Ultimate Load Testing: Statistical Method vs Simplified Method

When ultimate testing is required

Ultimate load testing determines the actual failure capacity of a post-installed anchor in a specific substrate by loading the anchor until it fails. Unlike proof testing, which verifies installation quality at a fraction of the design load, ultimate testing establishes the capacity of the anchor-substrate system from measured data.

Ultimate testing is required when:

• The substrate properties are unknown or variable (aged concrete, masonry, rock)
• The application falls outside the scope of the manufacturer's European Technical Assessment (ETA)
• No published design data exists for the specific anchor-substrate combination
• The engineer needs to establish a project-specific design capacity for a non-standard condition

The statistical method (minimum 15 tests)

BS 8539:2012+A1:2021 Annex B.2.3 defines the statistical method as requiring a minimum of 15 tests per substrate condition. The test results are analysed statistically to derive a characteristic resistance at the 5th percentile with 90% confidence.

This means that the derived capacity represents the value below which only 5% of the population would be expected to fall, providing a high level of confidence that the design capacity is achievable in practice.

The statistical method produces the most efficient (least conservative) design capacity because the larger sample size provides better statistical confidence. For large anchor installations where hundreds or thousands of production anchors will be installed, the cost of 15 additional tests is typically justified by the higher design capacity each anchor can be assigned.

The simplified method (5 tests)

The simplified method requires only 5 tests, but applies higher partial safety factors to the mean of the test results to compensate for the smaller sample size. This produces a more conservative design capacity than the statistical method.

The simplified method is appropriate when:

• The installation is small and the cost of 15 tests is disproportionate
• A conservative design capacity is acceptable
• Time constraints prevent a full statistical programme

Test configuration

Ultimate tests must be conducted in an unconfined configuration. In a confined setup, the reaction frame restrains the concrete cone breakout mechanism, preventing this failure mode from developing. Since the purpose of ultimate testing is to determine the actual capacity including all failure modes, confining the substrate would produce non-representative results.

Failure mode identification

The load-displacement curve recorded during each test provides diagnostic information that identifies the governing failure mode:

• Bond failure: produces gradual displacement with a load plateau, the adhesive debonds progressively from the substrate
• Cone breakout: produces sudden brittle displacement, the concrete tensile strength is exceeded and a cone of substrate separates
• Steel failure: produces yielding followed by fracture, the anchor rod reaches its tensile capacity

The failure mode tells the engineer whether the design is governed by the substrate, the bond, or the steel, which directly affects the anchor specification and installation methodology going forward.

References

• BS 8539:2012+A1:2021, Annex B.2.3
• AEFAC TN05 Volume 3, Guidelines for Site Testing of Anchors: Ultimate Tests
• VicRoads Section 680, Bonded Anchors