A675 GR90 Anchor Bolt Fatigue Failure
A675 GR90 Anchor Bolt Fatigue Failure
(OP)
I did a tensile test on a total of 28 3-inch diameter anchor bolts (set 30 inches into concrete with a nut at the end, an projecting up 30 inches into a steel hold down seat) holding down two water reservoirs 15 feet in diameter and 96 feet tall. They were only tightened with a wrench with a 6 foot cheater bar when new in 1996. After the Nisqually earthquake in February 2001 in the Seattle area I wondered for some time about the bolts,(many of the nuts looked loose), well I decided to prestress them all to 100-kips (.25 fy and 3 of them were broken off 4" below the concrete surface before I started), and we prestressed the remaining 25. It was very difficult to do this because the nuts only clear the reservoir by 3/8 inch. It looks like recent fatigue failure to me, no necking down and all broke the exact same way. I sent them into a lab for a chemical, tensil, brittleness test, and an opinion of failure mode. I had thought the concrete anchorage seat was pulling out slightly or the bolts deformed somehow during the earthquake - wrong. I am now quite concerned about the remaining bolts and how I could repair/replace the broken ones, or if I should replace them all (huge job).





RE: A675 GR90 Anchor Bolt Fatigue Failure
Ultrasonic examination is a nondestructive test (NDT) method that utilizes high frequency sound waves to detect internal flaws in material. This is a highly specialized test and would require an experienced Level II examiner. Call an NDT company in your area and have them evaluate the option of UT. I have seen this performed in the field numerous times and it might be feasible for your application.
Regarding the failed anchor bolts - you should have a licensed structural engineer review the situation to determine if there is enough redundancy to avoid having to replace the failed anchors. I would not recommend any butt weld repairs to failed anchors.
RE: A675 GR90 Anchor Bolt Fatigue Failure
The bolts broke off 4 inches below the concrete surface in a break that looks like a classical fatigue failure, roughly strait across and you can see how it propagated.
For repairs we have considered chipping down and attach long coupler nuts to the remaining imbedded section with a 40 inch shaft extension, or try to unscrew the bolt from the imbedded nut with an ultra high torque impact wrench, but not sure if this is possible. Welding on a piece is not an option