Eight floor systems, each constructed with a different subfloor materi
al, were covered with oak strip flooring and exposed to environmental
conditions typically found in residential buildings during constructio
n and the first 2-1/2 years of occupancy. The objective of the project
was to qualitatively and quantitatively explore the interaction of mo
isture cycling, dimensional stability, and fastener strength retention
on the performance of the floor systems. The exposure program include
d construction site conditions where the subfloor had been installed b
ut not protected from weather, an interior construction phase simulati
on, and a simulated occupancy phase. Construction variables included t
hree thicknesses of plywood, three thicknesses of oriented strandboard
(OSB), and traditional diagonally laid boards. Two types of flooring
fasteners (staple and nail) and two different finishes (polyurethane a
nd sander sealant) were included in the program. Based on gap developm
ent, surface profile, fastener strength retention, and lateral movemen
t, the results indicate that the thickest composite subfloor materials
and the solid wood subflooring provide the best underlayment for hard
wood flooring. Stapled connections were stronger and more variable tha
n nailed connections. Joints in OSB had more variability than those in
plywood. Also, moisture cycling increased the variability of joint st
rength for all subfloor materials. The lower permeability of the polyu
rethane finish allowed less movement and deformation due to environmen
tal moisture changes compared to the sander-sealant finish.