Tuned Mass Dampers and Frame Stiffening for a Medical Center Floor System

In recent years companies with medical imaging equipment have been leasing space in medical office buildings on the upper floors. This is problematic as medical imaging equipment can be very sensitive to vibration velocity. CT, CAT and PET scanners typically are sensitive to vibrations larger than 2,000 micro-in/s and MRI machines are affected by vibrations as little as 500 micro-in/s. Compare this to a typical office building which has a sensitivity of 16,000 micro-in/s (Murray 2016). Therefore, CT, CAT and PET scanners are 8 times more sensitive and an MRI machine is 32 times more sensitive than an office floor. Someone simply walking down the hall may induce vibrations in these machines that blur the images that they create. Since many office building floors may not even meet the office vibration criteria, meeting the criteria for such scanners is well beyond normal performance of an office floor system.

tuned mass dampers being installed

It was desired to place an MRI machine on an elevated floor of an existing office building with 50-foot spans. The floor was constructed of steel open-web trusses with concrete fill over metal deck. As discussed previously MRI machines are highly sensitive to vibrations; this particular unit required a maximum velocity of 0.07 micro-meters per second (micro-m/s) or 25,000 micro-in/s.
When the existing floor was modeled the velocity was predicted at 2.1 micro-m/s in the location of the MRI, thirty times what was allowed. To mitigate such a drastic difference, multiple strategies were implemented.

First the trusses were strengthened to three times (3x) their original stiffness. This stiffening only reduced the velocity to 0.62 micro-m/s, still far from what was required. Note that the velocity has decreased as well as the fundamental frequency increased to ~7.5 Hz. Stiffening the floor has a significant effect on vibration reduction, but even stiffening by a factor of 3 was not nearly enough for this situation.
Next a bridging member (a truss similar to the original trusses in this case) was added to the model. The natural frequency remained at ~7.5 Hz but the resonant velocity further decreased from 0.62 micro-m/s (with stiffening only) to 0.22 micro-m/s (with stiffening plus bridging). The reason for the natural frequency remaining unchanged is that bridging increases not only the stiffness of the structure but it also increases the portion of the floor mass participating in vibration. As noted earlier the maximum velocity at the unit location is reduced but still not enough.
Finally, TMDs were added to the model in addition to the stiffening and the bridging member. TMDs were included in the bay with the unit and the adjacent bay (perpendicular to the direction of the bridging truss). Placing TMDs in the adjacent bay was nearly as effective as placing it in the main bay as the two bays behaved as a continuous beam in the main modes. With 2 TMDs in each bay the velocity was reduced to 0.043 micro-m/s.

Following the installation of the TMDs, floor vibration at the target bays, without and with the tuned mass dampers operational, were measured. With two TMDs in place in each bay, the actual measured velocity in the floor system was approximately 0.07 micro-m/s which was the maximum allowed.