Shake-Table Fixture Amplification During Qualification Testing

Case classification

Published experimental context. This case study is based on phenomena documented in shake-table testing literature, where fixture dynamics interact with the table and the test article.

Scenario description

A piece of equipment was qualified by shake-table testing. The equipment was bolted to a steel fixture plate that was bolted to the shake-table. The fixture was designed to be rigid, but it had a natural frequency of approximately 25 Hz due to its span and thickness. During testing at a seismic input with significant energy at 25 Hz, the fixture amplified the table motion at its natural frequency. The equipment experienced accelerations at 25 Hz that were higher than the required response spectrum specified, but the test was considered passed because the test response spectrum (TRS) at the table level enveloped the RRS.

Analysis of the failure

The fixture amplification was not identified during the test because the control accelerometer was on the shake-table, not on the fixture. The TRS was computed from the table acceleration, which did not show the amplification. The equipment, mounted on the fixture, experienced the amplified motion at 25 Hz. The equipment had a component with a natural frequency near 25 Hz that was excited by the fixture amplification, causing it to fail functionally. The failure was attributed to the equipment not meeting its qualification, when in fact the fixture had introduced an amplified input that was not in the original RRS.

Root causes

• The fixture was not rigid as assumed — it had a natural frequency within the test frequency range
• The control accelerometer was on the table, not on the fixture, so the fixture amplification was not detected
• The TRS was computed from the table acceleration and did not reflect the actual input to the equipment
• The fixture design did not account for the frequency content of the test input
• The equipment had a component sensitive to the fixture frequency, creating a coupled resonance

Lessons learned

• The test fixture must be verified to be rigid (natural frequency above the test frequency range) or its dynamics must be characterised and accounted for
• A control accelerometer should be placed on the fixture near the equipment mounting to detect fixture amplification
• The TRS should be computed at the equipment mounting point, not at the table level, if the fixture is not rigid
• Fixture design should consider the frequency content of the test input — a fixture that is rigid for one spectrum may not be rigid for another
• If fixture amplification is detected, the fixture should be stiffened or the test input modified to compensate

Remediation

The fixture was stiffened with additional ribs to raise its natural frequency above 50 Hz, well above the test frequency range. A control accelerometer was added to the fixture at the equipment mounting point. The test was re-run with the TRS computed at the equipment mounting point, confirming that the input to the equipment matched the RRS without amplification. The equipment component that had failed was re-tested and passed, confirming that the original failure was caused by the fixture amplification, not by inadequate equipment design.