Why won't my analysis run when using pseudo-flexible diaphragms in RAM Frame?

The important thing to remember about the pseudo flexible diaphragm in RAM Frame is that it is no diaphragm at all. The behavior is exactly the same as if you set the diaphragm to "flexible", the program simply attributes the nodal loads for you when using the pseudo flexible option (based on the distribution percentages set under loads - pseudo flexible diaphragm properties). Consequently, if there is any portion of a nodal load acting normal to a 2D frame, because the frames are not parallel to the global axes, that will cause out-of-plane bending in the columns. An instability will occur if the columns are pinned or if the lateral wall out-of-plane stiffness is not considered under Criteria - General. Remember, gravity members are ignored in the RAM Frame analysis so you can't depend on them to stabilize the frames.

Fixing the columns against out-of-plane rotation will solve the instability (you could also assign column fixity along with a rotational spring for less resistance).  Linking frames together with additional lateral members (beams or horizontal braces) might also work.  Alternatively, model nodal loads in the Modeler - View - Elevation View making sure the resultant force is exactly in the plane of the 2D frame and then use the flexible type diaphragms in RAM Frame.  Note, Pseudo-flexible diaphragm percentages do no need to add up to 100%,  You could always assign 0% to the skewed frame and design it in another program. 

When it comes to calculating mode shapes, the program applies a tributary potion of the mass to the nodes of each frame. But for the same reason mentioned above, the calculated periods could become quite long if the frame is capable of free vibration in the out-of-plane direction. An eigensolution error is also likely when there is a lack of stiffness somewhere in the model. For more on eigensolution errors, refer to the topic below:

RAMSS Eigenvalue Error

In cases like this, I generally recommend running the model assuming a rigid diaphragm, record the principal period for the X and Y direction from the loads and applied forces report (or the periods and modes report), then modify the seismic load (or wind load) using these period values explicitly.

An alternative to using pseudo-flexible diaphragms is to use semi-rigid diaphragms. For more information on that, refer to the topic below:

RAM SS Semirigid Diaphragms

How do I assign frame numbers to a mutual column in two different frames?

In order for the pseudo flexible diaphragm properties to automatically recognize frames within the diaphragm boundary, frame numbers must be assigned to the vertical frame members. In many instances a single column is part of two different frames. In these situations, there are two different approaches. First, you could assign the mutual column the same frame number as one of frames. The total force in the frames will be correct and the overall behavior will be accurate. However, the axial force in the beam will be impacted and might not be correct. This may create an unconservative design and may require you to design the beam by hand for the proper axial force. The second approach is to assign a unique frame number to mutual column and assign it percentages in both directions. For example, imagine frames 1 is parallel to the X axis and frame 2 is parallel to the Y axis. Each frame takes 50% of the story shear in its respective direction. If they share a mutual column, it could be assigned to frame 100. Then, the percentages in the X direction would be entered as frame 1 and frame 100 each with 25% in the X direction and frame 2 and frame 100 each with 25% in the Y direction.  

See Also

Product TechNotes and FAQs

Structural Product TechNotes And FAQs

External Links

Bentley Technical Support KnowledgeBase

Bentley LEARN Server

Comments or Corrections?

Bentley's Technical Support Group requests that you please confine any comments you have on this Wiki entry to this "Comments or Corrections?" section. THANK YOU!