Deck Railings Guardrail Connections

Tested Guardrail Post Connections for Residential Decks

Joseph Loferski, Dustin Albright, Frank Woeste, Ph.D., P.E.

Guardrails are intended to prevent occupants from falling off a deck preventing serious injury or even fatality if from sufficient elevation. This article describes test results from a research program conducted to evaluate guardrail post-to-deck connections for residential applications. Testing was limited to residential guardrails with a 36-inch height, and do not apply to guardrails required by code to be at least 42 inches tall (IBC, Section 1012.2). It should be noted that reference to a test assembly as being “code-conforming” refers to the fact that the test procedure and results intend to satisfy provisions of the current model building codes.

The most recent IRC codes (International Code Council (ICC), 2000 and 2003a) specify in Tables R301.4 and R301.5, respectively, a minimum concentrated live load of 200 pounds for both guardrails and handrails. Footnote “d” to International Residential Code (IRC) Table R301.4 and Table 301.5 (ICC, 2000 and 2003a) defines application of the 200 lbs load: “A single concentrated load applied in any direction at any point along the top.” The evident question is “how can the code requirement be accomplished?”

A guardrail is a system of components connected together and fastened to the deck. The system includes the posts, rails, pickets, and connections between the various parts and deck framing. While there are many ways to design and construct a guardrail, including continuing column support as guardrail posts, this article focuses on guardrail posts connected to the band joists of the deck. Figure 1 shows two load cases that must be safely resisted by the post and the post connection to the deck structure. It shows a vertical load that might be caused by a person sitting on the top rail. A horizontal load could be caused by people leaning against the top rail, or by someone falling into the guardrail assembly. In both cases, the force must be transferred from the top rail to the post, and from the post to the connection of the post to the deck substructure.

Figure 1: The IRC (ICC, 2000 and 2003) require “guardrails and handrails” to be designed to support a minimum single concentrated load of 200 pounds “…in any direction at any point along the top.

The vertical load depicted in Figure 1 produces shear forces in the post-to-deck connection, and this force is relatively easy to resist by 1/2-inch lag screw or bolted connections. The horizontal load, however, is very difficult to accommodate in design because of the lever arm effect. The 200-pound load applied at the top of the post can easily produce a couple thousand pounds of load at the base of the post that must be carried by the connection.

For the tests described in this article, a horizontal load was applied to the top of the post in the outward and horizontal direction. The load was applied in the horizontal direction because it produces the maximum bending moment and resulting forces in the connection below. In other words, the “worst case scenario” was tested in anticipation that a person could accidentally fall into the guardrail producing a resultant force that is perpendicular to the post. Based on the code language, it is impossible to prove a detail by test when “a single concentrated load” is “applied in any direction” as any direction could mean any angle (0 to 360 degrees). The use of the word “any” includes the cases where the load is applied inward, towards the center of the deck floor, hereafter referred to as the “inward load case.” An example of inward loading on a guardrail (or guardrail post) might be the unlikely case that a tree would fall against the guardrail, collapsing the railing, and continue to fall into the residential unit. The inward loading case was not evaluated in this testing program, and this limitation will be discussed in the conclusions to this article.

For each specimen tested, the maximum load each connection configuration could carry before it failed was measured.The deflection of the post at 36 inches above the simulated decking surface was also measured, even though the building codes do not specify a deflection limit for this specific application. Based on observing numerous tests, the outward movement of the top of the post under load without breaking (ductility) is considered a positive attribute because it is likely to give warning to occupants that they may be overloading the assembly. Deflection limits are not believed to be necessary for traditional guardrail systems that are totally solid wood; however, it should be noted that the deflection of plastic and WPC guardrail systems is limited by provisions of AC174 (ICC Evaluation Service, Inc., 2005).

Guardrail Post-to-Deck Test Assembly

Requirements set by the IRC (ICC, 2000 and 2003a) were used to define the basic geometry for the test program. The railing was set at 36 inches above the deck surface in conformance with the minimum height specified by the IRC and that the deck boards in an actual application are at most 1.5 inches thick. Thus, the horizontal test load was applied to the post 37.5 inches above the top of the simulated deck joists.

Target Test Load for Passing the Test

Because the building code minimum design load requirement is 200 pounds, a code conforming post connection design, when tested in a laboratory, must be able to carry 200 pounds times an appropriate “safety factor.” The safety factor of a design is intended to “protect” the stated design requirement when realizing that actual field installations are not perfect as constructed in a laboratory test, that the connection can degrade in-service due to repeated loads and weathering, and that some in-service loads may exceed the design load. The safety factor in this application is not intended to account for (biological) decay of the wooden components, and if decay is detected in a post-to-deck connection assembly, the element should be immediately replaced with new PPT materials.

The deck post, joists, and band joists test assembly was considered to be a “proposed construction” that is “not capable of being designed by approved engineering analysis” as described in Section 1712.1 of the IBC (ICC, 2003b). Per IBC Section 1712.1, the “structural unit and connections” shall be tested as indicated in Section 1714. Section 1714 utilizes a safety factor of 2.5, a number that has been in the model codes for decades for testing structural assemblies. Therefore, in this testing program, a guardrail post-to-deck connection assembly was considered to be “code conforming” if the average test load for a design exceeded 500 pounds.


For this testing program, we used PPT (ACQ or CA-B) 2×8 Southern Pine simulated deck and band joists and 4×4 No. 2 Southern Pine posts. Some of the tests included a PPT 5/4×6 radius-edge-deck board attached to the joists and band joist. The lumber, both joists and posts, were purchased and kept “wet” before the assembly test so that a connection adjustment factor for “wet-use” would not have to be applied to the test data. Deck guardrail assemblies probably cycle at least once between wet (moisture content, MC, greater than 19%) and dry (MC less than 19%) conditions in-service. Therefore, these tests were conducted with wet PPT lumber.

The research program began with tests of commonly used post-to-deck connection configurations in Southwestern Virginia, including bolts, lag screws, notched and un-notched posts, and blocking attached in various configurations between the joists and band joist. None of these designs reached the target test load of 500 pounds needed to claim a code conforming connection assembly.

Figure 2: This guardrail post detail “looks” very strong as it utilizes two ½x6-inch bolts; yet, in a series of tests, the average failure load was only 47% of the test load requirement for a code conforming design.

Tests and Failure Modes

Five specimens of each configuration were tested. Two details were tested using 1/2-inch lag screws or 1/2-inch bolts as depicted in Figure 2. Both details relied on 2 inch decking screws and three 3-inch stainless steel screws to connect the band joist to the joists. Southern pine 5/4 deck boards were included in the tests to be representative of what contractors may be using in the field; however, it should be noted that relying on decking screws to stabilize a rail post connected to a band joist is an unreliable practice, as the deck board may be replaced with deck boards and fasteners significantly weaker than what was tested.

Table 1: Summary of guardrail post-to-deck-connection testing results for four residential rail-post-assemblies involving PPT 2×8 No. 2 Southern Pine joists and 4×4 No. 2 Southern Pine posts. (The average test load for an assembly as a percent of the 500 lbs. test load requirement must be greater than 100% to be considered a “test proven assembly” by the authority having jurisdiction for an actual construction.)

1/2-inch Lag Screws or Bolts Did Not Pass the Test

The lag screw connection detail failed by withdrawal of the threaded portion from the band joist at an average ultimate load of 178 pounds. The first row of Table 1 gives a summary of test results for the lag screw case. Deflection at the design load of 200 pounds was not recorded and reported because the assemblies, on average, failed before reaching 200 pounds. The average test load was only 35% of the requirement for a load tested assembly.

For the bolted deck rail post assembly shown in Figure 2, the connections failed at an average load of 237 pounds – barely surpassing the code required design load with almost no safety factor for the service life of the assembly. Referring to the second line of Table 1, the top of the post deflected, on average, 4.4 inches at 200 pounds of applied load. The average test load was only 47% of the test load requirement for a code conforming design (Figure 2). The bolted connections typically failed when the band joist “peeled” away from the deck joists as the screws that attached the band to the joists pulled out, or failed in withdrawal mode.

Other Designs Failed to Pass the Test

A variety of designs involving the use of 2×8 southern pine blocking that were “lag screwed” to the joists, and had bolts installed through the band, the post, and the blocking, were tested over a three month period. None of these designs met the required test load because the lumber components typically failed in tension perpendicular-to-grain due to extreme bearing stresses produced by the fasteners.

Some configurations included notched 4×4 posts attached with bolts to the band joist. Although none of the 4×4 laboratory test posts failed at the notch, notching reduces the strength of the post significantly; guard-rail posts should not be notched. Repeated in-service moisture cycles typically cause cracks to develop and propagate from the corner of the notch along the slope-of-grain. Thus, notches in a guardrail post, coupled with “grade permitted” slope-of-grain and moisture cycles, can produce a very weak guardrail post during the service life of the deck.

A Detail that Passed the Tests

As the testing program progressed, failure modes were analyzed to develop a better understanding of the high level of forces involved in the connection at the base of the post. Because all specimens failed by withdrawal of screws between the band and the joists, splitting of wood perpendicular-to-grain, screw head pull-through, or bending of the screws, it was realized that a successful design had to utilize bolts that are arranged in a way that they can transfer load from the post to the joist in shear (lateral loading) because bolted connections are very strong when resisting lateral, or shearing type loads.

Figure 3: Simpson Strong-Tie™ HD2A connector used to test an assembly with the post inside band joist (shown) and outside band joist.

Therefore, a connection design was sought that loaded bolts in lateral or shear mode. A commercial steel connector (Figure 3) was identified that is typically used to resist wind and earthquake loads in shear walls, but could also be used to attach a guardrail post to a deck joist. In this deck application, these connectors utilize three ½-inch diameter bolts: two bolts are installed in the joist and are loaded in shear, and the third bolt passes through the post, the band, and the connector itself. The third bolt is loaded in tension. These designs are depicted in Figures 4 and 5. In each case, another 1/2-inch bolt is installed in the lower part of the post and the band joist.

Figure 4: Guardrail post test assembly details for 4×4 southern pine posts inside of band joist.

Only one Simpson StrongTie™ HD2A connector was used per post, and the centerline of the connector was positioned 2 inches below the top edge of the 2×8 joist. Maintaining this dimension in the field is extremely important because it severely impacts the forces involved in the connection. In no case should the centerline of the HD2A be more than 2 inches from the top of the joist. A hot dipped galvanized (HDG) HD2A connector was “special ordered” for these tests, and as a minimum, only the HDG version of the connector should be used in construction due to the corrosion potential with new lumber treatments.

Figure 5: Guardrail post test assembly details for 4×4 southern pine posts outside of band joist.

Two Details Tested – Post Inside Band and Post Outside Band

Two joint configurations with five replications of each were tested: one with the post located inside the band as shown in Figure 4, and the other with the post located outside the band joist as shown in Figure 5. The flexibility in locating the post inside or outside of the band may be desirable or required for hollow plastic or wood-plastic-composite guardrail posts that slide over a solid-sawn 4×4 post and rely on the 4×4 for structural integrity. During testing, at least 650 pounds was applied to the top of the post; every specimen successfully survived this load. The test results are presented in the last two lines of Table 1. Note that the average test loads achieved were in excess of the 500-pound test load requirement necessary for a code conforming tested assembly. In fact, specimens were not able to be tested to failure without potentially damaging test equipment.

Figure 6 depicts a typical maximum deflection during one test for the case of “post outside the band joist.” From Table 1, the average deflection at the 200-pound design load at 37.5 inches above the simulated deck joists was 2.0 inches and 1.9 inches for the “inside” and “outside” details, respectively. Different failure mechanisms were observed for the two cases as the load increased up to the maximum of about 650 pounds. When the post was mounted inside the band, the washers under the bolt head embedded into the wide face of the 2×8 band joist, indicating the role of the compression-perpendicular-to-grain strength of the band joist. When the post was located outside the band, the bolt head and washer pulled well into the 4×4 post, crushing the wood fibers beneath the washer. Thus, for the post-outside-band case, the compression-perpendicular-to-grain strength of the 4×4 post was a factor in achieving the 500-pound test load requirement. Due to the role of compression-perpendicular-to-grain lumber strength in the tests, results for Southern Pine (G=0.55) may not directly apply to other lower density species of guardrail posts used for deck construction in the United States.

Figure 6: Outside band joist specimen loaded with 650 pounds. The specimens exhibited extreme “ductility” and this test characteristic is considered a desirable safety feature.

Inward Load Case not Tested

In this testing program, only one HD2A connector was used per post and the centerline of the connector was positioned 2 inches below the top edge of the 2×8 joist. It is possible that the guardrail post could be heavily loaded in the inward load direction, as by a falling tree. The HD2A connector post-to-deck assembly in the inward loading mode was not tested, and based on judgment, the assembly would not carry 500 pounds inward. However, it is believed that the assembly would carry 500 pounds in either direction by installing two HD2A connectors per post, one 2 inches from the bottom of the 2×8 band joist and one 2 inches from the top of the 2×8 joist. The contractor, professional designer, and property owner should consider the option of using two HD2A connectors to achieve a 500 lbs test load capacity in both directions – outward and inward ?

Joseph Loferski and Dustin Albright are, respectively, professor and graduate research assistant in the Department of Wood Science and Forest Products, Virginia Tech.
Frank Woeste, Ph.D., P.E. is Professor Emeritus, Biological Systems Engineering Department, Virginia Tech, Blacksburg. Frank can be reached via his email at

This article is reprinted with permission from the Forest Products Society’s Wood Design Focus, Summer 2006.

This article is available in Adobe PDF format:

PDF Tested Guardrail Post Connections for Residential Decks

Protected by Copyscape Originality Check