Vibration in composite floors

The physical response of a floor to imposed repetitive impact loads depends on a number of factors - span, deflection, stiffness, mass, etc. The tendency to reduce mass to obtain better structural economy (and hence cost) is at odds with the principles of avoiding vibration. While vibrations set up by people walking across the floor are highly unlikely to be anything more sinister than a psychological effect, the problem is still very real and needs to be addressed.

Generally, it is impracticable to provide damping to a floor that displays tendencies to vibrate. From a theoretical point of view, stiffening support beams will help, but it would be difficult to do this within an occupied building or at reasonable cost.

Providing floor finishes makes a negligible difference (other than by a very marginal increase of mass) although the cushioning effect may change a person's response to the problem even when in reality there has been very little change. Other subtle changes might help to reduce complaints; for example, removing or changing fittings that are inclined to rattle, securing loose floor tiles and avoiding cantilever light and computer screen supports.

Increasing the mass of the floor is unlikely to be possible in a floor that has already been constructed. Similarly, floors rarely incorporate any particular damping measures (although they may do where external vibration sources are likely to cause problems). In practice, a measure of damping may be provided by non-load-bearing elements of structure, the occupants (to a small extent) and by the performance of the structure where this differs from that which was intended in pure design terms. To provide rigidity to a structure, structural engineers often provide reinforced concrete walls (shear walls) or cores; this enables them to simplify the design process for the steel frame by treating structural connections as pin joints as opposed to moment connections. In practice, pin joints are often not pin joints in the true sense and some load transfer is inevitable - this can modify the performance of the floor.

Floors prone to vibration are best considered prior to fitting-out works. A detailed analysis will identify areas most at risk so that the design and location of corridors or walkways can be configured to avoid these. Measures such as the careful location of office equipment can also serve to keep foot traffic to more satisfactory routes, avoiding, for example, long span support beams and using areas that are naturally stiffer.

Problems of vibration in office floors are most often encountered in long-span steel floors with composite decks but precast concrete floors are also vulnerable in long-span configurations. The Steel Construction Institute's design guide on vibration of floors sets out calculation methods for the prediction of the effects of transient loads.

Case study - 1980s office building, West End of London

The building was a purpose built office of around 6,000m2 arranged over 9 floors with a basement level. The building was multi-occupied with a variety of different business types from general offices to trading floors.

The tenant of one of the trading floors complained that periodic vibrations were affecting their ability to read the small rows of figures displayed on their computer screens. Each trader had a display of 3 monitors, each arranged on a complicated bracket device attached to the individual workstation. No other tenants were affected, and the perception was that the vibration was due to the operation of a particular air-conditioning cassette unit on the floor above.

The landlord's initial response was to try and replicate the vibration by operating the air-conditioning unit and monitoring the response. No perceptible vibrations could be reproduced. The next step was to set up a series of monitoring points around the office using accelerometers wired back to a data logging unit. The fortuitous recurrence of the vibration incident enabled a graphical record of the condition to be obtained. The next step was to try and identify the source. By this time, tenants on 2 other floors had miraculously discovered vibration problems and so monitoring of those floors was put in place and a detailed study was undertaken over a weekend period when extraneous background 'noise' could be eliminated. Various items of main plant were turned on at roof level, macerators in female toilets were operated and local on-floor appliances turned on. All failed to replicate the vibration pattern discovered earlier.

Analysis of the key incident showed that it was significantly below the levels that would be considered unacceptable by comparison with published guidance. The most likely cause was human activity within the office, but the problem was magnified by the computer screen arrangements, each display arm tending to amplify the very small vibrations that were occurring. It was probably the magnified response that was alerting tenants to the problem rather than the vibrations themselves. Word of mouth between tenants led to concern that the building was in some way at fault, hence the 'discovery' of vibration on the other floors.

By designing damping mechanisms to the support arm it should have been possible to eliminate the vibration or at worst to bring it down to a level that did not affect the user. The building was not defective and landlord intervention was not required.