Subspace: An iterative simultaneous vector method to calculate the smallest eigenvalues and the corresponding eigenvectors. Suitable for quickly finding the lowest frequencies in medium to large models.
  • FEAST: uses the FEAST algorithm to effectively calculate all the eigenvalues within a specific range. Suitable for any size structure. For more information, see: http://www.ecs.umass.edu/~polizzi/feast/.

Mass Model

Allows you to select how the mass model of the analysis is displayed. The options are:
  • Consistent
  • Lumped
Mass reporting

For users that wish to determine a center of mass for each floor, the Simple mass option is available to simplify the calculation. A single value of mass for each node is listed in the Tabular Data for both the "Active Masses by Node" & "Total Masses by Node" view types.

The simple mass is calculated by averaging the mass values in the active directions.

Note: It is recommended that the simple mass option is used only when the lumped mass model is used.

Stopping Criteria

Stopping criteria prevent analysis continuing forever. If either of the criteria (Maximum number of modes or Stopping Frequency) are met, the analysis will not look for any more modes.

Modes

Allows you to modify the following options:
  • Initial number of modes: In order to speed up the analysis process, you can specify an initial number of modes you expect to be required to achieve the required participation. The value specified here should be close to the actual number required because if you enter too few or too many modes, the analysis may take longer.
  • Mass participation for RSA: Allows you to specify the mass participation required in each direction. If this isn't achieved before the stopping criteria apply, the RSA analysis will still be performed, but a warning will be issued.
  • Min. Mass participation for RSA: If the minimum participation isn't achieved before the stopping criteria apply, the RSA analysis is not performed.
Note:

The options in Stopping Criteria overrule both the number of modes and mass percentage.

Jacobi

Allows you to adjust the following settings:
  • Maximum number of sweeps: A sweep is a transformation of every off-diagonal in the global matrices. The option allows you to set the maximum number of sweeps allowed.
  • Sweep tolerance: At the end of each sweep, values are checked against the previous sweeps results. If the difference is less than the tolerance determined here the result is converged, and the analysis is complete.

Subspace

Allows you to adjust the following settings:
  • Maximum number of iterations: The number of iterations to perform.
  • Tolerance: At the end of each iteration, values are checked against the previous iteration results. If the difference is less than the tolerance determined here, the result is converged, and the analysis is complete.

FEAST

Allows you to adjust the following settings:
  • Initial search range: Specifies the initial range of values FEAST will search for eigenvalues in.
  • Overestimation multiplier: Specifies the initial guess for the subspace dimension within each range; an overestimate of the predicted number of modes in the range.
  • Maximum modes in range: The maximum number of modes in the range. If more modes are found in a range, the range is split into several smaller ranges.
  • Minimum search range: When a range is smaller than the value specified here, it will no longer be split, even if the maximum number of modes is greater than that allowed.

Modal Combination Method

To determine the representative maximum response of interest for a loadcase, the relevant values for each relevant mode are combined by using the modal combination method specified here. Note that once modes have been combined, the relative signs are lost. The options are:
  • Complete Quadratic Combination (CQC): Suitable for models where modes are closely spaced or well spaced.
  • Square Root of Summation of Squares (SRSS): Suitable only for models where modes are well spaced.

Iterative cracked section analysis settings

Button, command, or option Description

Global Convergence Criteria

Allows you to modify the following options:
  • Maximum number of iterations: The number of iterations to perform. Default = 200.
  • Tolerance: At the end of each iteration, values are checked against the previous iteration results. If the difference is less than the tolerance determined here, the result is converged, and the analysis is complete. Default = 0.00100.
  • Relative: Default: ON.
Tip:
To resolve global convergence failures:
  • Increase the value in Maximum number of iterations.
  • Decrease the value in Tolerance.

Local Convergence Criteria

Allows you to modify the following options:
  • Maximum number of iterations: The number of iterations to perform. Default = 500.
  • Tolerance: At the end of each iteration, values are checked against the previous iteration results. If the difference is less than the tolerance determined here, the result is converged, and the analysis is complete. Default = 0.000001.
  • Smoothing Parameter: A property only available for the Eurocode head code. Default = 0.005.
Tip:
To resolve local convergence failures:
  • Increase the value in Maximum number of iterations.
  • Decrease the value in Tolerance.
  • If you are using the Eurocode head code, increase the value in Smoothing Parameter.

Modification factors

Different factors can be applied for each of the different materials in the model in order to adjust the following properties:
  • E: Young's modulus
  • G: Shear modulus
  • Itorsion: Section inertia about local X
  • Imajor: Section inertia about local Y
  • Iminor - Section inertia about local Z
  • Area: Section area in compression or tension
  • Aparallel to minor: Section shear area in local Y
  • Aparallel to major - Section shear area in local Z
  • t: shell thickness (applicable to concrete only)

The factors also vary according to the member types, (and in the case of concrete members whether they are cracked or not).

For concrete members in particular, design codes can require that analysis stiffness adjustment factors are applied since the appropriate properties to use in analysis are load and time dependent.

For various other investigations it is also possible that you will want to apply an adjustment to material properties. One suggested example is the assessment of structures subject to corrosion. Another classic requirement in this regard relates to torsion, it is common engineering practice to assume that if it will work without assuming any torsional strength, then torsion can be ignored.

Although default modification factors for each material are provided in the settings sets to reflect the design code being worked to, you should check that these are appropriate for your particular analysis model.

Note:

If you make changes to any of the modification factors, analysis must be repeated.

Meshing settings

The meshing options can be adjusted in order to fine-tune the meshes that are produced in slabs and walls. The default settings are generally appropriate, but they could potentially be adjusted if 2D element quality errors are being created during validation.

Composite steel beams settings

Button, command, or option Description

Inertia used for loading analysis list

Allows you to specify the inertia to be used in the global analysis of the model for a composite steel beam:
  • Steel beam (default option)
  • Long term composite
  • Short term composite
Note:
  • If the model still fails to converge when using relaxation factors, expanding the allowable range (by changing the above minimum and maximum values) may produce convergence.

  • Relaxation factors are not activated by default as it is anticipated they will not be required for the majority of models. In addition, since the solution process is necessarily more complex, their use can increase the analysis time to a degree.

The inertia has different effects on different head codes:
  • US head code:

    Long and short term composite inertia only apply to rolled symmetric I sections without web openings. For all other sections, the standard steel beam inertia is used regardless of the analysis option selected.

  • EC head code:

    Long and short term composite inertia only apply to:
    • Symmetric I rolled sections without web openings
    • I plated sections without web openings
    • Westok plated sections without web openings
    • Fabsec sections without web openings

    For all other sections, the standard steel beam inertia is used regardless of the analysis option selected.

  • BS head code:

    Long and short term composite inertia only apply to:
    • Symmetric I rolled sections without web openings
    • Asymmetric I rolled sections without web openings
    • I plated sections without web openings
    • Westok plated sections without web openings
    • Fabsec sections without web openings

    For all other sections, the standard steel beam inertia is used regardless of the analysis option selected.

  • AUS and IS head codes:

    Design of composite beams is not currently supported for the head codes.

Was this helpful?
Previous
Next