There is no lubricant present at either stage. I believe the issue is that the torque check is not the same application as the assembly. The torque is nothing more than twisting of the stud. The application applies upward pull (away from the weld). Overall bad design. Could have went with a PEM...
I have a CD weld stud that requires a 3.1NM torque test. I torque the stud with out anything assembled and it passes the 3.1NM torque (weld holds)
When My team assembles the part and they attempt to torque the same stud with hardware the downward force causes the weld to fail and stud to fall...