Making a burst disc might seem straightforward, but the precision required to manufacture one is complex.
The challenge lies in ensuring the disc consistently ruptures at the correct pressure, every time. Burst discs must break between 90% and 100% of their rated pressure, and to achieve that, manufacturers must ensure consistency across several factors. The alloy used must be homogeneous, every disc must be the exact same thickness, and the torque applied during installation must be precise to the manufacturer's specifications.
It's the last two factors—thickness and torque—that you altered, which change the pressure threshold at which the disc fails.
Over-torquing during installation is the most common reason for premature burst disc failure. Manufacturers specify exact torque requirements for good reason. If the disc is torqued too much, it gets compressed at the edges, which thins it. A thinner disc is less able to withstand the pressure it was designed for. On the flip side, if a disc has been removed and reinstalled, it should not be reused. Loosening and then re-torquing the assembly weakens the disc, because each time it's re-torqued, it thins slightly at the mating surface between the disc and the assembly nut.
While the disc might not rupture immediately, you've still altered its structural integrity. By how much? That's hard to say. What’s certain is that the disc is no longer as reliable as it should be.
The original burst disc should have been rated to withstand 5/3 of the working pressure of the cylinder, but it’s now "rated" lower—by an unknown amount. Does this justify replacing it? I believe it does, but that decision ultimately depends on your personal tolerance for risk and how you use the cylinder.
If you’re the only one filling and using the cylinder, you might accept the unknown factor. However, if someone else is filling it for you, the right choice would be to replace the
entire burst disc assembly.