Wood Shoring Systems: How Do They Perform?

By DAVID J. HAMMOND

Because of its availability and versatility, wood shoring is widely used to temporarily stabilize damaged structures during urban search and rescue (US&R) operations. Shoring systems have been developed using wellaccepted engineering principles and design specifications. Testing conducted in California since 2000 has sought to verify predicted strengths and, more importantly, the failure characteristics of vertical and lateral wood shoring systems. The sites included the Menlo Park (CA) Fire District, the Federal Emergency Management Agency (FEMA) US&R California Task Force 3 (CATF3) training site, and the Moffett Field National Aeronautics and Space Administration/Ames Research Center (NASA/ARC) Disaster Assistance and Rescue Team (DART) training site.

Most of the testing occurred during advanced structures specialist training (StS2), funded by the Department of Homeland Security/Federal Emergency Management Agency (DHS/FEMA) US&R Program and the U.S. Army Corps of Engineers (USACE) US&R Program. All tested shoring systems conform to current FEMA US&R standard shoring or are intended to become standard shoring in the near future.

More than 30 tests each were performed on laced post (LP) systems and braced pairs of raker shores. The tests demonstrated that these shores have safety factors of three or higher compared to recommended design loads. For emergency shoring applications, an essential finding was that observers can clearly see visual overload signs at loads much less than those that will cause failure.

VERTICAL LOAD SUPPORTING SYSTEMS

Laced post shores. All LP shore testing was conducted at the NASA/ARC’s Moffett Field. The initial testing in 2001 was “proof of concept” testing, examining the overall behavior and capacities of standard shoring. Further testing (part of StS2 training) studied overloaded shoring behavior and its nearfailure characteristics. Later tests investigated new concept designs, such as those using plywood lacing/bracing instead of traditional 2 × 4 wood lacing, with the purpose of developing the new plywood laced post (PLP). Note the engineering shorthand for large loads or forces: 1,000 pounds = 1 kilopound or 1k.

FEMA design parameters. The following assumptions/characteristics are important for standard FEMA vertical shores:

• Vertical loads should be transferred by direct bearing of header to post and post to sole, using pairs of tapered 2× wood wedges to make height adjustments.
• Systems should be proportioned and braced so that when overloaded, the crushing of the header and sole can be observed under a much smaller load than that which will cause the posts to buckle and the system to fail.
• Member connections and containment gusset plates should be made from the connecters that are as small as reasonable.
• 2× wood wedges, used in pairs to allow for vertical adjustment, should start cupping (i.e., the wedge edges move upward as the post crushes the wedges and sole beneath) at about 50 percent of the load that will cause system failure. This should provide a “structural fuse” that will warn of an overload.
• The design strength of an LP shore using four 4 × 4inch posts is based on the perpendiculartograin compression design strength of No. 1 Douglas fir lumber. The LP design strength is 32k. Design strength = working load = design load.
• It is reasonable to expect that, on average, a properly constructed LP shore using No. 1 Douglas fir lumber will fail at three times the design strength.
• FEMA vertical shores have only minimal lateral stiffness and strength; therefore, if a structure is so badly damaged that it needs lateral support, then lateral shoring should be installed.

Vertical load testing devices. The initial tests were performed in 2001 at the NASA/ARC site using a 150ton bridge crane that had been part of the Outdoor Aeronautical Research Facility (OARF). Each shore was loaded with “free weights” consisting of a 38k slab placed on top of the shores and additional concrete blocks that weighed 25k a pair. The blocks were suspended in pairs and then lowered onto the 38k slab one pair at a time. As each pair of blocks came to rest on the 38k slab, its suspension chains became slack. The total possible load was 138k, and most of the LPs failed as the third pair of 25k blocks was being lowered (somewhere between 88k and 113k). Steel brackets restrained the 38k slab from moving laterally. Each laced post specimen was 12.5 feet high. The test setup was designated OARF1.

In November 2005, the OARF test setup was modified to more precisely measure the loading during the test and at failure. Four 50ton hydraulic rams supported on steel brackets were installed symmetrically under the slab and loading blocks. The rams would lift the load while the shore was placed under the slab; the load was applied to the shore by reducing the pressure on the rams (i.e., lowering the slab and blocks), and the change in pressure on the rams would accurately determine the load on the shore. As above, the total possible load was the 38k slab and four pairs of 12.5k blocks (138k total). This test setup was called OARF2 (photo 1).

(1) OARF2 test setup. (Photos courtesy of author.)

In May 2007, a new testing frame was fabricated at the NASA/ARC DART site using an existing rocket motor test stand. The stand was modified by adding a steel platform as a loading table at the ground level, supported by four 50ton hydraulic rams. A pair of steel channels were placed at each side of the stand to support two movable head beams for height adjustment. The ram pressure accurately determines the load on the shore specimen. This vertical load tester with a 280k capacity was designated 280kVT (photo 2).

(2) Vertical load tester (280kVLT) test setup.

LP and PLP construction.

• All were between 12.2 and 12.5 feet tall.
• All 4 × 4inch posts were visually graded unseasoned Number 1 or better grade Douglas fir. They were spaced four feet out to out (i.e., outside to outside).
• All lumber was purchased from a local lumberyard. The 4 × 4 posts were chosen for having the fewest knots and the straightest grain. The 4 × 4 headers and soles were cut from the remaining supply. 2× material was No. 1 or better grade Douglas fir.
• Until 2009, only ¾ inch CDX grade plywood was used. It is noted below when thinner plywood was used for tests conducted in 2009 and 2010.
• Instructors built the shores prior to StS2 training. In some cases, significant rain occurred between the time the shores were constructed and the time of their testing.
• The shores used hammer and gundriven nails. All gundriven nails had full heads but were slightly offcenter. The 16d nails were vinylcoated coolers (0.148 × 3.25 inch). The 8d nails were also vinylcoated (.131 × 2.375 inch).
• The design load for all LPs and PLPs is 32k.
  Table 1 shows the results of 13 tests of LP shores using 2 × 4inch and 2 × 6inch lacing (photos 3-5).

(3) The typical conditions at failure. Test specimen LP31’s rightfront post split at a knot near the intersection of the upper midbrace and two diagonals.

(4) At the upper right, the rightfront post has crushed the header to about half its thickness, which displaced the thenstandard 12 × 12 inch full gusset. The standard now recommends using 6 × 12 inch halfgussets, which are installed more quickly and allow a better view of the crushing of the header.

(5) The rightfront post at failure load (3.2 times the design load) crushed the wedges and caused them to cup. Crushing and cupping are quite obvious at about two times the design load and indicate that the shore is overloaded and personnel must take action.

PLP shores with 4 × 4foot layout. The StS2 testing program considered whether plywood strips could replace the 2 × 4inch lacing. Based on ease of construction, it was decided initially to use 24 and 12inchwide strips within the shores’ height and to tie them together at the top and bottom with 12inch strips. Initially, the tests used ¾inchthick plywood; later tests used 5⁄8inch and ½inch plywood. These systems have the posts spaced four feet out to out, can be more rapidly constructed than 2× LP systems, and weigh less. The tests demonstrated that the 12inchwide plywood strips were inadequate to brace the shores; shores laced this way failed in global buckling. Table 2 summarizes the results of six 4 × 4 foot LP tests using plywood (photos 6-8).

(6) Typical failure conditions. LP32’s two rear posts fractured at knots near the top of the upper plywood lacing, and the rightfront post crushed the header to about half its thickness, displacing the gusset.

(7)The plywood lacing deformed as the shore failed. However, note that the face grain of the plywood is running in the short direction instead of the long direction. The plywood’s misorientation was the shore builders’ error. The plywood lacing performed better in other tests when the face grain was aligned with the long direction (i.e., from post to post).

(8) The post has crushed the wedges and caused them to cup. Note how the plywood gusset has started to buckle as the sole is crushed. Placing the gussets about ¾ inch or so below the top of the header and above the bottom of the sole prevents this. If the shore’s actual load is close to the design load (as it should be), crushing would not occur.

PLP with 2 × 4 foot layout tests. Following the success of the 4 × 4 foot PLP testing, testers connected a pair of double Ts (with posts two feet out to out), spaced four feet out to out, to construct a 2 × 4 foot PLP. These shores are even lighter than 4 × 4 foot PLP and, because of their twofoot dimensions, are easier to prefabricate and carry into a damaged structure. This testing was intended to define the most efficient PLP and propose it as a new FEMA US&R standard shore.

The initial tests in 2005 and 2006 demonstrated that the sides of the shore with twofoot spaced posts would be more difficult to brace, and all of the tests ended with a buckling failure. In 2008, a 96 inchhigh piece of plywood was placed at midheight on the twofoot sides, and the shore behaved like the standard LP. Following 2008, since the 96inchhigh plywood piece was considered undesirable from an access point of view, other plywood configurations were tried on the twofoot sides while still using two 24inchhigh strips on the 48 inchwide sides.

In 2010, it was reasoned that the preferred configuration would be to place the 24inchhigh strips closer to the top and bottom of the shore, since the potential buckling curvature is larger near the shore’s ends. The PLP84 and 85 tests were successful and also demonstrated that 5⁄8inch plywood could be used. Table 3 summarizes the 18 PLP tests with 2 × 4 foot layouts (photos 9-10).

(9) PLP85, the final version of the 2 × 4foot PLP at failure. The posts fractured at knots, and the headers and soles gave ample warning of failure as they were crushed and split.

(10) PLP85 after failure, showing the extent of the crushing of the sole and the cupping of the wedges. The failure load was 140k (4.4 times design load), which is attributed to the uniform straightness and absence of knots in all four posts.

Overall LP and PLP findings. A total of 37 LP tests were performed as a part of FEMA/USACE StS training. The testing not only proved the shores viability and the appropriate design capacities but also allowed the StS students to observe their behavior. The tests demonstrated that, when properly configured, the LP, 4 × 4foot PLP, and 2 × 4foot PLP will indicate overloading at levels well below failure. The cupping of wedges and deformation (crushing and splitting) of soles and headers are “structural fuse” indicators, which were repeated in most all of the tests.

In addition, the tests will lead to the likely adoption of the PLP and the use of thinner plywood and oriented strand board (OSB) for connections in FEMA shores.

LATERAL LOADSUPPORTING SYSTEMS TESTS

Raker shore testing. All raker shore testing was conducted at the Menlo Park (CA) Fire District/FEMA US&R CATF3 training site. The initial test in October 1999 demonstrated the capability of the newly constructed raker testing device. Since September 2004, this facility has conducted raker testing as part of advanced structure specialist training (StS2). A total of 31 tests have been performed through 2010.

The lateral load testing device (“Raker Breaker”) consisted of a 1.5 foot thick, 20footsquare base slab; a ramactuated steel frame; and a 12 foot square tilting wall, which is hinged at its base. A single, 30ton hydraulic ram moves a single steel strut attached to the center of a torsion beam. The torsion beam is attached to a triangular steel frame at each end with a large steel pin with bearings; this allows for rotation. Two Lshaped steel struts were connected to the top of the torsion beam; when the torsion beam was rotated, the struts exerted a force against the tilting wall in two places. The load applied to the back of the wall by the Lshaped struts is at a height of about 10 feet above the hinge at the bottom of the tilting wall. The Raker Breaker is designed to exert a maximum force on the tilting wall of 24k (photo 11).

(11) Typical test setup for testing a pair of solid sole rakers.

(12) Raker pair after sole cleat flyoff; nails were reduced to six.

The raker shores to be tested were constructed in pairs, eight feet apart, with standard 2× lateral “X” bracing placed between them. All of the tested rakers were configured at an angle of 45°, with their insertion points (the point where the center of the raker intersects the wall plate) at eight feet above the bottom of their sole plate. Since the load applied to the back of the wall is 10 feet above the hinge, the force in the rakers is the ram force multiplied by a ratio of 10/8. Therefore, the total force on the pair of rakers is 1.25 times to force applied by the ram.

All rakers were attached to the tilting wall by lag screws placed through the wall plates and into a 6 × 6 inch wood sleeper. The two wood sleepers, spaced at eight feet on center, were bolted to the tilting wall. Large concrete blocks provided the sole anchorage.

Three types of raker shores were tested as braced pairs; the design strength for raker pairs is 5k.

• 19 pairs of solid sole rakers (including the initial test). Solid sole rakers are preferred, since they may be preconstructed as a complete triangle and carried to the damaged wall. Most of these rakers were constructed by students during Rescue Systems 2 training at the training site. All had an eightfoot insertion point (Table 4).
  
• 10 pairs of split sole rakers. This type of raker is normally used when rubble is on the ground or slab next to the damaged wall and the bottom of the shore must be sloped up to clear the rubble. These rakers were constructed by Rescue Systems 2 students as well as StS instructors (Table 5).
• Two pairs of pneumatic strut rakers. Each pair of struts came from different manufacturers (Rak24, Rak 34). StS instructors assembled these rakers just prior to the tests. Both had insertion points of between eight and nine feet.

Pneumatic strut raker testing. Pneumatic strut rakers from one manufacturer were tested as a braced pair in March 2005, and the system was loaded to 25k without any significant failures being observed. Initially, the system was assembled with a midbrace and lateral X bracing placed between the pair. The system is shipped with wood nailers connected to the struts with metal clamps. This allows for the nailed connections of the 2× wood bracing.

Since the test of the braced system resisted the 25k load without failure, the lateral bracing was removed and the system was reloaded to 25k. It appeared that the raker struts would begin to buckle if the load was increased much beyond 25k. This manufacturer’s strut raker system performed well enough to be used as a substitute for wood rakers. The braced system is shown in photos 13.

(13) Pneumatic raker system with lateral bracing.

Strut rakers from a second manufacturer were tested as a braced pair in November 2005. The system was assembled with a midbrace and lateral X bracing placed between the pair, and wood nailers were attached with metal clamps. The system was loaded to 12k, and the midbrace connections failed, causing the raker struts to buckle. The test was stopped at this point.

Following the test, the manufacturer stated that the midbrace connection would be redesigned. The company was subsequently sold to another company.

The redesigned raker system was tested in May 2011 during StS2 training and supported 25k without failure (photo 14). This strut raker system performed well enough to be used as a substitute for wood rakers.

(14) Test of redesigned pneumatic raker.

Wood raker shore testing. A total of 30 tests of raker shores were performed as a part of FEMA/USACE StS training. The testing not only proved the viability of the shores but also allowed StS trainees to observe their behavior. The tests demonstrated that FEMA raker shores, when adequately restrained, have a safety factor of at least 6. The large safety factor is justified since it is difficult to determine how much force the shores will need to resist during high winds and aftershocks.

The tests also demonstrated that, as rakers become highly loaded, the StS can observe deformations in the sole cleat (for solid sole raker) or the trough (for the split sole raker). These indications are not as dramatic as the “structural fuse” indicators for laced posts, but they do provide an observable warning of overload.

Test videos. Videos of the tests may be viewed in the Multimedia Section of www.disasterengineer.org. Also, PowerPoint® files summarizing the shoring tests are available in the Library Section of the Web site. Detailed information regarding the proper construction of these shores can be found in Sections 2 and 3 of the US&R Shoring Operations Guide and the US&R StS Field Operations Guide, also available in the Library Section. This site is supported by members of the FEMA US&R Structures Subgroup and Bracken Engineering, Tampa, Florida.

Author’s note: Leadership and individuals from the following organizations have provided invaluable support to this testing program: the NASA/Ames Disaster Assistance and Rescue Team; the Structural Engineers Association of Northern California; the Menlo Park (CA) Fire District, CATF3; and the DHS/FEMA US&R and USACE US&R Program, as well as structure specialist instructors.

● DAVID J. HAMMOND is a structural engineer and a member of California’s Urban Search and Rescue (US&R) Task Force 3 (CATF3). He has served on the Federal Emergency Management Agency (FEMA) US&R Advisory Committee and was chair of the Department of Homeland Security (DHS)/Federal Emergency Management Agency (FEMA) US&R Structures Subgroup. Hammond is a lead instructor for the FEMA/U.S. Army Corps of Engineers (USACE) Structural Specialists (StS) training as well as other FEMA US&R training courses.