Wood Joist Floor Research Articles

E90 round-robin—Status and path forward

A good understanding of the uncertainty of a test method is necessary to interpret the results of said test. The ASTM E90 test has many potential sources of variation that can lead to significant uncertainty in the test results, but this uncertainty is not well documented in the precision and bias statement within ASTM E90. The current precision and bias statement provides some limited repeatability data for a wood joist floor and a concrete slab and only provides reproducibility data for the ASTM E1289 reference specimen. The E1289 reference specimens are simple steel plates in single and double-layer configurations that yield STC ratings far below those of the wall and floor/ceiling assemblies that are commonly tested. The authors are in charge of a new round-robin that aims to provide more detailed repeatability and reproducibility data using common wall designs with STC ratings in the range more commonly seen in building code requirements and design guidelines. In this presentation, the authors will cover their proposed plan for the round-robin and will give an update on its status.

Quantitative comparisons of resilient channel designs in walls and ceilings

Resilient channel is common within lightweight construction to improve the airborne and impact noise isolation of wall and floor-ceiling assemblies. The authors have previously presented results from systematic studies of resilient channels in wood stud wall assemblies (Internoise 2009; International Congress on Acoustics 2010) and wood-joist floor assemblies (International Congress on Sound and Vibration 2015) in which the effect of brand and model of resilient channel was compared in various assemblies. These studies documented the clear and significant acoustical performance differences between makes and models of channel that would have been considered equivalent before the release of this work. Additional comparative studies have been performed adding to the data previously presented. Acoustical performance is compared and conclusions about products are presented.

Change in Canada’s national building code—Assessing flanking sound transmission in concrete-masonry constructions

The proposed 2015 edition of the National Building Code of Canada sees a major change in sound insulation requirements. Instead of prescribing requirements for the separating assembly only (in terms of STC values), the Code now sets requirements for the sound insulation performance of the complete system (in terms of Apparent Sound Transmission Class (ASTC) values), including flanking sound transmission. The National Research Council Canada is actively supporting the change in the Code by conducting various projects with industry associations from different construction sectors, in order to provide tools, guidance, and the necessary data for compliance. This presentation focuses on a joint project between the National Research Council Canada and the Canadian Concrete Masonry Producers Association. In the project, the direct and flanking sound transmission in concrete masonry and hybrid building systems were investigated. For masonry walls in combination with concrete floors, the ASTC values were calculated according to ISO 15712-1. For masonry walls in combination with wood joist floors, the ASTC values were measured according to ISO 10848. Furthermore, the effect of linings on concrete masonry walls was investigated. This presentation will provide an overview of each of these issues, including results and recommendations.

Fire Resistance of Wood Truss Floor Assemblies

This paper discusses the factors that affect the fire resistance performance of lightweight wood frame unrestrained floor assemblies protected with Type X gypsum board ceiling finishes. Twenty-two fire resistance experiments were conducted on full-scale load-bearing wood joist floor assemblies using the ULC standard fire exposure time–temperature curve that is similar to ASTM E119 standard. Parameters investigated in this study include the effects of gypsum board screws spacing from board edges, insulation installation, insulation type, joist spacing for assemblies with gypsum board attached to resilient channels, joist depth, resilient channel installation, resilient channel spacing, sub-floor topping, number of sub-floor layers, and load magnitude. The impact of these parameters on the fire resistance of wood joist frame floor assemblies is discussed. The fire resistance of wood frame floors appears essentially to be governed by the gypsum board screw spacing from the board edges and the type of insulation in assemblies with one layer of gypsum board and by the resilient channel spacing and the gypsum board screw spacing from the board edges in assemblies with two layers of gypsum board. The effects of other parameters in assemblies with two layers of gypsum board such as the joist spacing where the gypsum boards are attached to resilient channels, installation of resilient channels, insulation installation, insulation type, adding gyp-crete topping above the sub-floor and number of sub-floor layers on fire resistance are relatively insignificant.

Fire Resistance of Wood Joist Floor Assemblies

This paper discusses the factors that affect the fire resistance performance of lightweight wood frame unrestrained floor assemblies protected with Type X gypsum board ceiling finishes. Twenty-two fire resistance experiments were conducted on full-scale load-bearing wood joist floor assemblies using the ULC standard fire exposure time–temperature curve that is similar to ASTM E119 standard. Parameters investigated in this study include the effects of gypsum board screws spacing from board edges, insulation installation, insulation type, joist spacing for assemblies with gypsum board attached to resilient channels, joist depth, resilient channel installation, resilient channel spacing, sub-floor topping, number of sub-floor layers, and load magnitude. The impact of these parameters on the fire resistance of wood joist frame floor assemblies is discussed. The fire resistance of wood frame floors appears essentially to be governed by the gypsum board screw spacing from the board edges and the type of insulation in assemblies with one layer of gypsum board and by the resilient channel spacing and the gypsum board screw spacing from the board edges in assemblies with two layers of gypsum board. The effects of other parameters in assemblies with two layers of gypsum board such as the joist spacing where the gypsum boards are attached to resilient channels, installation of resilient channels, insulation installation, insulation type, adding gyp-crete topping above the sub-floor and number of sub-floor layers on fire resistance are relatively insignificant.

Experimental assessment of participating width of wood floor sheathing subjected to concentrated load: case study of OSB

An equation for the participating width (elastic) of one-way orthotropic wood joist floor sheathing subjected to concentrated loading has been recently developed. The participating width is the length of sheathing normal to the span or parallel to the joists that is structurally involved in supporting the concentrated load. In this paper, the participating widths of 18-mm thick one-way simply supported oriented strand board sheathings with spans of 300 mm and 600 mm (span–thickness ratios of 16.7 and 33.3, respectively), determined by laboratory tests are reported. The experimentally determined participating widths appear consistent with the analytically predicted participating widths of the sheathings. In addition, it has been shown that the participating width in the linear elastic stage of behaviour is roughly equal to the participating width near maximum load. Subject to experimental validation by other types, thicknesses and spans of sheathing, the analytically derived participating-width equation is recommended in modelling wood joist floor sheathing for concentrated load performance assessment.

Test-piece size for uniformly distributed and concentrated load performance assessment of wood floor panel

An examination of North American and European test floors for determination of uniform and concentrated load performances of wood joist floor panels shows that there is a wide variation in the required width to span aspect ratio (1.5 to 8.0) of the test floors. Although the test floors are intended to replicate real floors, there is a minimum width to span aspect ratio that is sufficient to simulate both uniform and concentrated load performances of floor panels. The minimum width to span aspect ratio is conservatively the effective width to span ratio of a one-way spanning plate model of the floor panel. A parametric study of the effective widths in bending of simply supported, fixed and continuous plate one-way models of floor panels has been carried out. An equation for the effective width of the models in terms of the span, thickness and mechanical properties has been derived. Theoretically, the upper bound of the effective width is of the order of 2.0 times the span. Subject to experimental validation, the derived effective width equation is recommended for specifying the size of test piece for determination of uniformly distributed and/or concentrated load performance of wood joist floor panels.

Estimating in-situ Material Properties of a Wood Joist Floor: Part 1 — Measurements of the Real Part of Bending Wavenumber

This paper outlines and evaluates a non-destructive experimental technique used to obtain in-situ measures of the real part of the bending wavenumber in the two principal directions of a wood joist floor. In-situ measured wavenumbers are compared to those obtained from beam samples cut from the floor sheathing to identify the frequency range when the joists significantly affect vibration response in the floor sheathing. Wavenumber measurements confirm the highly orthotropic nature. Measurements indicate that the wavenumbers in the direction parallel to the joists are a function of location between the joists. Lower wavenumbers in this direction can be expected closer to the joists. Some discussion is given regarding the rates of attenuation with distance that can be expected parallel and perpendicular to the joists given the variation in wavenumbers, however this will be the topic of a separate paper, Part 2.

Models for Concentrated Load Performance Assessment of Isotropic Wood Floor Sheathing

An important information required in the determination of models for the concentrated load performance assessment of wood joist floor sheathings is the minimum width of sheathing (effective width) that can be considered to support the load. In this paper, a theoretical study of the effective width in bending of simply supported, fixed, and continuous one-way models of isotropic wood joist floor sheathings is reported. The effective width has been found to be directly proportional to the span, inversely proportional to the thickness and degree of fixity at the supports, and almost independent of the wood sheathing’s mechanical properties. An equation has been developed for calculating the effective width of the models. For the critical simply supported model, the upper and lower bounds of the effective width are 2.0 and 1.4 times the span, respectively. Based on the effective width, a criterion for determining when a fully supported isotropic sheathing reduces to a one- or two-way plate model and geometric models for the concentrated load performance assessment of isotropic wood floor sheathings are proposed.

Flanking sound transmission in wood-framed construction

This paper presents results from the first phase of a project to characterize flanking sound transmission at wall/floor junctions in wood-framed constructions suitable for multifamily residential buildings. This phase of the study focused on the junction between a wood joist floor and a separating wall with two rows of 38×89-mm studs. The objective of the study was to identify details that simultaneously provide good noise control and fire resistance. This paper presents the experimental approach and key results from the acoustical measurements. In addition to comparing how five common fire control solutions for the floor/wall junction affect the airborne and impact sound transmission, several generic retrofit modifications to reduce the structural noise transmission were tested. Parameters studied included the floor-joist orientation (parallel or perpendicular to the separating party wall), and the type of subfloor (plywood or oriented strand board). The study has shown that suitable details can provide both adequate fire resistance and noise control. The technical approach provides a framework for extending beyond established laboratory measurement procedures, to predict performance in actual buildings.

Concentrated Load Design Procedures for Wood-Joist Floor Systems

Current light-frame wood construction design specifications recognize repetitive member effects in wood-joist floor systems. However, it is unclear whether these design stress increases apply equally to both uniform and concentrated load cases. To address this issue, eight floor systems were constructed and tested to determine their load-sharing properties when subjected to a concentrated load. The concentrated load was independently placed on each joist at quarter span and midspan of the floor while monitoring midspan deflections and reactions at one end of each joist. Statistical distributions were then chosen to represent percentages of load remaining at the end reactions of each joist. By comparing and combining experimental data with previous finite-element results, floor joist design equations were developed for a single joist. Factors addressing load distribution were introduced into these equations for calculating maximum moment and deflection. Additional design equations were developed to account for the “pseudo” concentrated load cases stipulated in ASCE standard 7-95, Minimum Design Loads for Buildings and Other Structures. Inherent load sharing in wood-joist floor systems was found to substantially reduce the moment and displacement in the joists directly subjected to a concentrated load.

Time-dependent service-load behavior of wood floors: analytical model

A preceding companion paper [1] examined the experimental time-dependent service-load response of wood joist floor systems, including an assessment of time-dependent deflections and load distribution. The results of the experimental program are used herein to develop an analytical model for the time-dependent response of wood joist floors. Time-dependent component (i.e. joist, sheathing and connection) models are developed from component tests. These component models are then integrated into a system model. The focus of this paper is to present the theoretical basis for the time-dependent system model, and illustrate its predictive capabilities. The system model approximates the time-dependent system creep of parallel-member wood systems to an acceptable degree of accuracy. Moreover, the model requires relatively little computer time, thus making it suitable for use in a Monte Carlo simulation

Time-Dependent Service-Load Behavior of Wood Floors: Experimental Results

It is well known that wood exhibits continued additional creep deflection under long-term load. Differential creep behavior among joists, sheathing material, and connection details in a wood joist floor system can affect the system performance of the assembly over time. For example, if the flexural stiffness of the partial composite T-sections consisting of wood joists and sheathing changes continuously over time, the deflection response of the wood assembly would obviously be affected. The objective of this research was to investigate the time-dependent behavior of wood floor systems within the service-load range as defined by time-dependent deflection behavior and time-dependent load-distribution behavior. The specific focus of this paper is on experimental tests and observations.

Floor research project at NRC Canada: An update on progress—Measurements with experimental impact devices

A major thrhust in this floor project is the collection of low-frequency impact information, below 100 Hz, using special impactors: a tire-dropping machine, two experimental rubber balls supplied by H. Tachibana, and a walker. The frequency range for the standard ISO tapping machine is also extended at 50 Hz. A preliminary analysis of the data collected so far in the project will be presented. Factors that were expected to change the shape of the impact spectrum drastically at low frequencies, so far, have had little or no effect. Changing joist depths from 185 to 285 mm had negligible effect on low-frequency transmission. Changing I-beam depth from 235 to 460 mm did have a significant effect. The best wood joist floor tested to date is still worse than a 150-mm concrete slab. So far in the project, the most effective remedy for excessive low-frequency impact sound is mass.

Measurements of airborne and impact sound through floors—A new project

A comprehensive project to study all aspects of sound transmission through floors has just been initiated at the National Research Council of Canada. As well as airborne sound transmission, measurements will include: impact sound transmission using a standard tapping machine, a special rubber ball, a walker, and a tire machine meeting the relevant Japanese standard. As well as these measurements on complete floor systems, all material properties will be measured to provide a complete characterization of the floor systems. Vibration measurements to determine modal characteristics and damping will also be made. These measurements will be made in the new floor facility at NRC which has been found to give more reliable results than the old facility, which is no longer in use. Floors in the new facility measure 3.8 × 4.7 m, a more realistic size than the 2.4 × 2.4 m floor in the obsolete facility. An outline of the project will be given. Initial measurements on wood joist floors to look at the effects of methods of attaching the floor and ceiling layers will be presented.

Reliability of Wood Joist Floor Systems with Creep

Light-frame wood floors are constructed from a number of parallel flexural members (joists), which provide the primary structural support. Sheathing material is attached across the top edges of these joists to form the floor surface. Because the joists are connected in this manner, load sharing occurs, which enhances the performance of the system as a whole. Current design procedures focus on the behavior of single members with only limited provisions account for the improved performance of systems. In this study, the reliability of wood joist floors is investigated considering both strength and serviceability limit states. The objective of the system reliability analyses is to determine appropriate system factors for use in single member load and resistance factor design checking equations. These factors are included to account for the beneficial effects associated with systems of parallel members. This study considers factors such as lumber species and grade, floor size, and the effects of creep deformations. A stochastic viscoelastic stress-strain relationship is used to model the time-dependent behavior of wood, the stochastic pulse process models are used to account for the temporal variability of the loads. Cumulative damage analysis using two common damage accumulation models is used to account for the time-dependent strength of wood flexural members.

Sound transmission through wood joist floor/ceiling systems: A study of the effects of sound absorbing materials and changes to the ceiling structure

A wood joist floor system was constructed in the Acoustics Laboratory at NRC. Measurements of airborne and impact sound transmission were made for different types and amounts of sound absorbing material in the cavity. Differences between materials were small. Different makes of resilient metal channels were installed; all gave about the same results. Methods of improving an unsatisfactory floor were examined. Over 20 floor systems were measured. The study was funded by Canada Mortgage and Housing Corporation. Measurements were made down to 63 Hz and revealed that in many cases there are resonances occurring below 125 Hz that have a strong influence on transmission loss values at and above 125 Hz. While the causes of the resonances are not always clear, measurements to these low frequencies give a clearer picture of the effect of changes to the structure. This paper will give a summary of the results obtained.

Reply: Design of lightweight wooden floors to avoid human discomfort

This paper gives "designer usable" methods for predicting the dynamical behaviour of lightweight wood-joist floors covered with semirigidly attached wood-based sheathings of materials such as plywo...

Design of lightweight wooden floors to avoid human discomfort

This paper gives "designer usable" methods for predicting the dynamical behaviour of lightweight wood-joist floors covered with semirigidly attached wood-based sheathings of materials such as plywood or waferboard. Included are methods for predicting natural frequencies and root-mean-square (rms) acceleration under a defined forcing function. It is proposed that this rms acceleration be the criterion for estimating user perception based tolerance of a projected design solution in the domestic setting. Case studies suggest that the economic cost of producing more rational designs is small. Consideration is, at this stage, limited to floors that are rectangular in plan and contain no openings nor intermediate support partitions. Suggestions are made as to methods of building floors with acceptable vibrational performances, together with recommendations for further research. Key words: floors, wood, vibrations, human tolerance, design method.

Wood Floors with Nonlinear Nail Stiffness

The results of an analysis of wood joist floors, including the effects of the nonlinear force‐slip relationship for nails, are presented. The finite‐element analysis procedure includes an iterative scheme for solving the resulting nonlinear equations, thereby satisfying equilibrium and compatibility. To assess the ability of the model to predict deflections above the service load level, the computed deflections were compared with those measured in tests of full‐scale point‐loaded and uniformly loaded floors. These comparisons were made at the known failure loads of the test floors. Also, a comparison was made between the results of a linear analysis of each test floor in which the connector stiffness was assumed constant throughout the load range and the experimental deflections. The nonlinear analysis proved superior, particularly for the uniformly loaded floors.