Structural Test of a Sliding Door: A Reference Frame That Flips the Verdict
A 2,400 mm wide × 2,400 mm tall sliding glass door assembly is installed in a test chamber for structural performance testing per ASTM E330. The design pressure is 2,400 Pa. The specification requires that deflection at design load not exceed L/175 of the member span, that the assembly withstand 1.5 × design load without permanent deformation or damage, and that the operable panel still operate after the test. At 2,400 Pa the technician reads 12.4 mm of lateral deflection at midspan of the vertical interlock mullion using a dial indicator on a magnetic base fixed to the chamber's steel reaction stand. The reading is taken 45 seconds after the load is reached. A later check shows the reaction stand's base moves 2.5 mm in the same direction as the specimen at that pressure. The load is released, and after 5 minutes the dial reads 0.9 mm, which the technician records as permanent set. Under full load a 2 mm gap opens at the interlock stile. The project team's summary reads 'passes: 12.4 mm < 13.7 mm'.
1. Is 12.4 mm the deflection to compare with L/175? Explain.
Read the worked answer
No. L/175 for a 2,400 mm span is 13.71 mm, and 12.4 mm is under it - but 12.4 mm is the displacement of the mullion relative to the reaction stand, not relative to a fixed datum.
The stand base moves 2.5 mm in the same direction as the specimen. The dial sees only the difference, so it under-reads: true deflection ≈ 12.4 + 2.5 = 14.9 mm. That is over 13.71 mm, and an apparent marginal pass becomes a fail. Had the stand moved the opposite way, the dial would have over-read and the correction would subtract. The sign is the whole argument, which is why the direction must be measured, not assumed.
Two related checks belong in the same breath. First, confirm what the span is - clear span between supports, not necessarily the nominal 2,400 mm width - because L/175 moves with it. Second, confirm the gauge is at the point of maximum deflection; on an interlock mullion that is midspan, but with an asymmetric load path it may not be.
The durable fix is a reference on an independent pier or a datum outside the load path, with a second gauge monitoring the stand to prove the datum is stable.
2. What procedural requirements does E330 impose that this test appears to have missed? Cover hold time, permanent set and operability.
Read the worked answer
Hold time and stabilization. A reading 45 seconds after the load is applied is premature for an assembly with gaskets, setting blocks and clearance take-up. Compression continues for minutes. The reading should be taken once successive readings differ by no more than the gauge resolution, with the load held steady - check the exact hold duration required by the edition of E330 your specification invokes rather than assuming one.
Permanent set. Five minutes of recovery is short, and the residual reading is only meaningful with the load fully removed and the gauge re-zeroed against the same datum used during loading. A 0.9 mm residual may be elastic recovery lag rather than permanent deformation. The standard specifies a recovery period and a method; follow it, or the number is not a permanent set measurement at all.
Operability. The specification requires the operable panel to work after the test. Cycle it, record the force and any binding, and document it. It is the most commonly skipped line on a structural test report.
Beyond that, the report needs the load steps applied, the 1.5 × design safety level and what was observed during it, gauge identification and calibration, and the method used to establish the deflection datum - the last being exactly the item that decided this test.
3. How should the 2 mm interlock gap under load be treated?
Read the worked answer
Not as an automatic E330 failure, and not as something to leave out of the report either.
E330 is a structural and deflection test. A 2 mm interlock opening under 2,400 Pa may be entirely within the intended engagement of the interlock and the design clearance. The questions that matter are:
- Does it close? If it closes fully on unloading, it is elastic movement and the structural criterion may be met. If a residual offset remains after the specified recovery period, that is permanent deformation and a different conversation.
- Does it compromise the air and water barrier? The interlock is a barrier detail, not just a structural one. In a normal mockup sequence, air leakage (E283) and water penetration (E1105/E331) follow the structural test precisely because the specimen has now been loaded. After a structural test, re-run those tests on the same, post-structural mockup. A gap that opens under 2,400 Pa and leaks under 300 Pa of water pressure is the finding that matters, and the structural test alone will never reveal it.
- Does the panel still operate? If the gap comes with binding or a change in operating force, that is a specification failure regardless of deflection.
Recommend recording the gap at each load step, re-running air and water on the post-structural specimen, and holding the pass/fail decision until both the corrected deflection and the post-structural leakage results are in hand.