RISA-3D

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. (ie., A member load that was pointing down which is applied to a member that gets flipped over will still point down ) The only noticable 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.

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.

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.

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.

What are the "Lb" and "Lc" fields on the AISC 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 "Lc" 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.

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 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 so flexible that it is not possible to run a P-Delta analysis.

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?)

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 the 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.

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.

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 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.

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 ASD code, or page 6-184 of the LRFD code. A full description of the limitations of the "K" approximation are given in the "Steel Design and Code Checking" section of the manual.

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.

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 Steel code check flag on the Global screen (Alt-G) set to "2" or "9" ? 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 LRFD code, 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 Parameters screen (Alt-W). You must also specify the wood parameter label in the Material/Wood Label field on the Sections screen (Alt-S) for the section that you want wood checks for. See also the "Timber Design and Code Checking" section of the manual for a step by step procedure on how to get wood code 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. (A current example might be going with a P3 500 Mhz system with 256The 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. (A current example might be going with a P3 500 Mhz system with 384 Mbytes of RAM, rather than a P3 1 Ghz system and 64 Mbytes or 128 Mytes) Mbytes of RAM, rather than a P3 1 Ghz system and 64 Mbytes or 128 Mytes)

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.
1) 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.
2) 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.
3) The problem could be with how you've saved the file in RISA Section. Check the FAQ list for RISA Section for further information.

RISAFloor

Why are my Camber values coming out lower than I expected?

The Solution Tab on the Global Parameters screen lists the various camber options. The "%DL for Camber" specifies what percentage of self weight to use to camber the beam. This includes the self weight of the beam and the deck, but does NOT include "DLPre" area loads or "ConstDL" deck loads.

Why isn‘t my DXF file importing?

Most DXF problems have to do with the way you've created your columns. When creating your DXF file it is essential to specify a column layer and to correctly define your columns. Only columns that are defined as "Blocks" will be recognized by the DXF import feature. The help documentation for your CAD systems should have more information on creating/defining blocks.

RISA-2D

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. (ie., A member load that was pointing down which is applied to a member that gets flipped over will still point down ) The only noticable 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 am I getting instability errors in RISA-2D?

RISA-2D will report an instability error when degrees of freedom in a model have zero or very little stiffness. A common case where this occurs is when all members attached to a node have the end rotation released (e.g., This would happen when trying to model a truss). If there are no external restraints to give the node rotational stiffness, RISA-2D will report an instability error. One way to solve the problem in this case would be to connect ONE member rigidly to the joint. Another way would be to apply a rotational boundary restraint to the node. (e.i., enter a "R" or "F" in the rotation field for that node)

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.

What are the "Lb" and "Lc" fields on the AISC 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 "Lc" 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.

Why don't my Reactions satisfy statics when I run a Response Spectrum Analysis?

The applied loads in a response spectrum analysis (RSA) are the applied masses times the accelerations in the response spectra. The modes of vibration, each of which contain a certain percentage of the mass, are combined using a modal combination method. Unless you use the "Dominant Mode" feature, any modal combination method will cause the results to be all positive. Because all the results are positive, statics will no longer satisfied. As a simple example, consider a symmetric portal frame, subjected to a lateral nodal load applied at one of the top corners. The net upward reactions should be zero and the net lateral reactions should equal the applied force. The vertical deflections of the top two nodes should have equal magnitudes, but opposite signs. Now, if the signs of the vertical deflections were forced to be same ( as would happen in an RSA ), you would get a net upward or downward force, instead of having the upward and downward forces in the opposite columns canceling each other out.

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.

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 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.

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 ASD code, or page 6-184 of the LRFD code. A full description of the limitations of the "K" approximation are given in the "Steel Design and Code Checking" section of the manual.

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.

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 Steel code check flag on the Global screen (Alt-G) set to "2" or "9" ? 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 LRFD code, 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 Parameters screen (Alt-W). You must also specify the wood parameter label in the Material/Wood Label field on the Sections screen (Alt-S) for the section that you want wood checks for. See also the "Timber Design and Code Checking" section of the manual for a step by step procedure on how to get wood code checks.

RISAFoot

My footing has uplift, why isn't RISAFoot providing top steel?

Version 2.0a of RISAFoot calculates the negative moment caused by overburden and slab self weight. If that moment exceeds the flexural strength of the plain concrete section (per section 22.5 of ACI),

RISABase

Why don't my RISA Section shapes appear in the RISA 3D shape selection database?

There are a few common mistakes people make when attempting to access RISA Section files in RISA 3D.

Common RISA Section Mistakes:
1) Don't confuse the name of the RISA Section "File" with the name of the "Section". The name of the "section" is what will appear in the RISA 3D database. Note: each file can contain many different sections. Hence the need to name each individual section.

2) Make sure the Sections names are unique. Since RISA Section always uses default names of "Section 1, Section 2...et cetera", you will probably have to rename your sections before saving your file. You can do this by clicking Section - Rename Sections from the Main Menu Toolbar (within RISA Section, not RISA 3D).

Common RISA 3D Mistakes:
1) 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.

2) 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.