Proof Testing Sample Selection for a Mixed Batch
A project has 300 post-installed anchors of two types: 200 adhesive anchors in the interior and 100 mechanical anchors along the perimeter. The specification requires proof testing of 5% of anchors, selected by the engineer. The contractor proposes to test 15 anchors, all from the interior adhesive group, because they are "easier to access." The perimeter mechanical anchors are in a high-wind zone. The inspector must decide whether the proposed sample is acceptable and how to document the test results.
1. Is the contractor's proposed sample acceptable? Why or why not?
Read the worked answer
No. A 5% sample (15 of 300) is numerically correct but not representative. The sample must reflect the population and the risk profile.
Reasoning:
- The anchors are two distinct types with different failure modes (adhesive bond vs mechanical expansion).
- The perimeter mechanical anchors are in a high-wind zone - a higher-consequence condition.
- Testing only interior adhesive anchors leaves the highest-risk group untested.
The inspector should reject the sample and require the engineer to select a sample that includes both anchor types and the perimeter high-wind group, with proportional or risk-weighted coverage.
2. What should the inspector document for each proof test, and what happens if one fails?
Read the worked answer
Documentation per test should include:
- Anchor ID/location and type.
- Test load applied and duration.
- Displacement readings (if required).
- Pass/fail against the specified proof load.
- Date, tester, and equipment calibration reference.
If one fails:
- Do not simply retest that anchor and move on. A failure indicates a possible systematic issue (installation method, cleaning, embedment, or concrete).
- The inspector should escalate to the engineer, who typically requires expanding the test sample or investigating the cause.
- Document the failure and the corrective action; do not conceal or reclassify it.
The inspector's role is to ensure the sample and results are honest and representative, not to minimize testing.