VisualAnalysis icon
Upgrade to VisualAnalysis 17.0 today. Here's why.

VisualAnalysis 12.0 Help

Analyzing for Design

Requires: Design Level

There are a number of considerations you should weigh carefully when deciding what to analyze and what options to use before you begin or complete your design.

Accuracy vs. Performance

Approximations

VisualAnalysis is an approximate, numerical tool for analysis. It uses the finite element method to calculate displacements and forces. The accuracy of the analysis is only as accurate as your model, and even then there are some approximations and simplifications made. There is no concept of 'load path' in a finite element analysis.

For example, member internal forces are calculated at discrete points along the member. If the moment varies in a nonlinear fashion and the number of calculation points is small, peak moments may be missed.

Control

VisualAnalysis lets you control the number of places along members where results are calculated. If you increase these numbers your results will be more precise whereas if you decrease them you will get better performance from the software.  VisualAnalysis by default will do a fairly good job of adjusting these values based on the size of your project. If you want to control it yourself, use Analyze | Performance vs Accuracy.

Performance Tip

If you have many load cases or many members, design checks can be relatively slow.  You may wish to operate in a "preliminary" design mode for a while, using just a few "key" load combinations, or adjusting the performance settings to "Fast" until you have essentially completed your design.  At that point you can change the load case settings or adjust the performance settings in order to get the most accurate results and design.

Inspect Your Results Carefully

Before design checks are made, you should carefully check the analysis results you have received from VisualAnalysis. If you have large displacements or rotations, running the design software may yield equally erroneous results. The software is all based on "small deflection theory"; so large results are usually garbage results!

In some cases you will see reasonable displacements, but member stresses may be larger than yield stresses for the materials. In other cases, buckling loads may have been exceeded, and for a first-order linear analysis they are not detected or flagged during analysis. This condition will likely be "caught" by the design checks, but you might more efficiently correct this problem in the model.

All member forces checked and reported in VisualAnalysis are with respect to the member element's local axes

It is recommended that you use the Analyze | Result Validation Checks command before starting the design process.

Analyzing Non-Design Load Cases

Load combinations that are not marked for design checks are simply ignored by the design software. Look at Load Case Manager on the Combinations tab for design settings. Service load cases are never checked for design.

In certain situations individual load cases will not analyze because they are unstable on their own.  This will happen for example if you have an 'overturning' load case which is unstable when analyzed separately from the self-weight of the structure.  By not analyzing load cases that are not required for design, you can improve the performance of the software and reduce unnecessary report output.

Frame Instability Analysis

Some of the design modules, like LRFD steel, expect that you have performed a frame-instability or second order (P-Delta) analysis. The AISC Direct Analysis Method is also an option. 

If you run only a first order analysis and the design software is expecting a second order analysis, you will need to define an approximate moment-magnification "B" factor in the steel LRFD design modules.

Design for Dynamic Analysis

VisualAnalysis will check results from a Result Superpostion case.  Dynamic Response analysis, Dynamic Time History analysis, moving load cases, and static load cases can be combined in a Result Superposition combination to produce "Envelope" resulst to be used for design. Because these are enveloped results, the Cb design parameter can not be calculated and is conservatively taken as 1. Also see notes here: Loading For Design.

Another “normal” approach for doing seismic design is to come up with static seismic forces to apply in a Service Load Case using some method acceptable to your building code. For example, use the Dynamic Response Spectra analysis, and then take the reported base shears from that analysis and distribute them back into the structure for a static analysis.