Ok, I skipped most of the posts in this thread after the first page or so as my head was about to explode.
Here's several key points in fairly simple terms to consider when we're talking about cave fills:
1) 3AA steel tanks and 3AL aluminum tanks are all hydro tested to 5/3rds the service pressure (1.67 times the service pressure) during re-qualification. Exempt and Special Permit tanks are hydro tested to 1.5 times the service pressure.
2) The purpose of a hydro test is to measure the elastic expansion that occurs when the tank is pressurized to the test pressure and ensure that the elastic expansion is within specified limits (expressed as the Rejection Elastic Expansion, or REE in cubic centimeters) and that there is also essentially no permanent expansion, known as "plastic deformation".
3) Re-qualification includes both a VIP and a hydro test. Most RINs do the VIP first as that's how it's outlined in the standards, and as it makes sense to take a look in the tank before you bother with the hydro test, drying process, etc. That's not an issue unless you're testing a 6351-T6 tank. I've seen 6351-T6 tanks that passed a visual eddy inspection prior to the hydro test, but then had a visible crack after the hydro test.
The moral here is that a dive shop should always do their own Eddy Current, Visual Plus or Optical Plus inspection on a 6351-T6 alloy tank when it comes back from hydro test. The DOT is confident that a crack will take an average of 7 years to propagate from the point of detection to catastrophic failure, and thus be detected by a re-qualification well before a catastrophic failure occurs. The lack of ANY catastrophic failure of a properly inspected 6351-T6 tank since the eddy current protocols are adopted supports that confidence. However, there have been a very few instances of non catastrophic failures (leaks) in properly inspected tanks and I attribute that to the performance of the hydro test after the eddy current test and the lack of a dive shop performing a Visual Eddy, Visual Plus or Optical Plus inspection on the tank before returning to service after it comes back from hydro.
4) Aluminum tanks have a fatigue life. In other words, each and every fill starts to shorten the life of the tank. Luxfer tests their tanks to 10,000 cycles, but fatigue life isn't a specific number, it's a fairly broad range. The more fills an AL tank has, the more it will fatigue. Overfilling an AL tank will significantly increase the fatigue that occurs and shorten the life of the tank as the fatigue that occurs is proportional to the additional stress. That's why NO ONE recommends overfilling an aluminum tank.
5) Steel tanks do not normally*** have a fatigue life. Carbon steel has an "elastic limit", which if exceeded will result in measurable permanent expansion/ plastic deformation of the tank, although that is still well below the limit where the tank will catastrophically fail.
However, steel tanks also have a "fatigue limit" below the elastic limit where no fatigue in the steel occurs and where a steel tank can (theoretically) withstand an infinite number of cycles without weakening and eventually failing.
Tanks made to the 3AA engineering standard have proven to be very conservatively engineered, and it appears a 3600 psi fill in a 2400 psi tank is below the fatigue limit for 3AA standard designed tanks, or at least so close to it that the fatigue life is still a very large number (thousands or tens of thousands) of cycles. That's why 3600 psi fills over the last 20-25 years in north FL have not resulted in 2400 psi steel tanks catastrophically failing, or even failing hydro tests.
6) ***Now, if the fatigue limit is exceeded on a steel tank, it will start to fatigue and the life of the tank will be limited, and at some point it will fail a hydro test.
This is what gives me pause regarding the "creep" in cave fills from 3600 psi, to 3800 psi and to then to the current 4000 psi you see in some shops. 4000 psi is the 5/3rds test pressure for a 2400 psi 3AA steel tank. 400 psi difference may not seem like much, but it's 10%, and that last 10% can be significant. It remains to be seen how well 3AA steel tanks stand up to this additional 10% overfill over the course of 10-20 years.
7) Exempt and Special permit tanks should not be over filled as the engineering standards they are made under are much less conservative.
Consider what happened with Worthington X series special permit tanks that were failing hydro testing by the traditional methods. Eventually Worthington, DOT and TC all agreed to new test criteria, where a "round out" procedure was used in a hydro test is to increase the tank pressure to 90% of the test pressure and perform a system check. In essence, the procedure rounds out any oval cross section in the tank and expands the galvanized zinc coating on the tank, and takes that into account by re-zeroing the expansion measurement before doing the full test at the full test pressure to measure the actual elastic expansion in the tank that occurs in the last 10% of the test pressure.
That was never an issue in 3AA galvanized steel tanks as the more conservative design the 3AA tanks didn't encounter the same permanent expansion issue due to the galvanized coating and the coating process. However, as most folks realize, special permit and exempt tanks are lighter than a similar size 3AA steel tank, and that's due to the thinner walls of the tank associated with the less conservative special permit design criteria. Overfilling an exempt or special permit tank by any significant percentage will most likely create a significant risk of exceeding the fatigue limit of the tank and begin shortening the life of the tank. The more the service pressure is exceeded, the more fatigue will occur and the shorter the tank life will be.
8) Work hardening is an artifact of plastic deformation and both work hardening and plastic deformation are the result of dislocations in the crystalline structure of the metal. However, provided the elastic limit isn't exceeded in a steel tank, work hardening won't occur.
9) Due to the absence of a fatigue life, the major killer of steel tanks is rust. A deep pit in a tank sidewall will create a stress riser that can cause the tank to fail at that point. A line of pits, such as might be formed at the edges of water pooling in a tank lying on it's side is the most common cause of a steel tank catastrophically failing. Large areas of rust that thin the tank significantly are also responsible for their fair share of failure. Pits in the bottom of the tank are less critical as the material is thicker there, but I'll still fail a steel tank that has a significant pit inside or out.
10) When aluminum tanks fail, they tend to fragment into a number of large and small fragments, creating very effective shrapnel patterns that kill and maim bystanders pretty effectively. In contrast, steel tanks tend to split when they fail, peeling open, but staying in one piece with minimal fragmentation and shrapnel. Of course the blast is still just as damaging and just as deadly, so it's still a bad day for all involved, but given a choice I'd rather be near a steel tank that fails than an aluminum tank that fails.
11) Tumbling steel tanks to remove flash rust and flecks of rust is a normal and well accepted practice. What you are doing is removing any surface rust that might otherwise be hiding a pit. A small fleck of rust may not be serious if it's removed in time. However if it is allowed to sit there, it will continue to attract moisture and cause more rust underneath. Left long enough, it will cause a pit.
A few hours to over night tumbling is enough to remove pretty much any rust that's not already causing a problem. However, you could tumble a tank for weeks and not remove enough steel to reduce the strength by any measurable degree. The steel used in steel tanks is very hard and very tough and while tumbling media will remove rust effectively it won't remove intact steel very well at all.
That also means that you can tumble a tank for weeks and you still won't be able to remove a pit. You might clean out the darker material in the bottom of the pit and shine it up so it looks less like a pit, but the pit will still be there. I don't bother tumbling a steel tank more than overnight before I make a call on whether it's a fleck of rust or an underlying pit.
Here's several key points in fairly simple terms to consider when we're talking about cave fills:
1) 3AA steel tanks and 3AL aluminum tanks are all hydro tested to 5/3rds the service pressure (1.67 times the service pressure) during re-qualification. Exempt and Special Permit tanks are hydro tested to 1.5 times the service pressure.
2) The purpose of a hydro test is to measure the elastic expansion that occurs when the tank is pressurized to the test pressure and ensure that the elastic expansion is within specified limits (expressed as the Rejection Elastic Expansion, or REE in cubic centimeters) and that there is also essentially no permanent expansion, known as "plastic deformation".
3) Re-qualification includes both a VIP and a hydro test. Most RINs do the VIP first as that's how it's outlined in the standards, and as it makes sense to take a look in the tank before you bother with the hydro test, drying process, etc. That's not an issue unless you're testing a 6351-T6 tank. I've seen 6351-T6 tanks that passed a visual eddy inspection prior to the hydro test, but then had a visible crack after the hydro test.
The moral here is that a dive shop should always do their own Eddy Current, Visual Plus or Optical Plus inspection on a 6351-T6 alloy tank when it comes back from hydro test. The DOT is confident that a crack will take an average of 7 years to propagate from the point of detection to catastrophic failure, and thus be detected by a re-qualification well before a catastrophic failure occurs. The lack of ANY catastrophic failure of a properly inspected 6351-T6 tank since the eddy current protocols are adopted supports that confidence. However, there have been a very few instances of non catastrophic failures (leaks) in properly inspected tanks and I attribute that to the performance of the hydro test after the eddy current test and the lack of a dive shop performing a Visual Eddy, Visual Plus or Optical Plus inspection on the tank before returning to service after it comes back from hydro.
4) Aluminum tanks have a fatigue life. In other words, each and every fill starts to shorten the life of the tank. Luxfer tests their tanks to 10,000 cycles, but fatigue life isn't a specific number, it's a fairly broad range. The more fills an AL tank has, the more it will fatigue. Overfilling an AL tank will significantly increase the fatigue that occurs and shorten the life of the tank as the fatigue that occurs is proportional to the additional stress. That's why NO ONE recommends overfilling an aluminum tank.
5) Steel tanks do not normally*** have a fatigue life. Carbon steel has an "elastic limit", which if exceeded will result in measurable permanent expansion/ plastic deformation of the tank, although that is still well below the limit where the tank will catastrophically fail.
However, steel tanks also have a "fatigue limit" below the elastic limit where no fatigue in the steel occurs and where a steel tank can (theoretically) withstand an infinite number of cycles without weakening and eventually failing.
Tanks made to the 3AA engineering standard have proven to be very conservatively engineered, and it appears a 3600 psi fill in a 2400 psi tank is below the fatigue limit for 3AA standard designed tanks, or at least so close to it that the fatigue life is still a very large number (thousands or tens of thousands) of cycles. That's why 3600 psi fills over the last 20-25 years in north FL have not resulted in 2400 psi steel tanks catastrophically failing, or even failing hydro tests.
6) ***Now, if the fatigue limit is exceeded on a steel tank, it will start to fatigue and the life of the tank will be limited, and at some point it will fail a hydro test.
This is what gives me pause regarding the "creep" in cave fills from 3600 psi, to 3800 psi and to then to the current 4000 psi you see in some shops. 4000 psi is the 5/3rds test pressure for a 2400 psi 3AA steel tank. 400 psi difference may not seem like much, but it's 10%, and that last 10% can be significant. It remains to be seen how well 3AA steel tanks stand up to this additional 10% overfill over the course of 10-20 years.
7) Exempt and Special permit tanks should not be over filled as the engineering standards they are made under are much less conservative.
Consider what happened with Worthington X series special permit tanks that were failing hydro testing by the traditional methods. Eventually Worthington, DOT and TC all agreed to new test criteria, where a "round out" procedure was used in a hydro test is to increase the tank pressure to 90% of the test pressure and perform a system check. In essence, the procedure rounds out any oval cross section in the tank and expands the galvanized zinc coating on the tank, and takes that into account by re-zeroing the expansion measurement before doing the full test at the full test pressure to measure the actual elastic expansion in the tank that occurs in the last 10% of the test pressure.
That was never an issue in 3AA galvanized steel tanks as the more conservative design the 3AA tanks didn't encounter the same permanent expansion issue due to the galvanized coating and the coating process. However, as most folks realize, special permit and exempt tanks are lighter than a similar size 3AA steel tank, and that's due to the thinner walls of the tank associated with the less conservative special permit design criteria. Overfilling an exempt or special permit tank by any significant percentage will most likely create a significant risk of exceeding the fatigue limit of the tank and begin shortening the life of the tank. The more the service pressure is exceeded, the more fatigue will occur and the shorter the tank life will be.
8) Work hardening is an artifact of plastic deformation and both work hardening and plastic deformation are the result of dislocations in the crystalline structure of the metal. However, provided the elastic limit isn't exceeded in a steel tank, work hardening won't occur.
9) Due to the absence of a fatigue life, the major killer of steel tanks is rust. A deep pit in a tank sidewall will create a stress riser that can cause the tank to fail at that point. A line of pits, such as might be formed at the edges of water pooling in a tank lying on it's side is the most common cause of a steel tank catastrophically failing. Large areas of rust that thin the tank significantly are also responsible for their fair share of failure. Pits in the bottom of the tank are less critical as the material is thicker there, but I'll still fail a steel tank that has a significant pit inside or out.
10) When aluminum tanks fail, they tend to fragment into a number of large and small fragments, creating very effective shrapnel patterns that kill and maim bystanders pretty effectively. In contrast, steel tanks tend to split when they fail, peeling open, but staying in one piece with minimal fragmentation and shrapnel. Of course the blast is still just as damaging and just as deadly, so it's still a bad day for all involved, but given a choice I'd rather be near a steel tank that fails than an aluminum tank that fails.
11) Tumbling steel tanks to remove flash rust and flecks of rust is a normal and well accepted practice. What you are doing is removing any surface rust that might otherwise be hiding a pit. A small fleck of rust may not be serious if it's removed in time. However if it is allowed to sit there, it will continue to attract moisture and cause more rust underneath. Left long enough, it will cause a pit.
A few hours to over night tumbling is enough to remove pretty much any rust that's not already causing a problem. However, you could tumble a tank for weeks and not remove enough steel to reduce the strength by any measurable degree. The steel used in steel tanks is very hard and very tough and while tumbling media will remove rust effectively it won't remove intact steel very well at all.
That also means that you can tumble a tank for weeks and you still won't be able to remove a pit. You might clean out the darker material in the bottom of the pit and shine it up so it looks less like a pit, but the pit will still be there. I don't bother tumbling a steel tank more than overnight before I make a call on whether it's a fleck of rust or an underlying pit.
