Troubleshooting SGCC Laminated Glass Ball Drop Test Failures – ASTM F3007 Help
We’ve been running into a frustrating issue at our facility and are hoping someone in the community has dealt with this or can offer insight
.
We’re currently performing in-house ball drop impact testing on laminated glass according to ASTM F3007-13, as part of SGCC production testing requirements. This test uses a 2.3 kg (5 lb) steel ball dropped from 3.66 meters (12 feet) for Class A products.
The issue is that our in-house ball drop tests are consistently failing — either complete penetration or post-breakage retention failure. However, when we send the same laminated glass batch out for third-party testing to meet CPSC 16 CFR 1201, which uses a 100 lb pendulum bag impactor, it passes.
We realize these are two different tests — one for SGCC (ASTM F3007) and one for CPSC safety glazing — but we’re surprised that glass which passes the more severe-seeming CPSC bag test could fail the ball drop test.
Here’s how we’re performing our in-house SGCC test:
· 2.3 kg (5 lb), 83 mm smooth steel ball
· Drop height: 3.66 m (12 ft ± 5 in)
· Glass conditioned for 4+ hours at 24°C ± 5°C (75°F ± 9°F)
· Support frame built by SGCC purchased from them. Their recommended apparatus.
· Light clamping or no clamping, but unsure if that’s affecting results
· Testing the thinnest configuration from the batch
Things we’re questioning:
· Is our drop height or release inconsistent?
· Could our ball surface be worn or damaged?
· Is the clamping or support method introducing extra stress?
· Are environmental conditions during testing affecting results?
· Could EVA cure variation be impacting break behavior?
Looking for feedback:
Has anyone else seen this kind of inconsistency between passing third-party CPSC impact tests and failing in-house SGCC ball drop tests?
We’d appreciate any tips on:
· How to properly set up a compliant ASTM F3007 drop test
· How much clamping is too much
· Whether you’ve seen EVA-based lami act unpredictably in ball drop tests
· Whether test rig calibration or support surfaces have played a role