I have a stair that I'm concerned may be susceptible to vibrational issues. I've often used RAM Elements to model a variety of stair configurations under gravity loads and ELF seismic loading, but can it also be used to check for vibration? Most of the information I'm finding online references using SAP to find the natural frequency for use in the AISC DG11 procedures, but DG11 also says that "several commercially available programs provide convenient features for computing the FRF magnitude". I know Elements can determine mode shapes via eigenvalue analysis, but what about the natural frequency? Any guidance on feasibility and/or ways to translate the procedures in the design guide to comparable inputs in Elements would be much appreciated. For instance, the amount of frictional damping on a stair plays a big role, and there's a damping input in the Earthquake Acceleration tab, but can any of the dynamic analysis functions for seismic loads be used for human-induced vibration, or would that not be applicable Thanks!
Jason
As you noted, Ram Elements can solve for the mode shapes which gives you the natural frequency (and period which is just the inverse). The seismic dampening variable is the only extra control the user has (besides the mass and model geometry). Ram Elements is not specifically designed for checking footfall or machine vibration checks, however. Here is a post from a few years ago: https://communities.bentley.com/products/ram-staad/f/ram-staad-forum/104819/rhythmic-excitation-of-monumental-stairs-in-ram-elements
Yeah, I did see that thread and had actually commented on it 4 years ago, and had seen another one asking about the mass to use in natural frequency determination, so I knew people have tried using Elements like this, but hadn't really seen anything indicating success with it. Since Elements isn't set up for vibration checks, I'm mainly just wanting to be able to determine the fundamental natural frequency of the stair for use in equations from DG11 and some of Brad Davis' seminars. Right now, I'm trying to simply match up a basic 1 stringer model in Elements with the results from the simple beam natural frequency formula in DG11. The DG11 estimate of natural frequency is 5.35 Hz for the values in the screenshot below. Elements gives me a first modal frequency between 4.5 Hz and 6.76 Hz depending on how I distribute the mass in nodes along the stringer (e.g. midpoint, quarter-points, tenth-points, etc). I also tried using the element self mass as the mass source in the analysis to see if that would be better, but that produced a frequency of 26.3 Hz compared to a calculated frequency of 6.8 Hz (adjusting the Ix in the formula below to match a section in Elements that would have 2k of total self-mass). I was hoping to get fairly close agreement on a simplistic model before using it on my real model. Is there something I need to be doing different in Elements?
I was helping another user with a similar question today, and we ended up rotating the model 90 degrees so that the +z axis is up (or +x). This makes it work better with Ram Elements eigen solution.
So when you say rotating the model to have +z up, is there a way to define what axis is vertical, or are you using the rotate command to rotate the physical model? Then would you manually change TY masses to be TZ? Loads defined globally like self-weight would need to be changed to the new vertical axis and so forth?
We rotated the physical model and swapped the mass directions in this case.