tank fill rate

[bl6394]

Tobin / Rjack - let's try to bring this back to a common sense discussion. Something that others will find useful.

We recognize that many industry practices factor in the limited skill or knowledge of the operator (FSO, diver, etc). For example, we don't expect recreational divers to understand bubble formation or the Bühlmann decompression algorithm - instead we give them reasonable limits and things tend to work out well.

We agree that some cylinders have been rated by their manufacturer's for fast fills - if they have been properly maintained and inspected. We agree that it's hard / impossible for a FSO to tell when a cylinder has been improperly overfilled or to what degree and frequency. We also can't tell, without a VIS, if there is considerable rust / corrosion or other defects that may impact the strength of that cylinder.

We agree that the most stressful way to fill a cylinder will be rapidly - without allowing time for much of the heat generated from adiabatic compression to dissipate. If there is the possibility for a failure to occur - you have maximized the probability with your fill technique.

So as a rule - we apply prudent practices to allow most FSO's to have a happy ending. Slow filling will provide the greatest safety factor for the FSO and provide a better fill for the customer. That's about all that needs to be said for most FSO's.

However, if a metallurgist is in the audience - we'll enjoy your input as well!

 

[rjack321]

We agree that the most stressful way to fill a cylinder will be rapidly - without allowing time for much of the heat generated from adiabatic compression to dissipate. If there is the possibility for a failure to occur - you have maximized the probability with your fill technique.

You keep confusing the temperature of the gas with metallurgical damage. The specific heat of a gas is so ridiculously low compared to the specific heat of both steel and aluminum that the gas temp is not important in scuba fills - it just creates a short fill. There is no damage from that hot gas. The temp of the actual cylinder increasing from (e.g.) 80F to 110F is basically irrelevant

Hot gas is not what is stressing the cylinder during a fill (fast or slow). The speed at which a load is applied will inversely influence how ductile the material is. Higher loading rates reduce the ability of a material to bend gradually. While not entirely independent of temperature, for the ranges seen in scuba fills, it is.

Tanks should be filled (with air or premix nitrox) at a modest rate - your don't need to agonize over that, the tank is likely to outlive all of us regardless of fill rate.

O2 filled at as slow a rate as possible - do agonize over this.

 

[cool_hardware52]

You keep confusing the temperature of the gas with metallurgical damage. The specific heat of a gas is so ridiculously low compared to the specific heat of both steel and aluminum that the gas temp is not important in scuba fills - it just creates a short fill. There is no damage from that hot gas. The temp of the actual cylinder increasing from (e.g.) 80F to 110F is basically irrelevant

Hot gas is not what is stressing the cylinder during a fill (fast or slow). The speed at which a load is applied will inversely influence how ductile the material is. Higher loading rates reduce the ability of a material to bend gradually. While not entirely independent of temperature, for the ranges seen in scuba fills, it is.

Tanks should be filled (with air or premix nitrox) at a modest rate - your don't need to agonize over that, the tank is likely to outlive all of us regardless of fill rate.

O2 filled at as slow a rate as possible - do agonize over this.

Exactly. Hot gas can be a problem, scuba tanks overheated from fast filling, no. No doubt tanks get hotter in the back of a SUV in August in SoCal than they ever will via a fast fill.

The Scuba industry has done a great job of scaring people about tanks and temperature. I remember a guy that was convinced an electric heat gun on the neck of a steel cylinder with a stuck NGT valve would destroy the tank. Um...........no, not really.

Tobin

 

[bl6394]

You keep confusing the temperature of the gas with metallurgical damage. The specific heat of a gas is...

... do agonize over this.

You're presumptuous. I haven't asserted a quantitative relationship between the increased risk of failure with a rapid fill versus a slow fill - based solely upon a higher temperature. I've asserted qualitatively that temperature is generally a factor contributing to metal fatigue.

We understand that metal fatigue is affected by a variety of factors including temperature, surface of the material, load, etc. However, as far as I can tell - neither of us are metallurgists - nor have we studied all of the factors specific to scuba cylinders. I'm not in a position to make a recommendation regarding fill rates based upon my own research. I doubt you are either.

This is why we follow manufacturer recommendations and industry standards.

However, you cited the speed at which the load is applied is a factor in the material's failure. Perhaps you'd like to cite a source for your claim?

In any event, most FSO's aren't going to be agonizing about which metallurgical theory describes the manner in which their cylinder will fail. Most would care less about what effect work hardening may or may not have with a 6 minute fill time versus a 30 second fill time.

The important goal is a fill rate that is easily understood and followed to create a safe environment for the FSO - and provide a good fill for the customer.

 

[Jack Hammer]

@bl6394 It read a little weird about the gas temperature, Bjorn, but I understood what you were trying to say about tanks getting pretty hot with an extreme fill rate and it generally being a bad idea.

 

[cool_hardware52]

Take a full scuba tank and crank the valve full open and time how long it takes to completely drain. That's roughly how long it will take to fill with an constant pressure source. (Filling may be a bit shorter if the banked gas is at a pressure well in excess of the tank pressure)

~Around 60 seconds depending on the exact valve and working and volume of the tank. Any load applied over *minutes* is very very slow. Makes little difference if it take 30 seconds or 90 seconds, either rate of change in load is glacial.

Now compare that to an auto suspension component on a vehicle that hits a pothole at 70 mph. Fully loaded in millseconds. That *might* make a difference.......

Tobin

 

[cool_hardware52]

**Disclaimer** the information below is for hoop wrapped SCBA cylinders, *not* Scuba tanks. Similar info exists for Scuba Tanks, but I'm having trouble finding it.

The fact remains that the apparatus used to pressure cycle SCBA cylinders and Scuba cylinders apply the max load over a very short period of time. They *have* to do so quickly or the required 10K cycles would take *forever*.

If each of the 10000 test cycles was performed as per a "slow fill" i.e. 20 minutes it would require ~139 *days*

Does *anybody* think Luxfer is running tests for ~nearly 6 months to verify that a given production batch is fit?

The reality is the pressure is applied and relieved in seconds. At 5 seconds each the test would take ~14 hours. At 2 seconds the test would need ~5 hours. A 2 second cycle is easily possible, it's likely far faster.

When is the last time you had your scuba tank filled in 2 seconds? or 5 seconds? Anybody? Bueller? Anybody?

https://www.luxfercylinders.com/img/rm_img/blog_img/454/attachments/1/luxcompinspectmanual.pdf

From the above link
"Pressure cycling test—This test is conducted on one cylinder per lot. The cylinder is subjected to a minimum of 10,000 cycles to service pressure followed by a further 30 cycles to test pressure.

Each test cylinder must withstand the cycling pressurization test without evidence of visually observable damage, distortion or leakage. After successfully passing the cycling test, each test cylinder is burst-tested and must demonstrate a minimum burst pressure of 3.06 times service pressure."

Tobin

 

[rjack321]

The important goal is a fill rate that is easily understood and followed to create a safe environment for the FSO - and provide a good fill for the customer.

As Tobin posted, fill rate (and the heat of pressurization generated, whether it dissipates or not) does not affect the life or safety of a cylinder in any meaningful way. Even with extremely rapid pressurization and depressurization that can't be achieved with a gas at all (hydros are done with an incompressible fluid for a reason) cylinders last a lot longer than divers.

 

[bl6394]

As Tobin posted, fill rate (and the heat of pressurization generated, whether it dissipates or not) does not affect the life or safety of a cylinder in any meaningful way. Even with extremely rapid pressurization and depressurization that can't be achieved with a gas at all (hydros are done with an incompressible fluid for a reason) cylinders last a lot longer than divers.

You misunderstood what I meant when I asked you to cite a source. I didn't hope that you would quote some other guy on the internet, Tobin, who has posted in this thread.

This is an example of citing a source. Luxfer, on page 22 of their guide for scuba cylinders, provides direction regarding fill rates:

WARNING: Do not fast fill* (1,000 psig or higher) scuba cylinders. A fill rate of between 300 - 600 psig/minute is recommended.

https://www.luxfercylinders.com/img/rm_img/blog_img/455/attachments/1/scubaguide.pdf

They repeat that warning on page 25. And again on page 49 "A Luxfer aluminum scuba high pressure cylinder should never be fast filled." I don't know why their engineers, lawyers, or technical writers chose to say this about their product. That's OK - I'm a software architect and scuba instructor not a scuba cylinder engineer. As an end user - I don't need to know why the manufacturer says this. I just need to follow their recommendations or risk the consequences.

I'd be happy to learn something new from you today. But please cite a source or credible reference for your belief.

Thanks!

 

[cool_hardware52]

I'd venture a guess that tank manufactures recommend against fast fills because it encourages overfilling in order that the final pressure after the gas and the tank have returned to ambient don't piss off the the guy paying for the fill........

The fact remains that the acceptance testing done during manufacturing involves the application and release of pressures many 1000's of times that are many many orders of magnitude faster than the fastest possible fill using compressible gases.

Tobin