Research Highlights

Technical Series 98-110

Testing of Air Barrier Systems for Wood Frame Walls

Introduction

A well designed and constructed air barrier system must be airtight, continuous, structurally capable of withstanding the air pressures exerted on the building envelope and constructed with materials that are durable, or at least maintainable. The air barrier systems will usually be constructed from different airtight materials joined with tapes or sealants to provide continuity over the entire building envelope. Over the life of a structure, the air barrier system will be subjected to pressure differentials resulting from stack effect, wind and possibly fan pressurization. From a structural point of view, pressure differentials resulting from stack effect are relatively small: typically, less than 20 Pa in a low-rise building will affect the air leakage rate through the air barrier. Pressure differentials greater than 1,000 Pa may be induced by wind, but because of the duration of the effect, these pressure differentials are generally less important in the consideration of total air leakage. However, if pressures are exerted on an air barrier system that has not been properly designed, damage may occur that could result in increased amounts of air leakage over the life of the building. Ultimately, this could cause the air barrier system to lose structural integrity. The National Building Code of Canada (NBC), in considering structural performance, requires an air barrier system to resist both sustained wind loads lasting up to one hour and higher gust wind loads lasting from three to five seconds.

Research Program

The purpose of this project was to design a test procedure to verify the air leakage and structural performance of air barrier systems intended for use in wood frame walls. Tests to evaluate the performance of windows, curtain walls and doors followed all the procedure outlined in ASTM E283, “Standard Test Method for Rate of Air Leakage through Windows, Curtain Walls and Doors,” using a pressure differential of 75 Pa. A similar test procedure was used in this project to measure the air leakage of air barrier systems. The Institute for Research in Construction (IRC) of the National Research Council Canada has proposed the following levels of airtightness for air barrier systems:

Air Leakage Air Leakage Rate

Classification (l/s-m2 @ 75 Pa)

Type 1 0.10 to < 0.15

Type 2 0.05 to < 0.10

Type 3 < 0.05

The procedure outlined in ASTM E331 was followed for determining the structural performance of the assemblies. The pressures used for the testing were derived from the design wind pressures in the Supplement to the National Building Code. The maximum loading required for the “worst” location in Canada is 1,000 Pa for one-hour loading and 2,500 for gust loading.

In this project, manufacturers were invited to provide 10 wall assemblies, each measuring 2.4 metres by 2.4 metres. These were tested for air leakage and structural performance under both negative and positive pressure differentials (see Table 1 for descriptions of the walls tested). For the purposes of the tests, pressure is defined as negative when the pressure would tend to pull the air barrier on to the studs; a positive pressure differential would push the air barrier off the studs. The negative pressure differential tests, both for air leakage and structural performance, were always performed first, as they were less likely to cause structural damage to the wall. Nine of the ten walls were constructed using a conventional 16 inches o.c. wood frame stud wall, with perimeter studs to facilitate sealing the wall to the test apparatus.

The air leakage of each wall was measured at pressure differentials ranging from approximately 10 to 100 Pa, and the results were plotted as flow-pressure curves, which are available in the full report. The pressure differential at each test point was held for a minimum of five minutes to allow time for the flow rate to stabilize.

To simulate sustained wind loading, the test walls were subjected to test pressures of 250, 500 and 1,000 Pa for one hour each. Gust loading was simulated by subjecting the walls to pressure differentials of 1,500, 2,000 and 2,500 Pa for five to ten se conds. Following each structural performance test, the wall was examined for visual signs of structural damage, and the air leakage rate was measured at approximately 75 Pa for verification of non-visible damage.

Table 1 describes each of the air barrier systems tested and the air leakage rate obtained at a pressure differential of 75 Pa. The table also provides the maximum pressure differentials, for both sustained load and gust load, to which the wall was exposed without exhibiting a significant loss in airtightness. Finally, it includes observations on the results of the structural wind load tests.

Table 1: Results of Air Leakage and Structural Load Tests
Wall # System Description Air Leakage Rate
l/s-m2@75 Pa
Sustained Load Gust Load
1 Fibreboard/ Tyvek/ Strapping 0.488 -1,000 Pa +1,000 Pa -1,500 Pa +1,500 Pa
  Fiberboard sheathing nailed to the wood structure with 1 3/4-inch galvanized roofing nails at 6 inches o.c., Tyvek paper (1990 version) stapled to the fibreboard every 4 feet, vertical wood strapping nailed to the perimeter and face of the studs with spiral nails at 12 inches o.c. There was no change to the air leakage rate after the structural tests. However, the wall was too leaky to obtain either positive or negative test pressures at 2,000 Pa and 2,500 Pa: under positive pressure, the Tyvek ballooned outward between the wood strapping and after the +1,000 Pa sustained load test, the Tyvek paper remained loose on the surface of the fibreboard sheathing.
2 Glasclad/ 3M tape 0.300 -1.000 Pa +500 Pa -2,300 Pa Not tested
  Glasclad insulation board nailed to the wood structure with 2 1/2-inch spiral nails with 1-inch square plastic washers, and 3M construction tape over the two vertical joints and perimeter. Two negative pressure differential tests were performed, one with nail perforations not taped and the other with nail perforations taped. The difference in air leakage rate between the two configurations was negligible. The wall was too leaky to perform the between the two configurations was negligible. The wall was too leaky to perform the 2,500 Pa test (maximum pressure -2,300 Pa). There was no change in the air leakage rate after the negative structural tests, although deflection of the Glasclad was quite noticeable during the -1,000 Pa sustained load test. The assembly failed during the first few minutes of the +1,000 Pa load test. Positive gust tests were not performed.
3 Exterior Gypsum/ Perm-A-Barrier Tape 0.015 -1,000 Pa +1,000 Pa -2,500 Pa +2,300 Pa
  Two exterior gypsum boards nailed horizontally to the wood structure using 1 3/4-inch galvanized roofing nails at 6 inches o.c., sealed at the joints with Perm-A-Barrier wall seam tape, after being primed with “Primer P-3000” one hour earlier. There was no change in the air leakage rate after structural tests. However, structural failure occurred during the +2,500 Pa test: the top gypsum board pulled off from the nails along the intermediate supports, after which the air leakage rate increased to 1,540 l/s-m 2 .
4 Extruded Polystyrene/ 3M Tape 0.002* -1,000 Pa +1,000 Pa -2,500 Pa +2,500 Pa
  Shiplapped, extruded polystyrene insulation boards nailed to the structure with 2 1/2-inch spiral nails with 1-inch diameter metal washers every 6 inches along the edges and 12 inches along the intermediate supports, with 3M construction tape over joints. Because of the airtightness of this assembly, the reference air leakage was taken at 185 Pa instead of 75 Pa. Two air leakage tests were performed under both negative and positive pressure differentials: first, with untaped nails at the centre of the insulation boards; then, with nails taped with 3M construction tape. No difference was noted with nails taped or untaped. The system failed structurally at some nail locations at +2,500 Pa gust load test, but the air leakage after the test was only 0.005 at 180 Pa.
5 Esclad/ 3M Tape 0.003 -1,000 Pa +500 Pa -2,500 Pa Not tested
  Esclad Energy Envelope boards installed vertically and nailed to the wood structure with 2 1/2-inch spiral nails and 1-inch diameter metal washers at 1 foot o.c. on the edges and 16 inches o.c. on the intermediate supports and with 3Mconstruction tape to seal the joints. The air leakage tests were done first with the nails along the intermediate supports of the Esclad board untaped, and then taped. Taping improved the air impermeability of the assembly by approximately six times, and these results are the ones that have been reported. There was no change in air leakage after the negative structural tests. The system failed 30 minutes into the +1,000 Pa sustained load test: the larger Esclad board pulled off the nails along the intermediate supports, except at two locations, where the nail pulled out of the stud.
6 Dryvit Wall Panel 0.003 -1,000 Pa +1,000 Pa -2,500 Pa +2,500 Pa
  Prefabricated wall panel supplied by Dryvit, consisting of steel stud structure, exterior gypsum board, expanded polystyrene insulation and a synthetic exterior finish. The air barrier system performed structurally through all the sustained load and gust load tests with no visible physical change of the assembly components. The gypsum board layer was made up of various differently sized pieces screwed to the steel stud structure about every four inches. The use of full-size gypsum boards might lead to a different performance than the one obtained.
7 Gypsum/ Joint Compound 0.002 -1,000 Pa +1,000 Pa -2,500 Pa +1,800 Pa
  Gypsum board (1/2-inch) attached horizontally to the wood structure with1 1/4-inch drywall screws every 8 inches along edge supports and 12 inches along the intermediate supports, with drywall joint compound (two coats) andpaper tape and surface-painted with two coats of latex paint. No change in air leakage was observed after negative pressure tests. Air leakage increased to 0.006 l/s-m 2 after reaching +1,800 Pa: the gypsum board pulled off somescrews but did not cause a complete break of the air seal. Because of the increased air leakage, a second +2,000 Pa test was attempted. Structural failure occurred during this test at approximately +1,600 Pa: the top gypsum board pulled away from all the screws along the intermediate supports, and the gypsum board joint broke along the whole length.
8 Plywood Skin Panel 0.004 -1,000 Pa +1,000 Pa -2,500 Pa +2,500 Pa
  Plywood skin glued with subfloor adhesive to a frame of 2 x 4 wood studs at 2 feet o.c. with a single top and bottom plate and 1 1/4-inch drywall screws at 6 inches o.c. The 2 1/2-inch glass fibre batt insulation against the inner plywood in thecavities left a 1-inch air space between it and the outer plywood. No change in air leakage rate nor any physical change was noted after all sustained load and gust load tests. To determine if a single plywood skin subject to the same attachment techniques would perform as well, 2-inch diameter holes were drilled in the exposed plywood skin at the top and bottom of each stud space, and the air leakage testwas repeated at intervals of from 10 to 500 Pa. The air leakage rate was measured as0.005 l/s-m 2 at 75 Pa. The 1,000 Pa sustained load test and the 2,500 Pa gust test were also repeated with no change in air leakage rate being observed.
9 Fibreboard/ Polyurethane Foam 0.019 -1,000 Pa +1,000 Pa -2,500 Pa +2,500 Pa
  Asphalt impregnated fibreboard (7/16-inch) nailed to the wood structure with 1 3/4-inch galvanized roofing nails at 6 inches o.c. along the edges and 12 inches o.c. along the intermediate supports, and stud cavities filled with at least 3 inches of standard density spray-applied polyurethane foam insulation cut flush to the face of the studs. The system performed structurally through all the sustained load and gust load tests with no change in air leakage rate being observed and no significant physical change to the assembly components. During the +500 Pa sustained load test, several nail heads became partially embedded in the fibreboard.
10 Fibreboard/ Polyethylene/ Gypsum 0.006 -1,000 Pa +1,000 Pa -2,500 Pa +2,000 Pa
  Non-asphalt impregnated fibreboard (7/16-inch) installed vertically with 1 3/4-inch galvanized roofing nails every 4 inches along edges and 8 inches along the intermediate supports, 6-mil polyethylene film with 4-inch overlap at the centre, stapled to fibreboard with 1/4-inch staples every 2 feet o.c., 1/2-inch horizontal gypsum boards installed on the polyethylene with 2-inch drywall screws every 8 inches along edges and 12 inches along intermediate supports. The gypsum was not sealed. No change in air leakage was observed after the negative tests, nor after the positive sustained load tests. The system failed structurally during the +2,000 Pa gust load test: the top gypsum board pulled off from the screws at one end.

Implication for the Housing Industry

When designers specify an air barrier system, in addition to materials, the designer must consider wind load conditions specified by the NBC, as these will affect the method of attaching and joining the different materials.

Most of the air barrier systems tested would function under local wind conditions in most urban centres in Canada and would achieve the air leakage rates proposed by IRC.

Several of the air barrier systems were airtight but failed structurally during the positive gust load testing. Clearly, firm attachment of the air barrier system to the supporting structure is very important. In most cases, a tighter nailing schedule or the use of nails with larger heads or strapping may solve the problem of attachment weaknesses.

Project Manager: Jacques Rousseau

Research Consultant: Institute for Research in Construction, National Research Council Canada

Research Report: Testing of Air Barrier Systems for Wood Frame Walls, 1988

A full report on this research project is available from the Canadian Housing Information Centre at the address below.

A full report on this research project is available
from the Canadian Housing Information Centre
.

The information in this publication represents the latest knowledge available to CMHC at the time of publication, and has been thoroughly reviewed by experts in the housing field. CMHC, however, assumes no liability for any damage, injury, expense or loss that may result from use of this information.

©1999 CMHC-SCHL. All rights reserved.