One-story buildings are much less vulnerable to wind damage than two- and three-story buildings.
- On average, hip roofs have demonstrated better performance than gable-end roofs.
Moderate roof slopes (4:12 to 6:12) tend to optimize the trade-off between lateral loads and roof uplift loads (i.e., more aerodynamically efficient).
Roof sheathing installation should be inspected for the proper type and spacing of fasteners, particularly at connections to gable-end framing.
The installation of metal strapping or other tie-down hardware should be inspected, as required, to ensure the transfer of uplift loads.
If composition roof shingles are used, high-wind fastening requirements should be followed (i.e., 6 nails per shingle in lieu of the standard 4 nails). A similar concern exists for tile roofing, metal roofing, and other roofing materials.
Consider some practical means of glazed opening protection in the most severe hurricane-prone areas.
For design purposes, snow is typically treated as a simple uniform gravity load on the horizontal projected area of a roof. The uniformly distributed design snow load on residential roofs can be easily determined by using the unadjusted ground snow load. This simple approach also represents standard practice in some regions of the United States; however, it does not account for a reduction in roof snow load that may be associated with steep roof slopes with slippery surfaces (refer to ASCE 7-98). To consider drift loads on sloped gable or hip roofs, the design roof snow load on the windward and leeward roof surfaces may be determined by multiplying the ground snow load by 0.8 and 1.2, respectively. In this case, the drifted side of the roof has 50% greater snow load than the non-drifted side of the roof. However, the average roof snow load is still equivalent to the ground snow load.
Design ground snow loads may be obtained from the map in Figure 3.3; however, snow loads are usually defined by the local building department. Typical ground snow loads range from 0 psf in the South to 50 psf in the northern United States. In mountainous areas, the ground snow load can surpass 100 psf such that local snow data should be carefully considered. In areas where the ground snow load is less than 15 psf, the minimum roof live load is usually the controlling gravity load in roof design. For a larger map with greater detail, refer to ASCE 7-98.
FIGURE 3.3 Ground Snow Loads (ASCE 7-98)
Map from American Society of Civil Engineers, ASCE http://publicecodes.cyberregs.com/icod/ibc/index.htm
This section provides a simplified earthquake load analysis procedure appropriate for use in residential light-frame construction of not more than three stories above grade. The lateral forces associated with seismic ground motion are based on fundamental Newtonian mechanics (F = ma
) expressed in terms of an equivalent static load. The method provided in this section is a simplification of the most current seismic design provisions. It is also similar to a simplified approach found in more recent building code development (ICC).
Most residential designers use a simplified approach similar to that in older seismic design codes. The approach outlined in the next section follows the older approach in terms of its simplicity while using the newer seismic risk maps and design format of NEHRP-97 as incorporated into recent building code development efforts (ICC); refer to Figure 3.4.
In general, wood-framed homes have performed well in major seismic events, probably because of, among many factors, their light-weight and resilient construction, the strength provided by nonstructural systems such as interior walls, and their load distribution capabilities. Only in the case of gross absence of good judgment or misapplication of design for earthquake forces have severe life-safety consequences become an issue in light-frame, low-rise structures experiencing extreme seismic events.
FIGURE 3.4 Seismic Map of Design Short-Period Spectral Response Acceleration (g) (2 percent chance of exceedance in 50 years or 2,475-year return period)
Map from American Society of Civil Engineers, ASCE http://publicecodes.cyberregs.com/icod/ibc/index.htm
Determination of Earthquake Loads on Houses
The total lateral force at the base of a building is called seismic base shear. The lateral force experienced at a particular story level is called the story shear. The story shear is greatest in the ground story and least in the top story. Seismic base shear and story shear (V
) are determined in accordance with the following equation:
When determining story shear for a given story, the designer attributes to that story one-half of the dead load of the walls on the story under consideration and the dead load supported by the story. For housing, the interior partition wall dead load is reasonably accounted for by the use of a 6 psf load distributed uniformly over the floor area. When applicable, the snow load may be determined. The inclusion of any snow load, however, is based on the assumption that the snow is always frozen solid and adhered to the building such that it is part of the building mass during the entire seismic event.
The design spectral response acceleration for short-period ground motion SDS
is typically used because light-frame buildings, such as houses, are believed to have a short period of vibration in response to seismic ground motion (i.e., high natural frequency). In fact, non-destructive tests of existing houses have confirmed the short period of vibration, although once ductile damage has begun to occur in a severe event, the natural period of the building likely increases.
Values of Ss
are obtained from Figure 3.7. For a larger map with greater detail, refer to ASCE 7-98. The value of SDS
should be determined in consideration of the mapped short-period spectral response acceleration Ss and the required soil site amplification factor Fa
The value of Ss
ranges from practically zero in low-risk areas to 3g in the highest-risk regions of the United States. A typical value in high seismic areas is 1.5g. In general, wind loads control the design of the lateral force-resisting system of light-frame houses when Ss
is less than about 1g. The 2/3 coefficient in Equation 3.8-2 is used to adjust to a design seismic ground motion value from that represented by the mapped Ss
values (i.e., the mapped values are based on a “maximum considered earthquake” generally representative of a 2,475-year return period, with the design basis intended to represent a 475-year return period event).
Table 3.11 provides the values of Fa
associated with a standard “firm” soil condition used for the design of residential buildings. Fa
decreases with increasing ground motion because the soil begins to dampen the ground motion as shaking intensifies. Therefore, the soil can have a moderating effect on the seismic shear loads experienced by buildings in high seismic risk regions. Dampening also occurs between a building foundation and the soil and thus has a moderating effect. However, the soil-structure interaction effects on residential buildings have been the topic of little study; therefore, precise design procedures have yet to be developed. If a site is located on fill soils or “soft” ground, a different value of Fa
should be considered. Nonetheless, as noted in the Anchorage Earthquake of 1964 and again 30 years later in the Northridge Earthquake, soft soils do not necessarily affect the performance of the above-ground house structure as much as they affect the site and foundations (e.g., settlement, fissuring, liquefaction, etc.).
TABLE 3.11 Site Soil Amplification Factor Relative to Acceleration (short period, firm soil)
The seismic response modifier R
has a long history in seismic design, but with little in the way of scientific underpinnings. In fact, it can be traced back to expert opinion in the development of seismic design codes during the 1950s (ATC, 1995). In recognition that buildings can effectively dissipate energy from seismic ground motions through ductile damage, the R
factor was conceived to adjust the shear forces from that which would be experienced if a building could exhibit perfectly elastic behavior without some form of ductile energy dissipation. The concept has served a major role in standardizing the seismic design of buildings even though it has evolved in the absence of a repeatable and generalized evaluation methodology with a known relationship to actual building performance.
Those structural building systems that are able to withstand greater ductile damage and deformation without substantial loss of strength are assigned a higher value for R
. The R
factor also incorporates differences in dampening that are believed to occur for various structural systems. Table 3.12 provides some values for R
that are relevant to residential construction.
TABLE 3.12 Seismic Response Modifiers for Residential Construction
1. The R factors may vary for a given structural system type depending on wall configuration, material selection, and connection detailing, but these considerations are necessarily matters of designer judgment.
2. The R for light-frame shear walls (steel-framed and wood-framed) with shear panels has been recently revised to 6 but is not yet published (ICC, 1999). Current practice typically uses an R of 5.5 to 6.5, depending on the edition of the local building code.
3. The wall is reinforced in accordance with concrete design requirements in ACI-318 or ACI-530. Nominally reinforced concrete or masonry that has conventional amounts of vertical reinforcement, such as one #5 rebar at openings and at 4 feet on center, may use the value for reinforced walls, provided the construction is no more than two stories above grade.
Seismic Shear Force Distribution
As described in the previous section, the vertical distribution of seismic forces to separate stories on a light-frame building is assumed to be in accordance with the mass supported by each story. However, design codes vary in the requirements related to vertical distribution of seismic shear. Unfortunately, there is apparently no clear body of evidence to confirm any particular method of vertical seismic force distribution for light-frame buildings. Therefore, in keeping with the simplified method, the approach used in this article reflects what is considered conventional practice. The horizontal distribution of seismic forces to various shear walls on a given story also varies in current practice for light-frame buildings. Until methods of vertical and horizontal seismic force distribution are better understood for application to light-frame buildings, the importance of designer judgment cannot be overemphasized.
Special Seismic Design Considerations
Perhaps the single most important principle in seismic design is to ensure that the structural components and systems are adequately tied together to perform as a structural unit. Underlying this principle are a host of analytic challenges and uncertainties in actually defining what “adequately tied together” means in a repeatable, accurate, and theoretically sound manner.
Recent seismic building code developments have introduced several new factors and provisions that attempt to address various problems or uncertainties in the design process. Unfortunately, these factors appear to introduce as many uncertainties as they address. Codes have tended to become more complicated to apply or decipher, perhaps detracting from some important basic principles in seismic design that, when understood, would provide guidance in the application of designer judgment. Many of the problems stem from the use of the seismic response modifier R
, which is a concept first introduced to seismic design codes in the 1950s.
Also known as “reserve strength,” the concept of overstrength is a realization that a shear resisting system’s ultimate capacity is usually significantly higher than required by a design load as a result of intended safety margins. At the same time, the seismic ground motion (load) is reduced by the R
factor to account for ductile response of the building system, among other things. Thus, the actual forces experienced on various components (i.e. connections) during a design level event can be substantially higher, even though the resisting system may be able to effectively dissipate that force. Therefore, overstrength factors have been included in newer seismic codes with recommendations to assist in designing components that may experience higher forces than determined otherwise for the building lateral force resisting system using methods similar to Equation 3.8-1. It should be noted that current overstrength factors should not be considered exact and that actual values of overstrength can vary substantially.
In essence, the overstrength concept is an attempt to address the principle of balanced design. It strives to ensure that critical components, such as connections, have sufficient capacity so that the overall lateral force-resisting system is able to act in its intended ductile manner (i.e., absorbing higher-than-design forces). Thus, a premature failure of a critical component (i.e., a restraining connection failure) is avoided. An exact approach requires near-perfect knowledge about various connections, details, safety margins, and system-component response characteristics that are generally not available. However, the concept is extremely important and, for the most part, experienced designers have exercised this principle through a blend of judgment and rational analysis.
The concept of overstrength is relative to the design of restraining connections for light-frame buildings by providing the designer with ultimate capacity values for light-frame shear wall systems. Thus, the designer is able to compare the unfactored shear wall capacity to that of hold-down restraints and other connections to ensure that the ultimate connection capacity is at least as much as that of the shear wall system. Some consideration of the ductility of the connection or component may also imply a response modification factor for a particular connection or framing detail. In summary, overstrength is an area where exact guidance does not exist and the designer must exercise reasonable care in accordance with or in addition to the applicable building code requirements.
The redundancy factor was postulated to address the reliability of lateral force-resisting systems by encouraging multiple lines of shear resistance in a building. It is now included in some of the latest seismic design provisions. Since it appears that redundancy factors have little technical basis and insufficient verification relative to light-frame structures, they are not explicitly addressed in this article. In fact, residential buildings are generally recognized for their inherent redundancies that are systematically overlooked when designating and defining a lateral force-resisting system for the purpose of executing a rational design. However, the principle is important to consider. For example, it would not be wise to rely on one or two shear-resisting components to support a building. In typical applications of light-frame construction, even a single shear wall line has several individual segments and numerous connections that resist shear forces. At a minimum, there are two such shear wall lines in either orientation of the building, not to mention interior walls and other nonstructural elements that contribute to the redundancy of typical light-frame homes. In summary, redundancy is an area where exact guidance does not exist and the designer must exercise reasonable care in accordance with or in addition to the applicable building code requirements.
Deflection amplification has been applied in past and current seismic design codes to adjust the deflection or story drift determined by use of the design seismic shear load (as adjusted downward by the R
factor) relative to that actually experienced without allowance for modified response (i.e., load not adjusted down by the R
factor). For wood-framed shear wall construction, the deflection calculated at the nominal seismic shear load (Equation 3.8-1) is multiplied by a factor of 4. Thus, the estimate of deflection or drift of the shear wall (or entire story) based on the design seismic shear load would be increased four-fold. Again, the conditions that lead to this level of deflection amplification and the factors that may affect it in a particular design are not exact (and are not obvious to the designer). As a result, conservative drift amplification values are usually selected for code purposes. Regardless, deflection or drift calculations are rarely applied in a residential (low-rise) wood-framed building design for three reasons. First, a methodology is not generally available to predict the drift behavior of light-frame buildings reliably and accurately. Second, the current design values used for shear wall design are relatively conservative and are usually assumed to provide adequate stiffness (i.e., limit drift). Third, code-required drift limits have not been developed for specific application to light-frame residential construction. Measures to estimate drift, however, are in terms of nonlinear approximations of wood-frame shear wall load-drift behavior (up to ultimate capacity). In summary, deformation amplification is an area where exact guidance does not exist and predictive tools are unreliable. Therefore, the designer must exercise reasonable care in accordance with or in addition to the applicable building code requirements.
Another issue that has received greater attention in seismic design provisions is irregularities. Irregularities are related to special geometric or structural conditions that affect the seismic performance of a building and either require special design attention or should be altogether avoided. In essence, the presence of limits on structural irregularity speaks indirectly of the inability to predict the performance of a structure in a reliable, self-limiting fashion on the basis of analysis alone. Therefore, many of the irregularity limitations are based on judgment from problems experienced in past seismic events.
Irregularities are generally separated into plan and vertical structural irregularities. Plan structural irregularities include torsional imbalances that result in excessive rotation of the building, re-entrant corners creating “wings” of a building, floor or roof diaphragms with large openings or non-uniform stiffness, out-of-plane offsets in the lateral force resistance path, and nonparallel resisting systems. Vertical structural irregularities include stiffness irregularities (i.e., a “soft” story), capacity irregularities (i.e., a “weak” story), weight (mass) irregularity (i.e., a “heavy” story), and geometric discontinuities affecting the interaction of lateral resisting systems on adjacent stories.
The concept of irregularities is associated with ensuring an adequate load path and limiting undesirable (i.e., hard to control or predict) building responses in a seismic event. Again, experienced designers generally understand the effect of irregularities and effectively address or avoid them on a case-by-case basis. For typical single-family housing, all but the most serious irregularities (i.e., “soft story”) are generally of limited consequence, particularly given the apparently significant system behavior of light-frame homes (provided the structure is reasonably “tied together as a structural unit”). For larger structures, such as low- and high-rise commercial and residential construction, the issue of irregularity--and loads--becomes more significant. Because structural irregularities raise serious concerns and have been associated with building failures or performance problems in past seismic events, the designer must exercise reasonable care in accordance with or in addition to the applicable building code requirements.
A key issue related to building damage involves deformation compatibility of materials and detailing in a constructed system. This issue may be handled through specification of materials that have similar deformation capabilities or by system detailing that improves compatibility. For example, a relatively flexible hold-down device installed near a rigid sill anchor causes greater stress concentration on the more rigid element, as evidenced by the splitting of wood sill plates in the Northridge Earthquake. The solution can involve increasing the rigidity of the hold-down device (which can lessen the ductility of the system, increase stiffness, and effectively increase seismic load), or by redesigning the sill plate connection to accommodate the hold-down deformation and improve load distribution. As a non-structural example of deformation compatibility, gypsum board interior finishes crack in a major seismic event well before the structural capability of the wall’s structural sheathing is exhausted. Conversely, wood exterior siding and similar resilient finishes tend to deform compatibly with the wall and limit observable or unacceptable visual damage. A gypsum board interior finish may be made more resilient and compatible with structural deformations by using resilient metal channels or similar detailing; however, this enhancement has not yet been proven. Unfortunately, there is little definitive design guidance on deformation compatibility considerations in seismic design of wood-framed buildings and other structures.
As a final issue, it should be understood that the general objective of current and past seismic building code provisions has been to prevent collapse in extreme seismic events such that protection of life is reasonably provided, but not with complete assurance. It is often believed that damage can be controlled by use of a smaller R
factor or, for a similar effect, a larger safety factor. Others have suggested using a higher design event. While either approach may indirectly reduce damage or improve performance, it does not necessarily improve the predictability of building performance and, therefore, may have uncertain benefits, if any, in many cases. However, some practical considerations as discussed above may lead to better-performing buildings, at least from the perspective of controlling damage.
Other Load Conditions
Other “forces of nature” may create loads on buildings. Some examples include:
- frost heave;
- expansive soils;
temperature effects; and
In certain cases, forces from these phenomena can drastically exceed reasonable design loads for homes. For example, frost heave forces can easily exceed 10,000 pounds per square foot. Similarly, the force of expanding clay soil can be impressive. In addition, the self-straining stresses induced by temperature-related expansion or contraction of a member or system that is restrained against movement can be very large, although they are not typically a concern in wood-framed housing. Finally, the probability of a direct tornado strike on a given building is much lower than considered practical for engineering and general safety purposes. The unique wind loads produced by an extreme tornado may exceed typical design wind loads by almost an order of magnitude in effect. Conversely, most tornadoes have comparatively low wind speeds that can be resisted by attainable design improvements. However, the risk of such an event is still significantly lower than required by minimum accepted safety requirements.
It is common practice to avoid the loads noted above by using sound design detailing. For example, frost heave can be avoided by placing footings below a “safe” frost depth, building on non-frost-susceptible materials, or using other frost-protection methods. Expansive soil loads can be avoided by isolating building foundations from expansive soil, supporting foundations on a system of deep pilings, and designing foundations that provide for differential ground movements. Temperature effects can be eliminated by providing construction joints that allow for expansion and contraction. While such temperature effects on wood materials are practically negligible, some finishes, such as ceramic tile, can experience cracking when inadvertently restrained against small movements resulting from variations in temperature. Unfortunately, tornadoes cannot be avoided; therefore, it is not uncommon to consider the additional cost and protection of a tornado shelter in tornado-prone areas. A tornado shelter guide is available from the Federal Emergency Management Agency, Washington, D.C.
As noted at the beginning of this article, this article does not address loads from flooding, ice, rain, and other exceptional sources. The reader is referred to ASCE 7 and other resources for information regarding special load conditions.