Confined vs Unconfined Anchor Testing: Why Configuration Matters

What the configuration controls

When a hydraulic jack applies tensile load to an anchor, the reaction frame pushes down on the surrounding substrate. The position of those reaction points relative to the anchor determines which failure modes the test can detect.

In an unconfined configuration, the reaction frame spans beyond the perimeter of the theoretical concrete cone. This allows all failure modes to develop, including cone breakout, which is often the governing failure mode for anchors designed under AS 5216. The reaction points must sit outside the cone's circular face, which means the span increases with embedment depth. For chemical anchors, the critical edge distance is calculated as 1.5 times the effective embedment depth per AS 5216:2021.

In a confined configuration, the reaction frame bears close to the anchor. The compressive force applied by the test rig to the surrounding substrate restrains the concrete cone, preventing breakout from occurring. The test only evaluates the bond between the anchor and the substrate, not the concrete tensile capacity.

When to use confined testing

Confined testing is appropriate for:

• Reinforcing bar connections: designed under AS 3600, where cone breakout is inhibited in service by compressive struts from surrounding reinforcement
• Installation quality checks: on chemical anchors where the primary concern is bond integrity, not substrate capacity
• Space-constrained: locations where an unconfined setup cannot physically be deployed

However, AEFAC TN05 Volume 2 notes an important limitation: a confined setup may not detect gross errors in embedment depth, because the restraining mechanism prevents the cone from forming even when the anchor is too shallow.

When to use unconfined testing

Unconfined testing is required for:

• All ultimate tests: , confined configuration prevents cone breakout, which would produce non-representative capacity data
• Anchor-theory connections: designed under AS 5216, where the in-service loading condition allows the full cone breakout mechanism to develop
• Masonry substrates: , AEFAC TN05 Volume 4 recommends unconfined testing because confined configuration restrains the mortar joint failure mechanism
• Rock substrates: , unconfined testing is essential to detect piston pull-out and cone lift-out failure modes

The non-conservative error

The most common configuration error is specifying confined testing when the in-service condition is unconfined. The test restrains the cone, producing higher apparent capacity than the anchor will deliver when loaded by the actual structure. The anchor passes the test but may not sustain the design load in service.

This error is difficult to detect after the fact because the test results appear satisfactory. Only an engineer who understands the relationship between the design theory, the in-service loading geometry, and the test configuration will recognise the mismatch.

Practical guidance

• The test specifier, not the testing contractor, should determine the test configuration
• Document the design theory (anchor or reinforcing bar) and the corresponding configuration in the test specification
• When uncertain, default to unconfined, it is conservative and will detect all failure modes
• For masonry and rock, always use unconfined configuration

References

• AEFAC TN05 Volume 1, Guidelines for Site Testing of Anchors: General
• AEFAC TN05 Volume 2, Section 7.1: Test Configuration
• AEFAC TN05 Volume 4, Testing in Masonry
• AS 5216:2021, Design of post-installed and cast-in fastenings in concrete