Schierle, G Goetz¹ and Vergun, Dimitry²

Suggestions for testing light wood structures and components are presented based on failures caused by the Northridge earthquake of January 17, 1994, and other recent earthquakes in Southern California. Observations in conjunction with research regarding quality control for seismic resistant construction are also included. Findings based on forensic investigations over several years, prior to and following the Northridge earthquake are covered as well. Light wood structures were assumed to be resistant to seismic load because of their resilient qualities. However, most of these assumptions have been shattered due to the damages inflicted by the Northridge tremor. While relatively few lives were lost, the amount of property losses has been strikingly high. The testing program will hopefully remedy seismic hazards and provide better understanding of structural component behavior and their interaction as systems for improved performance.

The Northridge earthquake demonstrated that current practice in residential wood construction and design is relatively effective regarding life safety but much less effective to prevent loss of property. While causing only 58 fatalities, many residential wood buildings had been damaged to various degrees. Some even collapsed. As in other earthquakes, structural failures also caused substantial non-structural damage that may be equally costly. A study about non-structural damage (Lagorio, 1990) found that it can amount to as much as 70 percent of future repair and replacement costs. In Los Angeles alone, over 100,000 structures were damaged, many of them residential. Considering the collateral damage, this represents a staggering economic loss to the city and region.

Observations after the Northridge earthquake and research on residential light wood structures conducted prior to it showed that some failures have probably been caused by missing or flawed seismic safety items (Schierle, 1993. 1996). This study included mail surveys of architects and engineers, and site surveys of buildings, after completion of framing, but prior to the application of sheathing. The study found that one third of seismic safety items were missing or flawed in over 40 percent of surveyed structures. It is alarming that key items to resist wind and seismic forces are among those missing or flawed. These include shear wall hold-downs, nailing and proportion (aspect ratio), wall-to-wall straps and tie-downs, diaphragm blocking and nailing, drag strut splice, and roof-to-wall anchors. The Los Angeles Times reported on its April 21, 1994 cover page, that at least one third of the Northridge earthquake failures could be attributed to flawed construction. Much of this may be ascribed to the lack of quality control during construction. Therefore, it is imperative that construction crews and designers gain an understanding of the important role seismic safety items play in the prevention of building failures. Books for architects (Ambrose and Vergun, 1999; Arnold and Reithermen, 1983; Lagorio, 1990) remain significant resources in the education of aspiring designers.

There remains some uncertainty regarding design procedures because the load path in wood structures is not yet fully understood. This is partly due to variation in wood quality and because of the aforementioned missing or flawed seismic safety items. Incorrect load path assumptions cause unexpected levels of stress and displacement that may result in failure. Design decisions based on erroneous assumptions are not reliable. Therefore, validation is needed, for example, regarding the behavior of flexible diaphragms. While tests have been conducted to explore this issue for light wood structures, they are not reliable due to flawed assumptions. Equal loads were applied directly to resisting walls, rather than uniformly distributed over the diaphragm as in actual load conditions. The predictable outcome of equal resistance of the walls suggested behavior like a rigid diaphragm.

The proposed tests outlined in our paper attempt to address three issues: 1) provide better insight of the behavior of wood structures for the design of both new and retrofitted structures; 2) improve the performance of wood structures under seismic load; and 3) provide improved seismic safety by cost effective means. The latter point is most relevant since the appropriate level of seismic safety is often a political and societal issue, weighing economics versus life safety. The affordability of housing is an important factor in such considerations. The testing proposals on the following pages are intended to address these issues.

Fig. 2. Panel orientation parallel to joists

Blocking of floor and roof diaphragms is often missing in residential structures (Schierle, 1993, 1996), probably due to high costs. The reduced strength without blocking may be critical for soft-story and similar conditions, prevalent in apartment buildings. Framing anchors or similar devices could be an economic alternative to blocking. Assemblies with such devices should be tested to determine their relative strength compared to blocking.

Fig. 3. Framing clip for shear transfer

Glued gypsum board

Gypsum wallboards have frequently been used as shear walls. However, tests by Edwin Zacher have shown that their strength is greatly reduced after a few cycles of seismic loading because nails cut slots in boards which drastically reduces their strength. Gluing gypsum boards to the framing, in addition to nailing, might eliminate this problem and provide a cost-effective means of lateral resistance. Glued gypsum board shear panels should be tested to determine their strength and behavior under cyclic seismic loads. Appropriate gluing methods could be developed in cooperation with the respective industry. If the tests prove successful, glued plywood panels could also be tested.

The aspect ratio of plywood shear walls is limited to 1:3.5 by the Uniform Building Code. Since the Northridge earthquake, Los Angeles limits the ratio to 1:2. These ratios are frequently assumed from floor to floor rather than the full height of shear walls. This practice would require the shear walls to be effectively attached at each level to a rim beam capable to resist the overturn moment in bending. However field surveys have shown aspect ratios of 1:18 for shear walls connected to rim joists that are spliced right at the wall (Schierle, 1993, 1996). The resulting bending capacity of such rim joists is drastically reduced. Tests should be conducted to validate the strength of such assemblies.

Fig. 4. Shear wall aspect ratio test model

Shear wall anchor bolts are frequently missing or flawed due to improper alignment, over-drilling, base plate splitting, and other conditions. Shot pin anchors are less subject to such problems because they are installed after the foundations are completed, allowing installation under more favorable working conditions. They could, therefore, provide a cost effective and potentially more reliable alternative to anchor bolts. Shear walls anchored with shot pins should be tested to determine their strengths for various pin spacing.

Fig. 5. Misplaced anchor bolt

Problem configurations

The collapse of the Northridge Meadows apartment complex was a textbook example of configuration problems (Arnold and Reitherman, 1982). Most intersections of connecting wings were ripped apart because of differential movement. While the problem is well documented, effective solutions, except for seismic joints which may not always be desirable, are not fully understood. Methods for seismic strengthening should be developed and tested.

Fig. 6. Roof diaphragm failure due to irregular configuration

Thousands of apartment buildings in Southern California have tuck-under parking supported by steel columns without bending resistance. This soft-story condition puts such buildings at great risk in a major earthquake. Retrofit solutions should be tested to determine their effectiveness.

Fig. 7. Failed soft-story tuck-under parking

Recent code provisions assume reduced wood strength, based on moisture content, with higher values for dry wood. In certain cases this seems to be contradicted by actual performance. Field observations have shown that wood rafters, dried under constant heat, tend to split due to apparently reduced strength. Tests should be conducted to determine strength degradation of wood under very dry conditions as frequently encountered in attics and similar conditions.

Fig. 8. Failed roof beams due to dry wood

Ambrose J and Vergun D (1999) Design for Earthquakes, John Wiley & Sons, Inc.

Arnold C and Reitherman R (1982) Building Configuration and Seismic Design, John Wiley & Sons

Lagorio, H J (1990) Earthquake, An Architects Guide to Nonstructural Seismic Hazards, John Wiley & Sons, Inc.

Schierle, G G (1993) Quality Control in Seismic Resistant Construction, Report to National Science Foundation

Schierle, G G (1996) "Quality Control in Seismic Design and Construction," Journal of Performance of Constructed Facilities, American Society of Civil Engineers