Stability Issues in RAM Frame

Question

I'm modeling a performing arts center with BRBFs for the lateral system. When running the model in RAM Frame I've been getting errors about the model being unstable causing it not to run. I followed the steps listed in this link, http://communities.bentley.com/products/structural/structural_analysis___design/w/structural_analysis_and_design__wiki/ram-frame-troubleshooting-tn , and can reach step 5 but not go any further. 
 We've modeled a very similar performing arts center before using regular braced frames and that model had no issues running, so I'm not sure if the issue is the BRBFs not providing enough stability or I'm missing something glaringly obvious in the settings. 
 
 Thanks

Karl G · Accepted Answer

The model runs for all loads cases if P-Delta is turned off. However, I see some large deflections associated with local effects of the semirigid diaphragm. See below. 
 
 Here are my recommendations: 
 1. Change the diaphragms on the Tech Mezz Level to Flexible/None diaphragms and set the Exposure in RAM Frame - Loads - Exposure to "None". These diaphragms are small partial diaphragms connected to only a few frame columns and not appear to be subject to wind loading. The mass of these diaphragms has already been combined to a diaphragm on an upper level. 
 2. In RAM Frame - Criteria - Diaphragm, check the box for including out-of-plane stiffness of semirigid diaphragms with one-way decks. The semi rigid diaphragms have a low flexural stiffness due to a small effective thickness and a relatively low Elastic Modulus. The effect of including the out-of-plane stiffness should be small. However, since the top of the columns are fixed, you may see a small moment transfer between the diaphragms and columns. If this is a concern, you can release moment at the top of the columns that are connected to a semirigid diaphragm. 
 3. Change the wind and seismic load cases to use specified frequencies and periods instead of calculated values. The P-Delta associated instability is closely tied to the Eigen Analysis used to determine these values. The Eigen Analysis is not completed if the frequencies and periods in each direction are specified. You can either use specified values from an analysis run without P-Delta or use a frequency of 1 Hz for wind load (rigid structure equations used) and T= Ta for seismic load. 
 4. Change the solver type in RAM Frame - Criteria - Analysis to the In-Core Direct Solver. The direct solver is more rigorous than the sparse solver and the more rigorous solution method will help bypass the error.

Karl G · Answer

There is an option in RAM Frame to exclude the BRBF's when analyzing the gravity load cases (Criteria - General). If the column bases and beam ends are pinned and there are no shear walls, this will result in no lateral stiffness and result in an instability when attempting to analyze those load cases. A similar problem occurs for tension-only members, which are also excluded when analyzing gravity load cases. 
 The frequently asked question, "Why am I getting large deflections under gravity loads?" on the web page below discusses this and some suggestions for avoiding the instability: 
 http://communities.bentley.com/Products/Structural/Structural_Analysis___Design/w/Structural_Analysis_and_Design__Wiki/ramss-tension-only-faq 
 Please post back if this does not apply to your model or if you follow the suggestions and continue to see the error.