RISA-3D FAQ
Common support and usage questions.
RISA-3D
Why is RISA-3D locking degrees of freedom?
When RISA-3D solves a model, it looks for degrees of freedom that have zero or “very little” stiffness. Since all degrees of freedom must have some stiffness, RISA-3D will lock these degrees of freedom so that it can finish the model solution. When the model solution is finished, a window will “pop up” and report all the locks that RISA-3D had to apply to solve the model. Sometimes these locks are trivial (such as the case where a member can spin about it’s own axis, caused by AllPin’s at both ends), and sometimes these are quite serious. The locked degrees of freedom should always be studied to determine if the original model was unstable. It’s a good habit to always resolve any locked degrees of freedom by modifying your Member End-Releases or Nodal Boundary Conditions.
What is the difference between Member Releases & Boundary Conditions?
Probably the greatest source of confusion to people doing stuctural modeling is the difference between a Member Release and a Boundary Condition. A Boundary condition describes how a node is attached to the external world around it. The boundary condition says whether or not a NODE can translate or rotate. In contrast, a Member Release describes how a BEAM element is attached to a NODE. The member release says whether or not the beam to node connection will transfer shear, moment, or axial forces. In application, a boundary condition applied to a node would be used to model a fixed or pinned support condition, while a member release applied to a beam element would be used to model a fixed or pinned beam to column connection.
Why do some of my members “flip” over when a R2D model is read into R3D?
The way the local-y axis of a member is defined is different between RISA-2D and RISA-3D. Thus, members in RISA-2D models may get turned over 180 degrees when read in RISA-3D. Any member distributed loads or point loads are automatically adjusted back to their original direction (i.e., A member load that was pointing down which is applied to a member that gets flipped over will still point down). The only noticeable affect this will have is that the sign of the shears shown in RISA-3D for any “flipped” members will be opposite that shown in RISA-2D.
Why is my P-Delta analysis diverging?
See the P-Delta topic in the Modeling Tips section of the manual for a more complete explanation of this topic. By far, the most common cause of P-Delta convergence problems is local instabilities. (See the Modeling Tips section of the manual for information on dealing with local instabilities.) If you are trying to model P-Delta effects on a 2D frame, you will want to make sure that you restrain the out-of-plane degrees of freedom. This is most easily accomplished using the ALL code on the Boundary Conditions screen. In some cases, a model may be so flexible that it is not possible to run a P-Delta analysis.
How is Warping Torsion in RISA-3D calculated?
RISA-3D’s torsion calculations are based on the AISC publication Torsional Analysis of Steel Members. The program currently only models warping fixed conditions. The program does not model a warping pinned condition. (A warping pin is restrained against twist, but free to warp) RISA-3D models all warping conditions as a CASE 2 as shown in the AISC publication. This is correct for columns of buildings, and any other case where torsional loads are applied only at the ends of the member.
For the cases where internal torque loads are applied as nodal torques, the solution given by RISA-3D will be conservative for the maximum warping stresses at the ends of the members. If an accurate solution is required (as opposed to a conservative solution) for a point torque or a distributed torque on a member, then it is recommended that the engineer combine RISA-3D’s direct stresses with the warping stresses obtained from a hand solution. The above mentioned AISC publication on torsion contains charts for various boundary conditions and loading conditions to make the hand solution much faster.
What are the Lb and Lcomp fields on the Hot Rolled Steel Design Parameters screen?
These fields are the unbraced lengths for a member and control the AISC code checking that RISA-2D/3D can perform on AISC hot rolled steel shapes. For a detailed description, see the appropriate section in the program manual. In a nutshell, the Lb values (unbraced length) control the calculation of the allowable axial stress (Fa), while the Lcomp value (unbraced compression flange length) controls the calculation of the allowable bending stress (Fb).
These fields are provided because in many cases the actual unbraced lengths are different than the member lengths in a structural model. For example, a column with K-bracing framing in on one side only would need to be composed of two model members, yet the unbraced length out of the plane of the bracing would be the full column height.
How are the K value approximations calculated by RISA-2D/3D?
RISA-3D is able to approximate the K values for a member based on that member’s sway condition and end release configuration. The K-factor approximation is based on Table C-C2.1, found on page 5-135 of the 9th Edition ASD code, or page 6-184 of the 3rd Edition LRFD code. A full description of the limitations of the K approximation are given in the Hot Rolled Steel – Design section of the manual.
Why can’t I get an AISC steel unity check for certain members in my model?
The usual things to look for are as follows : Is the shape you using for the member a database shape? RISA only does steel code checks for database shapes and the online shapes. Is the shape for the member in question an Arbitrary database shape? RISA cannot do code checking for sections defined with the Arbitrary shape. For the 3rd Edition LRFD code and 13th Edition ASD/LRFD, you must also do a P-Delta analysis to get code checks.
Why can’t I get a NDS wood unity check for certain members in my model?
RISA only does wood unity checks for sections that are defined in the Wood database. You must also specify the wood species on the Wood tab of the Materials spreadsheet for the section that you want wood checks for. See also the Wood – Design section of the manual for a step by step procedure on how to get wood code checks.
How do I model the situation where one member is crossing over another?
See the related topic in the Modeling Tips section of the manual for a more complete explanation. In general, you would model the two members at their actual centerline locations. You would then use a rigid link or a slave node to connect the upper and lower nodes at the point where the two members would intersect.
How do I model a composite section using beam members and the plate/shell elements?
See the related topic in the Modeling Tips section of the manual for a more complete explanation. In general, you would model the beam and the plates at their actual centerline locations. You would then use rigid links to connect the beams to the plates.
Why can’t I use nodal slaving to build a rigid diaphragm?
It is not accurate to use nodal slaving to try to create a diaphragm. While the nodal rotations can be slaved and the correct diaphragm behavior maintained, slaving the in-plane translational degrees of freedom will produce incorrect diaphragm behavior. A diaphragm that has a load applied to a location other than the center of stiffness should experience both a translation and a rotation. A diaphragm that is created by slaving in-plane translations will not rotate under such a loading, it will only translate, and be much stiffer than it should be.
How do I tell if my Plate/Shell Finite Element Results are accurate?
Assessing the accuracy of a finite element (FE) model is not a simple thing because it requires an understanding of how finite elements work, as well as some practical experience in FE modeling. The underlying principle is that the model should be able to accurately represent the deflected shape of the “real” structural element being modeled. As a simple example, a single 4 node quadrilateral finite element cannot be used to model a floor slab loaded out of plane. The reason for this is that the single element cannot accurately represent the true deflected shape of the slab.
A useful rule of thumb is as follows: Build a model and get results. Then submesh the model into smaller elements and compare the new results to the previous results. If the results don’t change by more than 10%, you’re probably close to the correct solution. This rule of thumb is not fool-proof. It is possible to build a bad model to start with and then have the submeshed model still be a bad model. If you’re in doubt about a particular model’s accuracy, you can always send us your model via email and the staff here at RISA will be able to give you some comments on the model.
Why can’t I get a Deflected Plot or a Member Detail Report when I run an Envelope Solution?
When you get results from an envelope solution, the results are often from different load combinations at different locations on the same member. This means that a deflected shape would be very discontinuous and wouldn’t really have any physical meaning. As for the Member Detail report, we may add the ability in a future release for the Member Detail report to show the enveloped forces and unity checks.
What’s the fastest way to run RISA-3D?
The fastest way to run RISA-3D is with lots of RAM. The more RAM you have, the better chance you have of keeping your whole stiffness matrix off of your hard drive during the solution. If you have large models to run, it’s generally more cost effective to buy a slower processor, but load up on all the RAM you can.
Why can’t I see the Add button on the WindowTool Bar?
The RISA toolbars are sized for a screen resolution of 1024x768. If some of the toolbar is chopped off then your screen resolution is probably set to 800x600 or less. To correct this, you should change your screen resolution to 1024x768 or better. This can generally be done within the Display setting in the Windows Control Panel.
Why doesn’t the arbitrary section created in RISASection appear in the Shape Selection database?
There are a couple of common mistakes people make when attempting to access RISA Section files in RISA 3D:
- Check the File Location settings for RISA Section Files under Tools-Preferences. When RISA 3D launches, it will search for RISA Section files ONLY in the directory listed in Tools-Preferences. If your files are stored in another directory, they will not be read into RISA 3D.
- RISA 3D only reads in Section files when it is first launched. Therefore, you must close and Re-start RISA 3D after creating the file in RISA Section.
- The problem could be with how you’ve saved the file in RISA Section. Check the FAQ list for RISA Section for further information.
Why am I not getting 90% mass participation for my Response Spectrum Analysis?
See the topic Dynamics Convergence – General in the Modeling Tips section of the manual for a more complete explanation of this topic. In general, you may need to just run more modes to get 90% mass participation. If you’ve already tried more modes, then you would need to look at how many local modes you are getting as opposed to global modes. Another thing to look at is how much of your applied mass is getting lost into your boundary conditions. (i.e., is your ammount of “active” mass 90% of your total mass?)
Why can’t I get more than xx modes when I run my Dynamic Analysis?
See the topic Dynamics Convergence – General in the Modeling Tips section of the manual for a more complete explanation. Usually, the cause is simply not having enough mass degrees of freedom that are free to vibrate in the model. Also, if some of your modes are much stiffer than others – like the axial mode of a column, compared to the flexural modes – you can also get this message.
Why does the Deflected Plot from a Response Spectrum Analysis look “strange”?
The results from any Response Spectrum Analysis (RSA) will all be positive. This is due to the modal combination methods used to obtain the RSA results. Since the results are all positive, ( instead of being positive and negative ) the deflected plot will be discontinuous at locations where the result would normally have been negative if it had been computed by a regular static analysis.
To get a better feel for this, you can compare the deflected shapes of two simple models. Take a single bay portal frame subjected to a lateral load applied to one of the corners and perform a static analysis. Then subject the same model to dynamic loading via an RSA. If you zoom in on the results for the RSA, you will see that both columns are going UP, instead of one column going up and one going down. You should see a discontinuity at the location where the column in the statics solution would be going down.
What are the hardware requirements for RISA-3D?
Minimum
- Any Windows compatible computer with a Pentium 3 or better processor
- Windows 2000\XP\Vista
- 256 MB of RAM
- 200 MB of hard disk space
- Two or three button mouse
- USB port (required for Stand-Alone version or the Network Host computer)
Recommended
- Windows XP
- As much extended RAM as possible
- As much free disk space as possible
- Two button mouse with wheel
Structural Engineering Software, Structural Analysis Software, Structural Design Software
©2010 RISA Technologies, LLC. All Rights Reserved.
949.951.5815 | E-mail RISA Technologies, LLC | Privacy Policy