Low-amplitude floor vibrations have become a governing serviceability performance design consideration for floors constructed with low mass-to-stiffness ratio materials such as wood. Studies reported here were conducted at the University of Ottawa to assess vibration serviceability performance of glued-laminated-timber (glulam) beam-and-deck floor systems. Such floors are applicable in non-residential buildings having spans up to about 10 m. The primary goal was to use test and numerical analysis methods to investigate how construction variables (e.g. beam span, beam spacing, addition of nonstructural overlays) affect the vibration responses of such floors. A secondary goal was to assess applicability of vibration serviceability design criteria proposed for other types of floors to glulam beam-and-deck floors. Apart from tests aimed at characterizing responses of laboratory built rectangular plan floors, focus groups were asked to subjective rate acceptability of the performances of those floors. Focus group ratings determined ability of humans to discern alterations in floor motions resulting from construction modifications, based on an opinion survey technique developed by other researchers. This determined that humans can detect and rate performance of floors having different engineering design characteristic, but cast doubts on the consistency of the employed opinion survey technique.
Laboratory tests revealed that mid-span displacements of floors are functions of two-way deflected shapes of floors and are reduced by adding nonstructural overlays and extra beams. Adding non-structural overlays reduces fundamental natural frequency demonstrating gain in modal mass was greater than for modal stiffness. There was inconsistency between the result of focus group evaluations and predictions of acceptability of floors made using available suggested vibration serviceability design criteria.
Finite Element (FE) models of glulam beam-and-deck floor systems were created and verified using laboratory test data. Based on those models it was concluded that fundamental natural frequencies and mid-span displacements of floors are relatively insensitive to variations in floor width-to-span ratios. However, higher order natural frequencies are strongly affected by floor width-to-span ratios. Increasing thickness of deck elements can decrease natural frequencies and cause them to cluster in ways that amplify surface motions caused by dynamic forces like human footfall impacts.
Field vibration tests were conducted to investigate the dynamic behaviour of a large glulam beam-and-deck office floor having a complex plan shape and support conditions. That floor has long beam spans and partial continuity between bays defined by a mixture of column and wall supports. It was tested before non-structural floor toppings were added and after building completion and occupation. FE modeling of the floor was created and predicted modal characteristics (i.e. mode shapes, natural frequencies) compared with experimentally derived ones. Controlled walking tests were conducted to assess the dynamic response under office occupation conditions. It was concluded the vibration serviceability response of the floor was satisfactory based on peak acceleration measurements and lack of office worker dissatisfaction. Importance of this is the floor has low order natural frequencies less than 8Hz, which means existing proposed design practices created for lightweight timber floor would incorrectly classify its performance. The discrepancy is indicative that such design practices fail to capture effects of construction variables and damping characteristics of large floors.
In general, vibration characteristics of lightweight floors are highly related to effects of construction details such as plan aspect ratio, boundary conditions and presence of nonstructural elements. Apart from clarifying specifics of how glulam beam-and-deck floors vibrate, this thesis is intended to contribute to Canadian and international efforts to create engineers design methods that robustly predict whether or not specific floors will have adequate vibration serviceability performance under defined floor occupancy conditions.
| Identifer | oai:union.ndltd.org:uottawa.ca/oai:ruor.uottawa.ca:10393/36443 |
| Date | January 2017 |
| Creators | Ebadi, Mohammad Mehdi |
| Contributors | Doudak, Ghasan |
| Publisher | Université d'Ottawa / University of Ottawa |
| Source Sets | Université d’Ottawa |
| Language | English |
| Detected Language | English |
| Type | Thesis |
Page generated in 0.004 seconds