august 29, 2005 hurricane katrina - norbord€¦ · the hurricane shortly after landfall and a map...

THE ENGINEEREDWOOD ASSOCIATION

APA

A U G U S T 2 9 , 2 0 0 5

Hurricane Katrina

ByThomas D. Skaggs, Ph.D., P.E. – APA Technical Services Division

Bryan T. Readling, P.E. – APA Field Services Division

STRUCTURAL PERFORMANCE OF WOOD-FRAME BUILDINGS IN THE AFTERMATH

Hurricane Katrina – Structural Performance of Wood-Frame Buildings in the Aftermath

Form No. SPE-1125 ■ © 2005 APA – The Engineered Wood Association ■ www.apawood.org

2

BACKGROUND

On August 29, 2005, Hurricane Katrina made landfall at 6:10 a.m. CDT in Plaquemines Parish, Louisiana, between GrandIsle, Louisiana, and the mouth of the Mississippi River. Although Hurricane Katrina had reached Category 5 status while inthe Gulf of Mexico, the storm had been tempered to a Category 3 storm by the time it made landfall, with sustained windspeeds of approximately 120 mph (Graumann et al., 2005; Knabb et al., 2005). Figures 2 and 3 show a satellite image ofthe hurricane shortly after landfall and a map depicting the path and sustained wind speeds of the hurricane, respectively.According to a special report by McGraw Hill Construction, Hurricane Katrina is the most expensive U.S. natural disasterin history. The total estimated cost of the destruction is in excess of $125 billion; by contrast, 1992’s Hurricane Andrewdamage was estimated to be $37 billion, adjusted to 2005 dollars (Murray, 2005).

The authors of this report were dispatched in two independent teams approximately three weeks after the hurricane.Although there is obvious concern with the loss of perishable data due to routine clean-up work, including temporaryweatherproofing of damaged structures, both groups intentionally waited several weeks to not interfere with emergency aid.

APA Field Services Engineers Bryan Readling, P.E., and David Lewis toured Biloxi, Bay St. Louis, and Waveland, Mississippi,and Slidell, Louisiana, from September 20 to 22, as part of a 43-person investigation team affiliated with the Institute forBusiness & Home Safety (IBHS), a nonprofit insurance industry group (Figure 4). APA Technical Services Senior EngineerThomas D. Skaggs, Ph.D., P.E., toured 300 miles along the Gulf Coast, from as far east as Biloxi to as far west asDiamondhead and Waveland, Mississippi, from September 23 to 25, as part of a 5-person investigation team affiliated withthe National Science Foundation (NSF) (Figure 5). The findings and recommendations from these two teams are includedin this report. The main focus of the two expeditions was to study the structural behavior of wood-frame constructionaffected by Hurricane Katrina winds.

OBSERVATIONS

Much of the damage from Katrina was related to either flooding from the breaches in the New Orleans levees or stormsurge in excess of 25 feet. Engineering News Record has estimated that more than 90,000 square miles in the Gulf Coast areawere near total devastation (Post, 2005). Figures 6, 7, 8, and 9 demonstrate typical damage observed on the Mississippicoast and in the vicinity of Lake Pontchartrain, Louisiana. Unfortunately, many coastal homes were completely destroyed,as shown in Figure 6. Flooding was not isolated to homes geographically close to the coast; Figure 10 shows a home farinland from water that was still subjected to 8 feet of storm surge, which is believed to have caused the brick veneer to fail.

No quick fix for reducing flooding damage exists, but APA believes improved construction techniques and recommenda-tions could mitigate structural damage due to wind. Figure 1 is a ground wind speed map that one of the NSF team mem-bers developed based on National Oceanic and Atmospheric Administration (NOAA) wind speed data. Although theauthors of this report have seen several wind speed maps that show higher wind speeds, it is believed that the mapsdemonstrating higher wind speeds used results from radar based at 3,000–4,000 feet above sea level. The wind speed mapin Figure 1 has been projected to ground level and adjusted for ground-based weather stations. It should be noted that thatBasic Wind Speed according to ASCE 7 for the coastal area is 130–150 mph. Also on Figure 1 are the sites the NSF teamvisited, which are more fully documented in the NSF Report by van de Lindt et al. (2005).



3

FIGURE 1

GROUND WIND SPEED CONTOURS BASED ON NOAA DATA, MAP GENERATED BY ANDREW GRAETTINGER,UNIVERSITY OF ALABAMA, FOR THE NATIONAL SCIENCE FOUNDATION REPORT. ST LOUIS BAY, MISSISSIPPI, ISTHE BAY ON THE LEFT HAND SIDE OF THE MAP, AND THE BILOXI, MISSISSIPPI, BAY IS ON THE RIGHT HAND SIDEOF FIGURE 1.



4

Although it is beyond the main objective of this report, it should be noted that the observed damage was not limited onlyto wood-frame construction. Figures 11 and 12 show damage to metal buildings, and Figures 13 and 14 show damage to acommercial building with concrete masonry unit (CMU) walls and wood-framed roof. The only connection between theCMU walls and the wood roof was the light gauge steel flashing; essentially, this was a freestanding CMU wall next to thewood roof. The backside of the building experienced similar failure, but only the gable area of the wall collapsed – appar-ently ties were sufficient to prevent complete wall collapse at the second story. The failure mechanism in this building isquite similar to poor wall-to-roof details as highlighted by numerous seismic events.

Figure 15 illustrates a similar failure to Figures 13 and 14. This apartment complex was built in the mid-1980s. Note the 4-foot-long braced wall panels at each end of the wall and at each story height. Apparently, this apartment complex was builtfollowing prescriptive construction provisions. The remainder of the wall was sheathed with foam sheathing. The finish onthis wall was a brick veneer that was inadequately tied to the wall studs. These buildings had an architectural parapet,which appeared to have allowed pressurization of the walls. Once the brick veneer was lost, the foam infill panels were alsolost, resulting in a significant amount of water damage. Figure 16 shows another two-story apartment complex that wasbuilt in 1978. The construction techniques of these two apartment complexes were quite similar. In this case, the infill pan-els were fiberboard instead of foam sheathing. On the wall line illustrated in Figure 16, the stucco became detached fromthe wood framing and collapsed. Stucco ties were connected directly to the fiberboard. Although this apartment complexwas subjected to 6-foot storm surge, the loss of fiberboard on the second story caused additional loss to that story and thefirst story ceiling.

An important observation was that recently built single-family dwellings generally performed well (Figure 18). It was impos-sible to determine if completed new homes were built to modern codes but empirical evidence suggests that code changesimplemented after Hurricane Andrew resulted in improved structural performance. Figure 18 shows a large residence thathad recently been framed. This home was in a subdivision of large homes in Gulfport, Mississippi. Although this subdivi-sion was located less than 2.5 miles from the Gulf Coast, very little damage was observed. In the home shown in Figure18, there was a complete load path from the foundation to the roof, and all walls were fully sheathed with wood structuralpanels. It should be noted that this home had very narrow garage fronts, and it is unlikely that this garage front would meetprescriptive code requirements, including the APA Narrow Wall Bracing Method. Although this home was well built, thegarage area of this home may be vulnerable to future high wind events.

Figure 19 shows a Biloxi apartment complex built after Hurricane Andrew. Given the dimensions of the apartment complexand the individual units, these structures were most likely engineered. This apartment complex sustained virtually no struc-tural damage. This complex was 1/4-mile north of another apartment complex (Figure 15), built pre-Hurricane Andrew,that sustained heavy damage.

A notable exception to the good performance of new homes can be seen in Figures 20–25. Figures 20–23 were taken in alarge subdivision built between 1999 and 2005, located west of Gulfport, Mississippi. Based on wind speeds presented inFigure 1, these homes were subjected to wind speeds ranging from 95 to 100 mph. Even though the estimated winds werenot much different than the wind speeds that govern conventional construction in most of the U.S., approximately 75% ofthe homes in this subdivision were damaged. The vast majority of this damage could be classified as loss of vinyl siding andloss of foam sheathing in the wall and gable areas of the structures.

http://www.apawood.org/bracing



5

Of the homes that were damaged, all were partially sheathed with wood structural panels infilled with foam sheathing.Upon closer inspection, the siding was attached with 3/4-inch staples directly into both the OSB and the foam sheathing,which was attached to the wall studs with staples. There was no weather-resistant barrier between the siding and thesheathing, which can lead to other long-term durability problems highlighted in Figure 26. The foam sheathing did nothave sufficient withdrawal resistance to prevent the siding from being lost. Also note in Figure 21, vinyl siding remainedintact in portions of the wall sheathed with OSB, and the siding was non-existent in areas with the foam sheathing. Oncethe siding was lost, it appears that the foam sheathing could not withstand negative wind pressure pulling the staplesthrough the face of the foam sheathing. After the foam sheathing was blown off the structures, a significant amount ofwater intrusion resulted in damage to the interior of the homes and contents. In one case, the estimated repair cost of thehome was equal to the original purchase price of the home in 1999 (van de Lindt et al., 2005).

The homes in Figures 24 and 25 were close to Biloxi, Mississippi. The loss of foam sheathing in the gable area is essentiallyidentical to the type of damage observed in Figures 20–22. It should be highlighted that in terms of structural perform-ance, many of these homes were not significantly damaged, but the loss of non-structural sheathing obviously led to a largefinancial loss due to water intrusion. Some researchers have suggested that wall sheathing should have the capability ofresisting 30 psf wind pressure, when design wind speeds would generate this level of pressure. This requirement was in the1995 CABO One and Two Family Dwelling Code, but is not included in the International Residential Code. The require-ment in the International Residential Code, however, is that sheathing must meet the wall component and cladding designpressures, which in some cases may be less than 30 psf.

The home shown in Figure 26 was consistent with an entire community built around 1990, before Hurricane Andrew.These homes were located in Gulfport, Mississippi, less than 4 miles east of the homes highlighted in Figures 20–22. Windspeeds for these homes were also in the 95–100 mph range. Many homes in this community lost a significant amount ofvinyl siding. This siding was attached with nails and no weather-resistant barrier was between siding and the OSB. Uponclose inspection, it appeared that the source of the failure was not nail withdrawal but rather nail heads pulling through thenailing flanges of the vinyl siding (intact nails can clearly be seen in some of the photos). This might indicate that the per-formance of vinyl siding has improved after Hurricane Andrew. In Figure 26, note the water damage below the second-story window and under the porch roof. Although these homes appear to be heavily damaged, the authors of this reportestimate damage to be significantly less than the homes featured in Figures 20–22, since water damage from failed wallsheathing was avoided in the interior of the house.

Another light-frame structure that lost siding is shown in Figure 27. This detached garage was apparently built using pre-scriptive provisions. Plywood was observed at each end of the garage front. The remainder of this wall and gable areas weresheathed with fiberboard. It is speculated that the wind breached the garage door and pressurization resulted in the loss ofgable sheathing and siding. Even with the narrow garage front (oriented E-W), little structural damage was observed withthe exception of the loss of infill fiberboard sheathing. Also note that in areas of the garage where the vinyl siding wasattached to plywood, the siding remained intact, yet in the areas of the fiberboard sheathing, the siding was lost.

Of the homes observed in the Gulf Coast area, most homes performed well. Homes that lost shingles were common, how-ever (Figures 7, 17, 19, 28, and 34), as were homes that lost trim (Figures 17, 20, 21, 22, 24, 26, and 29). It was uncom-mon to see roof sheathing loss. The home in Figure 29 was one of the most severe cases of roof sheathing loss observed bythe teams. More typically a roof might lose one piece of sheathing. Nonetheless, loss of roof sheathing, independent fromthe framing, was a relatively rare occurrence as compared to other large and forceful hurricanes. In both of the authors’independent observations, when loss of roof sheathing occurred, it was related to inadequate attachment to roof framing.In some cases fasteners were spaced in excess of the code minimum (Figure 30), and in other cases the fasteners were misapplied.



6

It should be noted that the design wind speeds in this area were in the range of 130–150 mph (2003 IRC), depending onthe distance from the Gulf Coast. These wind speeds are beyond the scope of prescriptive provisions in both theInternational Building Code (ICC, 2003a) and the International Residential Code (ICC, 2003b). APA has roof fastener rec-ommendations for these higher wind speeds in Roof Sheathing Fastening Schedules for Wind Uplift (APA Form T325) that sug-gest fasteners should be spaced closer than 6 inches on center on the panel edges and 6 inches on center in the field of thepanel.

Even more rare than loss of roof sheathing was significant wind damage to gable-end walls. Figure 31 is a photo of the lossof a gable end truss, which was a common problem in Hurricane Andrew. This failure was associated with a gross under-use of fasteners in the gable end area of this truss. As noted in Figure 15, 16, 20, 21, 22, 24, 25 and 27, a more commontype of damage associated with gabled roofs was the loss of non-structural sheathing in the gable areas after the exteriorcladding was lost.

One of the more dramatic failures noted is illustrated in Figure 32. These condominiums were originally built in the 1970swith flat roofs (based on interviews with occupants). Around 1980, the flat roofs were retrofitted with a sloped truss roof.Upon inspection of the roof failure in Figure 32, it was noted that truss was tied to a top plate, but the connection from theplate to the rest of the structure was obviously inadequate. Figure 33 shows a similar detail to the roof portion that pulledthe top plate off the structure. The homemade roof clip was fastened with two 6d nails. Current hurricane ties use four 8dnails for a similar type of code-recognized connector.

http://www.apawood.org/level_c.cfm?content=pub_searchresults&pK=t325&pT=Yes&pD=Yes&pF=Yes&CFID=1496624&CFTOKEN=98490594



7

CONCLUSIONS AND RECOMMENDATIONS

It should be emphasized that wood-frame structures subjected to high wind (i.e., no surge damage) in the Gulf Coastregion performed very well. There were some trends that were observed, resulting in the following conclusions and recommendations:

1.) Non-structural wall sheathing is not adequate to resist design pressures expected along a hurricane coastline or in areaswith lower estimated wind speeds of 100 mph, based on the observed damage and the estimated wind speeds shown inFigure 1 of this report. Costly water damage to building interiors and contents could be minimized by fully sheathing struc-tures with wood structural panels. Wood structural panel sheathing will meet the wall component and cladding pressurerequirements of the IBC and IRC.

2.) All roof sheathing loss observed by the APA teams was due to attachment schedules or practices that did not meet eventhe minimum code requirements for any basic wind speed. For areas prone to hurricane-force winds, such as most GulfCoast regions, additional nailing above the tabulated, code minimum is required. Refer to Roof Sheathing Fastening Schedulesfor Wind Uplift (APA Form T325) and Retrofitting a Roof for High Wind Uplift (APA Form A410) for more information.

3.) In coastal areas, high design wind speeds often result in homes that are outside the scope of conventional constructionprovisions. Regardless, many structures that were built according to conventional construction provisions performed verywell in the areas that experienced winds in excess of 105 mph.



8

REFERENCES

APA DocumentsAPA Data File: Roof Sheathing Fastening Schedules for Wind Uplift, Form T325

APA Data File: Retrofitting a Roof For High Wind Uplift, Form A410

Narrow Walls That Work, Form D420

Other DocumentsASCE. 2002. Minimum Design Loads for Buildings and Other Structures – ASCE 7. American Society of Civil Engineers,Reston, Virginia.

CABO. 1995. CABO One and Two Family Dwelling Code. Council of American Building Officials (CABO), Falls Church, Virginia.

Graumann, A., T. Houston, J. Lawrimore, D. Levinson, N. Lott, S. McCown, S. Stephens, D. Wuertz. 2005. HurricaneKatrina – A Climatological Perspective – Preliminary Report – Technical Report 2005-01. NOAA’s National Climatic Data Center,Asheville, NC.

ICC. 2003a. International Residential Code For One- and Two-Family Dwellings. International Code Council (ICC), Country Club Hills, Illinois.

ICC. 2003b. International Building Code. International Code Council (ICC), Country Club Hills, Illinois.

Post, N. M. 2005. “Hard-Hit Areas Face Tough Question.” Engineering News-Record. October 10, 2005: 10–12.

Web SitesKnabb, R.D., J.R. Rhome, and D.P. Brown. Tropical Cyclone Report: Hurricane Katrina 23-30 August 2005. National HurricaneCenter. Available at: www.nhc.noaa.gov/pdf/TCR-AL122005_Katrina.pdf

Murray, R. A. 2005. Hurricane Katrina: Implications for the Construction Industry. Special Report, McGraw-Hill Construction.Available at: www.construction.com

Van de Lindt, J. W., A. J. Graettinger, R. Gupta, S. E. Pryor, T. D. Skaggs and K. J. Fridley. 2005. Damage Assessment ofWoodframe Residential Structures in the Wake of Hurricane Katrina. A Preliminary Report to National Science Foundation.Available at: www.engr.colostate.edu/~jwv/hurricane-Katrina-woodframe.htm

www.engr.colostate.edu/~jwv/hurricane-Katrina-woodframe.htm

www.engr.colostate.edu/~jwv/hurricane-Katrina-woodframe.htm

www.nhc.noaa.gov/pdf/TCR-AL122005_Katrina.pdf

http://www.construction.com/AboutUs/20050909pr.asp



9

FIGURE 2

NASA Satellite Image, August 29, 2005, 7:20 a.m. CDT, one hour after Katrina had made landfall. Photo courtesy of NASA.

FIGURE 3

Hurricane Katrina intensity and path, courtesyof Steven Babin and Ray Sterner of the JohnsHopkins University Applied Physics Laboratory.

FIGURE 4

APA representatives participated in Institute forBusiness & Home Safety (IBHS) investigationteam. This multifamily structure in Slidell,Louisiana, was no match for storm surge.



10

FIGURE 5

The National Science Foundation DamageAssessment Team used a mobile base camp for structural observation along the MississippiGulf Coast.

FIGURE 6

Only the foundation remained of a home inWaveland, Mississippi, directly off the GulfCoast. This failure is representative of more than35 miles of coastal devastation from stormsurge.

FIGURE 7

This OSB-sheathed house in Bay St. Louis,Mississippi, located approximately 100 feet from its original foundation, was another victimof storm surge.



11

FIGURE 8

Storm surge destroyed all but the steel columnsin the lower portion of this gulf-front home inWaveland, Mississippi. This home was the leastdamaged for at least a mile on either side alongthe coast.

FIGURE 9

Older homes at the edge of Lake Pontchartrainfared poorly under the force of water.

FIGURE 10

Single family home far inland from the gulf butflooded with approximately 8 feet of water. Theauthors of this report believe that the walls werepushed out not by wind (note shingles intact),but by water quickly exiting the home.



12

FIGURE 11

This airport hangar near Hancock Co. HighSchool, Kiln, Mississippi, sustained damage.

FIGURE 12

APA representatives that participated in IBHSinvestigation team stand in front of the ruins of acommercial metal building.

FIGURE 13

A commercial building in Biloxi, Mississippi,used a wood truss system and wood structuralpanels under a standing seam metal roof. Theonly observed tie from the CMU endwall to theroof was flashing. The end wall separated fromthe roof as a complete unit. The backside of thisbuilding is pictured in Figure 14.



13

FIGURE 14

The gable area of the two-story portion of thebuilding shown in Figure 13 also collapsed as a unit.

FIGURE 15

This Biloxi, Mississippi, apartment complex wasbuilt to conventional construction provisions.The infilled panels were foam sheathing. Theends of these apartments were finished with abrick veneer that was inadequately tied to thewall. Once the veneer was lost, the foam sheath-ing was also lost allowing significant buildingdamage caused by rain infiltration.

FIGURE 16

This Gulfport, Mississippi, apartment complexwas located 1-1/2 miles from the Gulf Coast andwas subjected to a 6-foot storm surge. The exte-rior finish was stucco with a brick-like appear-ance. The stucco ties to the fiberboard did notremain intact, perhaps due to the wetting of thefiberboard, resulting in a total loss of this unit.Although flooding damaged the first floor ofthese units, the loss of the stucco finish resultedin even further flood damage. The apartmentsdid not appear to have significant structuraldamage.



14

FIGURE 17

This new development in Gulfport, Mississippi,was adjacent to areas that sustained vinyl sidingloss and roof damage. This home experiencedsome of the worst damage in the neighborhoodwith relatively minor shingle damage and sometrim loss. Just to the north of this home was ahome with vinyl siding that was “Dade CountyApproved.” The vinyl siding on the adjacenthouse was intact. Although some homes wereobserved with vinyl siding damage, in this area,the vinyl siding performed well.

FIGURE 18

Homes in the Brighton Place subdivision inGulfport, Mississippi, experienced very littlestructural damage. It is believed that this homewas representative of the construction in thisdevelopment. There was a complete load pathfrom foundation to roof and the walls were fullysheathed with wood structural panels.

FIGURE 19

This three-story wood framed apartment com-plex in Biloxi, Mississippi, was built post-Hurricane Andrew (1998) and sustainedvirtually no damage. The picture shows a win-dow breach. There was some shingle loss in thecomplex, and one unit lost a minor amount offiber-cement siding on the third story.



15

FIGURE 20

This image is representative of a common failurefound in this neighborhood north of I-10, westof Gulfport, Mississippi. Walls were partiallysheathed with wood structural panels and fiber-glass covered foam (Figure 23), and wall sheath-ing was stapled to framing under vinyl siding.Wall systems failed due to wind pressure onapproximately 75% of the homes in this neigh-borhood, resulting in water infiltration and damage to interior of homes and contents.

FIGURE 21

This home was in the same neighborhood asFigure 17. These structures were built between1999 and 2005. The vinyl siding was attachedwith 3/4-inch staples that withdrew from thefoamboard. Intact staples were noted in someportions of the OSB that did in fact lose vinylsiding. Note intact vinyl siding in garage areathat was sheathed with wood structural panels,and lack of vinyl siding in areas of the foamsheathing. Home in background with housewrap was post-Katrina repair work. The home-owner chose to fully sheath the home with OSB.

FIGURE 22

This home was in the same neighborhood asFigures 17 and 18. Foam sheathing was inade-quate to resist the wind pressures on the gableends. Massive water infiltration and damage tohome and contents resulted. Note debris piledbeside the road for deposit into landfill.According to a homeowner in this neighbor-hood, the estimated repair cost of his home wasequal to the original purchase price of a newhome in 1999 (van de Lindt et al., 2005).



16

FIGURE 23

A close-up of the foam sheathing that blew offhomes illustrated in Figures 17, 18 and 19.

FIGURE 24

Foam sheathing and vinyl were lost on gable-endin Sunkist (Biloxi), Mississippi, just south ofI-10. Massive water infiltration and damage to contents resulted.

FIGURE 25

More sheathing was lost at a gable end inSunkist, Mississippi.



17

FIGURE 26

This neighborhood in Gulfport, Mississippi, wasbuilt in 1990, pre-Hurricane Andrew. Thisneighborhood experienced significant loss ofvinyl siding. There was no weather-resistant bar-rier between the siding and the OSB. Note theareas under the second story window and theroof/wall interface that experienced deteriorationdue to poor detailing and repeated moistureintrusion. Vinyl siding manufacturers typicallyrecommend installation of a weather-resistantbarrier between the vinyl siding and the substrate.

FIGURE 27

This single car garage in Gulfport, Mississippi,was partially sheathed with wood structural pan-els at each end of the walls, and the remainderof the walls were sheathed with fiberboard pan-els. It appears that the garage door was breachedallowing pressurization of the garage. Vinyl sid-ing was also lost in the event; note that the vinylsiding remained intact in the ends of the garagefront that contained wood structural panels.Also note that there was little structural damageto this garage; no wall racking could bedetected.

FIGURE 28

On these OSB roofs in Slidell, Louisiana, shin-gles are stripped but roof sheathing held fast.Older homes in the foreground were swept awayby water from Lake Pontchartrain.



18

FIGURE 29

The APA Disaster Assessment Team evaluates arare case of roof sheathing loss in Slidell,Louisiana. See next photo for a close-up.

FIGURE 30

Staples were located greater than 12 inches oncenter at panel ends and many were missing the framing with one or both legs on this damaged roof.

FIGURE 31

This is one of the few gable-end trusses foundthat was damaged, a vulnerability to gable roofconstruction in high winds. On this house, theroof sheathing was grossly under-fastened in thearea of the gable end.



19

� FIGURE 32

Condominiums in Diamondhead, Mississippi,showed damage similar to that documented inthe IBHS assessment photos. The original con-dominiums were built in early 1970s with flatroofs. The condos were architecturally upgradedaround 1980 with a sloped truss roof. The trussroof was inadequately tied to the existing flatroof or wall, thus resulting in the roof becomingcompletely detached from the condominium.

FIGURE 33

Minimal size homemade roof clips with 2-6dbox nails (very small) into the truss at bearingcontributed to roof failure.



THE ENGINEEREDWOOD ASSOCIATION

APA

We have field representatives in many major U.S. cities and in Canada who can help answer questions involving APA trademarked products.

For additional assistance in specifying engineered wood products, contact us:

APA – THE ENGINEERED WOOD ASSOCIATION

HEADQUARTERS7011 So. 19th St. ■ Tacoma, Washington 98466 ■ (253) 565-6600 ■ Fax: (253) 565-7265

PRODUCT SUPPORT HELP DESK(253) 620-7400 ■ E-mail Address: [email protected]

DISCLAIMERThe information contained herein is based on APA – The Engineered Wood Association’s continu-ing programs of laboratory testing, product research, and comprehensive field experience. NeitherAPA, nor its members make any warranty, expressed or implied, or assume any legal liability orresponsibility for the use, application of, and/or reference to opinions, findings, conclusions, orrecommendations included in this publication. Consult your local jurisdiction or design professionalto assure compliance with code, construction, and performance requirements. Because APA has nocontrol over quality of workmanship or the conditions under which engineered wood products areused, it cannot accept responsibility for product performance or designs as actually constructed.

Issued December 2005

http://www.apawood.org