Figure 1: Illustration of the apparent definition of the deflection limits. ASTM E985 includes post deflection, while AC273 and ASTM D7032 do not include post deflection.

Building codes require a guard along most elevated, open-sided walking surfaces. Analysis is one method of proving code compliance of new or existing guards, but there are challenges, as discussed in Part 1 (Civil + Structural Engineer, October 2014, page 59). Testing is an alternate method that may demonstrate code compliance not shown by analysis.

This two-part article focuses on wood-framed exterior decks typically found on residential buildings (e.g., single family homes, townhouses, condominiums, etc.). Part 1 provided a review of the building code requirements related to guards, commonly found non-code-compliant guard post details, corresponding code-compliant guard post details proven by analysis, as well as other guard design considerations. Part 2 provides a review of the codes and standards that define testing procedures, loads, and limits for code compliance, and a discussion of in-situ field testing.

Figure 2: Guard rail testing with load applied horizontally “outward” at the midspan of the top rail.

Figure 3: Guard rail testing with load applied “vertically downward” at the top rail connection.

IBC 2012 basis of testing The code requirements referenced in this article are based on the 2012 International Building Code (IBC); the information is generally applicable to the International Residential Code as well.

The IBC recognizes testing as a method of obtaining code compliance in Chapter 17 – Special Inspections and Tests, Sections 1709 InSitu Load Tests and 1710 Preconstruction Load Tests. Both sections require testing to the appropriate “referenced standards,” and if no such standard exists, the test procedure must be developed by a “registered design professional.” Both sections provide minimum test cr

Test standards and acceptance criteria There are several test standards with acceptance criteria published by American Society for Testing and Materials (ASTM) International and code agencies (AC/standard) related to guard code compliance, including the following:

  1. Where h is the rail height and l is the effective rail length (e.g., distance between the edges of the posts).
  2. AC273: “…the deflection at the midspan of the top rail (guard) is measured relative to the center of the two posts (i.e. it does not include post deflection).” ASTM D7032 similar.
  3. Where h is the distance from the top of the top rail to the first point of fastener connection to the supporting construction.
  • ASTM E935 — Standard Test Methods for Performance of Permanent Metal Railing Systems and Rails for Buildings
  • ASTM E985 — Standard Specification for Permanent Metal Railing Systems and Rails for Buildings
  • ASTM D7032 — Standard Specification for Establishing Performance Ratings for Wood-Plastic Composite Deck Boards and Guardrail Systems (Guards or Handrails)
  • ASTM E894 — Standard Test Method for Anchorage of Permanent Metal Railing Systems and Rails for Buildings
  • AC174 — Acceptance Criteria for Deck Board Span Ratings and Guardrail Systems (Guards and Handrails) — This criterion is for “guardrail systems (guard and handrails)… of any shape and thickness… manufactured from materials not prescribed by the applicable code.”
  • AC273 — Acceptance Criteria for Handrails and Guards — This criterion “is limited to handrails and guards… produced from metal or wood.”

Table 1 provides a summary of the key requirements of the AC/ standards; refer to the AC/standards for other important requirements not included in the summary.

The AC/standards have similar core elements, but varying load factors, loading direction, and deflection limits. It is the registered design professional’s responsibility to determine whether the testing requirements meet the building code.

The IBC requires the test loads to be 2x (in-situ) and 2.5x (preconstruction) the unfactored design loads. Three of the listed AC/ standards above require 2.5x, one requires zero, and one does not specify. The IBC (by reference to ASCE-7) requires that the guard resist design loads applied at any point and any direction. However, the AC/standards typically specify testing in one or two directions.

Figure 4: Guard rail testing with load applied horizontally “outward” at the top rail connection

Figure 5: Guard rail post testing with load applied “outward” at the top of the post.

Four of the AC/standards have similar deflection limits of h/24+l/96 (top rail loading) and h/12 (post loading). However, the criterion of top rail deflection measurement varies. Design professionals could reasonably interpret the h/24+l/96 top rail midspan limit to include the post and top rail deflections since the h/24 term accounts for the top rail height and the end user will “feel” the cumulative deflection.

However, AC273 and ASTM D7032 include the language in Table 1 Footnote 2, apparently to exclude the post deflection (Figure 1). Registered design professionals who are utilizing existing guard testing AC/standards or designing a test for a specific application (possibly based on an existing AC/standard) must pay close attention to the deflection limits that can cause guards to fail testing.

Additionally, the AC/standards are not always clear on what constitutes a non-deflection failure. For example, does a top rail that is permanently bent after application of the maximum load (e.g., 500-pound concentrated load) constitute a guard failure if the components do not break? We suggest that it does not because the top rail has performed its intended purpose of keeping the occupant alive

Field testing Once the decision is made to utilize testing to show compliance of an existing guard system, we recommend that the registered design professional identify the most appropriate AC/standard for the testing or use as the basis for a designed test.

Whichever standard is used, registered design professionals must consider how to construct the loading apparatus. The apparatus must be designed and anchored to resist the test loads, minimizing deflection that could affect results or damage force-resisting components (e.g., existing building). For in-situ testing on existing structures, anchorage may prove difficult with minimal disturbance of surrounding building components (Figures 2 through 5).

Consider the number of tests needed to account for the different guard configurations on the project and to address statistical significance. Also, consider testing the “worst case” configurations that will permit approval of more robust configurations based on engineering judgment/ analysis. If a project has too many in-situ configurations, testing may not be a viable option.

Additionally, consider the time required to perform each test. Even though the actual load test cycle is relatively short, it can take hours to set up/break down the test apparatus at a location. Depending on the guard configuration, a test apparatus that can apply force by pushing and pulling may reduce the number of times the apparatus must be set up and broken down.

Summary We have evaluated many questionably code-compliant guard systems. The system components are often code-compliant, but the critical post-to-structure connection often does not appear to meet code by analysis. Testing can be a reasonable approach to demonstrate code compliance where analysis does not. If testing is considered, the parties involved must understand the costs and risks, and the registered design professionals must correctly identify and apply a test that demonstrates code compliance.


SCOTT A. TOMLINSON, P.E., and ERIK W. FARRINGTON, P.E., are senior project managers at Simpson Gumpertz & Heger Inc. Reach them at satomlinson@sgh.com and ewfarrington@sgh.com. MAKOTO S. WEINSTEIN, EIT, is a staff I at SGH. Reach him at msweinstein@sgh.com.

X