Gas cylinder treatment

Abstract

A method of treating a metal or metal alloy gas cylinder 10 comprises receiving the gas cylinder 10, the gas cylinder 10 comprising a body 12 defining an internal chamber 14 for storing gas 16 and a neck 18 connected to the body 5 12 and providing access to the internal chamber 14, applying a strain to the internal chamber 14 of the body 12 of the gas cylinder 10, the applied strain being less than an elastic strain limit of a material from which the body 12 of the gas cylinder 10 is formed but greater than 0.2%, forming a surface layer 20 on the internal chamber 14, and releasing the applied strain from the internal 10 chamber 14 of the body 12 of the gas cylinder 10.

Description

[0001] DESCRIPTION

[0002] GAS CYLINDER TREATMENT

[0003] This invention relates to a method of treatment of a gas cylinder.

[0004] Gas cylinders are widely used for the storage of compressed fluids such as gasses and liquids. A typical gas cylinder will be formed of aluminium or an aluminium alloy and will have an internal chamberwhere the fluid is stored under pressure. Gas cylinders are provided with an internal layer, which forms either naturally or is coated onto the metal or metal alloy that forms the internal chamber. This internal surface layer or coating is present in order to protect the material of the internal chamber and / or to improve the stability of the gas being stored within the gas cylinder. Since gas cylinders are filled with fluid under pressure, this layer will fracture when the surface is put under sufficient strain. This typically occurs, for example, when the cylinder is pressurized with the gas contents.

[0005] It is therefore an object of the invention to improve upon the known art. According to the present invention, there is provided a method of treating a metal or metal alloy gas cylinder, the method comprising the steps of receiving the gas cylinder, the gas cylinder comprising a body defining an internal chamber for storing gas and a neck connected to the body and providing access to the internal chamber, applying a strain to the internal chamber of the body of the gas cylinder, the applied strain being less than an elastic strain limit of a material from which the body of the gas cylinder is formed but greater than 0.2%, forming a surface layer on the internal chamber, and releasing the applied strain from the internal chamber of the body of the gas cylinder.

[0006] Owing to the invention, it is possible to provide a gas cylinder with a layer of material on the internal chamber in order to protect the metal or metal alloy that has been used to form the gas cylinder, which has an improved performance in relation to its fracture, when compared to existing gas cylinders. As stated above, the internal surface layer will fracture when the surface is put under strain, for example, when the cylinder is pressurized with the gas, or other fluid, contents. In order to reduce the likelihood of this fracture, a small amount of strain is applied to the surface of the internal chamber, during the forming of the internal surface layer or coating. This strain is at a low level, for example 0.3% strain, in the elastic range of the substrate material and below the yield point of the material. This application of a strain during the forming of the surface layer protects the layer from subsequent fracture during use, as the strain at which the surface layer fractures is now higher than it would otherwise have been. Preferably, the strain applied during the step of applying a strain to the internal chamber of the body of the gas cylinder is less than 80% (and more preferably less than 60%) of the elastic strain limit of a material from which the body of the gas cylinder is formed.

[0007] The improved method can be used as a process for improving the %strain to fracture of, for example, a thin aluminium oxide (alumina) layer that forms on the inner surface of gas cylinders made of aluminium and aluminium alloys. The fracture of this oxide layer has a detrimental effect on the cylinders and, in particular, on the stability of any gas stored within them. The improvement in the fracture resistance of the inner surface is dependent on the strain imposed during the forming of this layer, and even low strains lead to marginal improvements. The method of present invention is a process to improve the % strain needed to fracture a thin layer formed on the internal surface of a gas cylinder, for example a gas cylinder constructed using aluminium and aluminium alloys. The internal layer (which is either an additional layer or a modified layer) provided on the surface of the internal chamber has a protective (“passivating”) effect that helps prevent further oxidation and / or corrosion of the gas cylinder.

[0008] In relation to aluminium alloy all-metal high pressure gas cylinders, the improved process results in a protective layer on the internal chamber that provides additional positive effects over and above the minimization of oxidation and / or corrosion internally within the gas cylinder that has been treated in this way. An additional enhancement of the gas cylinder is the improved long-term gas stability for both high purity gases and specialized gas mixtures contained in high pressure gas cylinders. The improved process prevents the fracture of the internal surfaces formed in gas cylinders, which will otherwise inevitably be fractured during cylinder pressurization in all commercial cylinders filled to their design working pressure. The improved method involves causing a “local disruption” of, for example, the originally formed oxide / metal interface on the inner surface of the cylinders and this is carried out while the cylinders are under strain (for example by being pressurized).

[0009] Preferably, the step of applying a strain to the internal chamber of the body of the gas cylinder comprises applying a predefined pressure to the internal chamber of the body of the gas cylinder. A strain needs to be applied to the surface of the internal chamber of the gas cylinder, while the surface layer of the internal chamber is being formed. This strain can be generated by applying a pressure to the gas cylinder while the surface layer is being formed. This provides an effective solution to the requirement to provide the strain needed to treat the gas cylinder during the process of forming the surface layer of the internal chamber. The gas pressure can be controlled to provide the low level of strain needed to impart a strain to the internal chamber which is less than an elastic strain limit of a material from which the body of the gas cylinder is formed.

[0010] Advantageously, the strain applied during the step of applying a strain to the internal chamber of the body of the gas cylinder is less than 0.5%, and more preferably, less than 0.3%. The strain applied to the surface of the internal chamber of the gas cylinder needs to be sufficient to provide some elastic strain to the material of the internal chamber, while providing a strain that does not cause any plastic deformation of the material that is used to form the gas cylinder. Since gas cylinders are commonly formed using aluminium or an aluminium alloy, the limit of the applied strain can be set at a level such as 0.5%, and more preferably, less than 0.3%. This will achieve the necessary improvement in the damage resistance of the surface layer of the internal chamber while not causing any deformation of the metal used to form the gas cylinder. Embodiments of the present invention will now be described, by way of example only, with reference to the accompanying drawings, in which:

[0011] Figure 1 is a schematic diagram of gas cylinder,

[0012] Figure 2 is a schematic diagram of a method of forming a gas cylinder, Figure 3 is a flowchart of a process of treating the gas cylinder, Figure 4 is a graph showing the effects of pressure on a gas cylinder, Figure 5 is a schematic diagram of a test sample taken from a gas cylinder,

[0013] Figures 6a and 6b are a pair of graphs showing the current versus time response during cylinder anodization, and

[0014] Figure 7 is a schematic diagram of a 4-point load strain testing device.

[0015] Figure 1 shows a gas cylinder 10. The gas cylinder 10 comprises a body 12 defining an internal chamber 14 for storing gas 16. A neck 18 is connected to the body 12 and provides access to the internal chamber 14. The body 12 is formed from a metal such as aluminium or an aluminium alloy. A surface layer 20 is provided on the inner surface of the internal chamber 14. The neck 18 of the gas cylinder 10 is closed by a valve 22 that is mounted at the neck 18 of the gas cylinder 10. The gas cylinder 10 can be filled with the gas 16 under pressure via the valve 22.

[0016] Figure 2 shows schematically how the gas cylinder 10 can be formed. A metal or metal alloy ingot 24 is cut to create a block 26, which is then lubricated. The lubricated block 26 is then introduced into a mould 28 and a punch-tip 30 is applied to the block 26 under extremely high pressure. This process of effectively backward extrusion creates a partially formed component 32, which is essentially the gas cylinder 10 before the creation of the neck 18 of the gas cylinder 10. At this point in the process, the partially formed component 32 is washed using a steam lance and a nylon brush with a citric acid-based detergent. The washed component 32 is then rinsed with clean uncontaminated water and air dried.

[0017] The component 32 is then introduced into a head forming device 34, where at least the upper part of the component 32 is heated while pressure is provided from below. The shape of the head forming device 34 and the effect of the heat and pressure on the component 32 results in the shoulders and neck 18 of the gas cylinder 10 being formed and it is at this point in the process that the component 32 is turned into a gas cylinder 10. The gas cylinder 10 is cooled and a neutral detergent wash is performed both internally and externally on the gas cylinder 10.

[0018] The gas cylinder 10 is then passed to a neck thread forming stage during which the neck is machined to form a screw thread on the internal surface of the neck 18, in order to be used for the subsequent mounting of the valve 22. The screw thread can also be machined onto the external surface of the neck 18, if required. At this point, the gas cylinder 10 can be additionally treated, for example with heat treatment, quenching, or the application of a coating material 20 on the surface of the internal chamber 14. The gas cylinder 10 is then inspected and any further customisation can be applied after the inspection.

[0019] Figure 3 shows a method of treating the gas cylinder 10. The method comprising the steps of, firstly step S1, which comprises receiving the gas cylinder 10, secondly step S2, which comprises applying a strain to the internal chamber 14 of the body 12 of the gas cylinder 10, the applied strain being less than an elastic strain limitof a material from which the body 12 of the gas cylinder 10 is formed but greater than 0.2%, thirdly step S3, which comprises forming a surface layer 20 on the internal chamber 14, and finally step S4, which comprises releasing the applied strain from the internal chamber 14 of the body 12 of the gas cylinder 10.

[0020] The treatment process of Figure 3 is designed to provide a gas cylinder 10 that has been provided with a thin surface layer of material 20, which has an improved performance in relation to the fracture of the surface layer 20. As stated above, during the process of manufacturing the gas cylinder 10, the cylinder 10 can be internally coated with a material 20 or the surface layer 20 forms automatically (for example when polished aluminium is exposed to air). This surface layer 20 has two primary purposes, firstly to protect the internal chamber 14 from damage (forexamplewhen the gas cylinder 10 is pressurised) and secondly to protect the gas contents of the gas cylinder 10 from degradation through contact with the metal or metal alloy used to form the body 12 of the gas cylinder 10.

[0021] The inner surface layer 20 of the internal chamber 14 will fracture when the surface is put under strain, for example, when the cylinder 10 is pressurized with the gas contents. The treatment process described above will reduce the likelihood of this fracture, and this is achieved by a small amount of strain (above a level of 0.2%) being applied to the internal chamber 14 while the surface layer 20 is being formed. This strain is at a low level, for example 0.3% strain, in the elastic range of the substrate material and below the yield point of the material. There is no plastic deformation of the surface layer 20 as a result of the strain being applied. The level of strain should be no more than 80% and ideally no more than 60% of the elastic strain limit of a material from which the body 12 of the gas cylinder 10 is formed. This application of a strain during the forming process protects the surface layer 20 from subsequent fracture during use, as the strain point at which the surface layer 20 fractures is now higher than it would otherwise have been.

[0022] Figure 4 shows a graph that illustrates the effect of pressure on a gas cylinder 10, which is a 111 mm OD, 6.8 mm Wall 25.58 MPa (Working Pressure) -371 1 psi. The lower axis shows the amount of pressure being applied (in MPa) to the internal chamber 14 of the gas cylinder 10. The left-hand axis shows the hoop stress (in MPa) suffered by the gas cylinder 10 and this is plotted on the graph with the dots labelled “Stress”, the upper line of dots. The solid dots represent a thin-walled cylinder 10 and the open dots represent a thick-walled cylinder 10. What this shows, for example, is that a conventional gas cylinder placed under a pressure of 20 MPa, suffers approximately 150 MPa of hoop stress.

[0023] The right-hand axis shows the hoop strain (in %) suffered by the gas cylinder 10 and this is plotted on the graph with the dots labelled “% Strain”, the lower line of dots. Again, the solid dots represent a thin-walled cylinder 10 and the open dots represent a thick-walled cylinder 10. What this shows, for example, is that a conventional gas cylinder placed under a pressure of 25 MPa, suffers approximately 0.3% hoop strain. This is the normal working pressure of a gas cylinder 10. Also shown on this graph is the hoop strain % required to cause an internal surface oxide fracture, which is slightly above 0.1%, labelled “existing cylinder performance”. This will occur at a cylinder pressure of approximately 12.5 MPa, which is well below the working pressure required for a conventional gas cylinder 10.

[0024] The flowchart of Figure 3 is a process to improve the % strain to failure / fracture provided by the thin layers 20 formed on the surfaces of aluminum and aluminum alloys. The % hoop strain needed to fracture the surface layers generated on the inner walls of commercial all-metal high pressure gas surfaces (which have not been treated by the process of Figure 3) are generated by pressures within a cylinder well below those promoted by a cylinder designed working pressure, as can be seen in Figure 4. Also shown in this Figure, labelled “New Innovation”, shows the % hoop strain required to cause a fracture in a treated gas cylinder 10, which is now above the strain that the gas cylinder 10 would undergo at normal working pressure. The % strains to fracture measurements are consistent with the published experimental values and recent modelled data for aluminium and aluminium alloys.

[0025] In order to test the effectiveness of the treatment described in the process of Figure 3, various working examples were created and tested as described below.

[0026] EXAMPLE 1

[0027] The first working test used gauge length regions of convention rectangular cross-section tensile test samples taken from a 250 mm diameter, 17.2 mm wall thickness, AA6061 aluminum alloy high pressure gas cylinder, as shown in Figure 5. Each cylinder sample 36 was coated with beeswax to mask-off all external surfaces other than a central area of the cylinder internal wall measuring approximately 1.5 mm x 1.5 mm.

[0028] Figure 5 shows the gauge length section of a longitudinal tensile sample 36 that has been machined from a 250 mm diameter, 17.3 mm wall thickness commercial high pressure AA6061-T6 cylinder and has a test cell 38 surrounding the relevant part of the sample 36 to contain an aqueous test solution 40. The non-beeswaxed surfaces were exposed to an aqueous solution (pH 8.2) of 7.1 g / liter boric acid + 8.1 g / liter disodium tetraborate (both ANALAR grade) dissolved in deionized water and subjected to an anodizing treatment by applying +2 volts (relative to the standard Silver / Silver chloride reference electrode potential) using saturated potassium chloride (KCI) as the internal electrolyte (Ag / AgCl / Sat KCI) to generate a “barrier” alumina film around 3 nm thick. Prior to the beeswaxing treatment the tensile specimen gauge length surfaces to be subjected to anodization were given a light mechanical polishing terminated using fine particle size alumina particles suspended in light mineral oil that resulted in a “mirror” finish. Any organic polishing residue was removed via degreasing with isopropyl alcohol.

[0029] A series of tests were conducted on the individual tensile specimens 36 such that each sample 36 was subjected to a standard anodization treatment for thirty minutes while continuously held under fixed displacement control with initial loads of zero, 50, 100, 150, 200 and 225 MPa in a conventional screw-driven mechanical tensile testing machine. During the anodizing treatment, the electrochemical current required to maintain the applied +2 V relative to Ag / AgCl / Sat KCI electrode potential was continuously monitored. The current was also continually monitored during load removal over a minute afterthe thirty-minute period had expired. Within two minutes, each sample was reloaded to 225 MPa using a nominal strain rate of around 10’4 / s, again with the current associated with maintaining anodization potential (+2 V relative to Ag / AgCl / Sat KCI) being monitored.

[0030] Analysis of the final stages of this data as a function of the strain imposed during straining clearly reveals a major increase in the anodizing current associated with the initial fracture an anodized barrier surface layer, as illustrated in Figures 6a and 6b. These Figures shows two graphs that show the current flowing during the final stages of anodization process while increasing strain was applied at a nominal strain rate of 10'4 / s. The % strain to fracture surface layer identified by a sudden increase in current. As stated above, the tests on the samples 36 were performed using different initial loads of zero, 50, 100, 150, 200 and 225 MPa. The two graphs shown in Figures 6a and 6b show the results in relation to the penultimate of these tests, where the sample 36 was anodized while under a strain load of 200 MPa. The two graphs in Figures 6a and 6b show the current flowing against time after the surface layer under strain had been completed and then reloaded after two minutes with increasing strain up to a maximum stress of 225 MPa. The point at which the current rapidly increases in the graphs indicates the point at which the anodized layer has fractured, since a much larger current is required at this point to maintain the 2 volts used in the anodizing, indicating a rapid increase in the uncoated / unprotected area. What is shown in the graphs (Figure 6a prior to baseline correction and Figure 6b subsequent to baseline correction) is that a strain of 0.32% is required to cause the fracture that is indicated by the capacitive spike in the current flowing.

[0031] Table 1 below shows the data for the % strain to fracture of the surface layer as a function of the loading applied during the thirty minutes anodizing treatment. The data provided in Table 1 clearly demonstrates the marked beneficial increase due to the imposition of an elastic stress > 100 MPa during the generation of the barrier surface layer.

[0032] Initial Stress at % Strain during Elastic Strain to start of Anodizing Anodizing Fracture Anodized (MPa) Layer 0 0 0.1, 0.13,

[0033] 50 0.073 0.14, 0.15

[0034] 100 0.145 0.23 150 0.217 >0.3 200 0.290 >0.3, >0.35

[0035]

[0036] 225 0.326 0.32, 0.38

[0037] Table 1

[0038] The rows in the table illustrate the different performances of the samples 36 used in the tests described above. The first column indicates the amount of stress (in MPa) applied to the sample 36 while the sample 36 was being anodized to create the surface layer 20 on the sample 36. The second column shows the % strain being applied to the respective sample 36 while the anodizing was occurring under the stress. The third column shows the subsequent % strain required to fracture the surface layer 20 after the testing was completed.

[0039] The first row in the table represents the current gas cylinder technology (a control sample), in that no stress, and hence no strain, is applied to the gas cylinder 10, while the surface layer 20 is formed (the anodizing). As can be seen from the third column, after this process is completed, the surface layer 20 will fracture at levels around 0.1 and 0.13 % strain applied, as was found for internal cylinder surfaces of control samples not subjected to anodizing. Since this is below the working pressure of current gas cylinders, this implies that in general surface layers will fracture during using of the gas cylinders. In the second row of the table, 50 MPa is applied to the sample 36 while the sample 36 is being applied. As can be seen from the third column, there is very little improvement in the fracture resistance of the surface layer. At 100 MPa a reasonable improvement in the fracture resistance has been shown, but the most significant improvement is demonstrated when the pressure applied during anodization is at 150, 200 or 225 MPa.

[0040] EXAMPLE 2

[0041] Figure 7 shows a testing device 42 that was used in a second set of tests relating to the material treatment process of Figure 3. The device 42 comprises a sample 36 which is located between two pairs of bars 44 that apply a defined load according to the direction of the arrows in the Figure. The sample 36 is placed between the bars 44, part of which is exposed and part of which is covered. A tray 46 is present that allows fluids to be held in contact with the sample 36. Here the sample 36 is an aluminium alloy AA6061-T6 commercial rectangular extruded bar (25.4 mm by 12.7 mm), rather than samples taken directly from a commercial gas cylinder, as in the tests of Figures 5 and 6. The 4-point bend load test specimens were coated with beeswax 48 to mask-off all external surfaces other than a central area 50 (~2.2 mm x 2.5 mm) of the upper surface in between the inner loading points were subjected to various loads subjected to the anodizing treatment involving the aqueous environment and electrochemical conditions used in previous tests. Individual test specimens were used for each applied load evaluated in the previous tests, whereas for the tests of type shown in Figure 7, multiple loading conditions were assessed on several specimens. After two hours anodization loading was removed and almost immediately re-applied up to 225 MPa, using a nominal strain rate of 5 x 10'4 / s. The observed % strains needed to fracture barrier anodized layers generated during these tests are broadly equivalent to those found in the previous test. Table 2 below shows the outcome of the tests taken directly on the metal alloy.

[0042] Initial Stress at Elastic Strain to start of Anodizing Fracture Anodized

[0043] (MPa) Layer

[0044] 0 0.074

[0045] 50 0.14

[0046] 100 0.17

[0047] 150 0.24

[0048] 200 >0.35

[0049] 200 + 30 mins

[0050] 0.064

[0051]

[0052] without load

[0053] Table 2

[0054] The results shown in Table 2 for the second set of tests, taken on a bar of metal alloy are comparable with the first set of tests, the results of which are shown in Table 1. As before, the control sample in row 1, which has no stress applied during the anodizing process has a very low strain to fracture, with a fracture occurring at 0.074%. Noticeable improvement in the performance of the anodized layer occurs when the stress applied during the anodizing is at least 150 MPa, with the best performance occurring in the layer that was anodized while the sample 36 was under a pressure of 200 MPa, consistent with the results from the first set of tests on the sample taken directly from a gas cylinder 10.

[0055] In the final row of the table, two hours additional anodizing without an applied load following an initial 30 minutes anodizing with an applied load was carried out. This resulted in a total loss of the beneficial improvement % strain to failure of the barrier layer, indicative that continuous rather than intermittent load application is required for beneficial effects.

[0056] EXAMPLE 3

[0057] In light of the result of the process shown in the final row of Table 2, a series of tests were conducted on the tensile test samples used in the first set of tests (Figures 5 and 6), applying the same anodizing treatment for thirty minutes under strained conditions using initial applied stress of 200 MPa. The difference from the first testing being additional anodizing times without an applied load for times in excess of thirty minutes consistently reduced the % strain to failure to 0.12 % and thereby eliminating beneficial improvements, reproducing the result shown in the final row of Table 2 above, for a commercially AA6061-T6 extrusion alloy.

[0058] EXAMPLE 4

[0059] A further fourth set of tests were also carried out in which elastic stress / strain was imposed on the internal walls of a commercial AA6061-T651 high-pressure gas cylinder using those promoted during the gas pressurization of such cylinders. The hoop stresses and strains generated during pressurization may be estimated by considering cylinders as either “thin-walled” or “thick-walled” vessels, as shown previously in relation to Figure 4. Consideration of the strain data as a function of the cylinder fill-pressure, which will be similar for commercial cylinders irrespective of the national or international design code used to design the gas cylinder, reveals the hoop strains generated on cylinder internal wall surfaces during pressurization will exceed those known to fracture the surface layers when internal pressures are significantly below those developed at their designed working pressures.

[0060] In this final test, a commercial AA6061 aluminum alloy high pressure gas cylinder 10 was pressurized during the final stages of being subjected to an internal surface generation treatment disclosed in US Patent 9,085,062, the contents of which are incorporated here by reference, where the interfacial surface region between the metallic alloy and the developing surface oxide was continually disrupted during a tumbling operation and the local polishing process leads to a tailored hydrated oxide layer. While improved gas stabilities of unstable gas mixtures (for example, low concentrations of hydrogen sulfide, H2S in air) stored in such aluminum alloy high pressure gas cylinders have been achieved using the process described in this US Patent without pressurization of the cylinder during the final polishing process, a significant further improvement in gas stability will result when the cylinders 10 are pressurized to generate internal stresses of around 200 MPa and the internal cylinder surface oxide facture is now prevented by use of the treatment process of Figure 3.

[0061] In this process, the gas cylinder 10 is rotated at a defined speed for a defined period of time. This can be carried out in a single step or multi-step process. An abrasive mixture, preferably in the form of a slurry, is introduced into the internal chamber 14 and moved across the inside surface 20 of the cylinder 10 in order to abrade or polish the surface to reduce its surface roughness. A preferred method of achieving this is to introduce the abrasive slurry into the cylinder 10 and then rotate the cylinder about a horizontal axis. The speed of rotation and the composition of the slurry are set so as to promote a continuous motion as the cylinder 10 is rotated, so that the media abrades the inside surface 20. The abrasion exposes a freshly created aluminium surface that reacts directly with the “local environment” that is established at the metal / solution interfacial region to form a thin layer of aluminium oxides and / or oxyhydroxides on the inside surface 20 of the cylinder 10. As mentioned above, the gas cylinder 10 is pressurized during the final stages of being subjected to the internal surface generation treatment, in order to apply the necessary strain, while the surface layer 20 is being formed.

Claims

CLAIMS1. A method of treating a metal or metal alloy gas cylinder (10), the method comprising the steps of:• receiving the gas cylinder (10), the gas cylinder (10) comprising a body (12) defining an internal chamber (14) for storing gas (16) and a neck (18) connected to the body (12) and providing access to the internal chamber (14),• applying a strain to the internal chamber (14) of the body (12) of the gas cylinder (10), the applied strain being less than an elastic strain limit of a material from which the body (12) of the gas cylinder (10) is formed but greater than 0.2%,• forming a surface layer (20) on the internal chamber (14), and • releasing the applied strain from the internal chamber (14) of the body (12) of the gas cylinder (10).

2. A method according to claim 1, wherein the step of applying a strain to the internal chamber (14) of the body (12) of the gas cylinder (10) comprises applying a predefined pressure to the internal chamber (14) of the body (12) of the gas cylinder (10).

3. A method according to claim 1 or 2, wherein the strain applied during the step of applying a strain to the internal chamber (14) of the body (12) of the gas cylinder (10) is less than 0.5%.

4. A method according to claim 3, wherein the strain applied during the step of applying a strain to the internal chamber (14) of the body (12) of the gas cylinder (10) is less than 0.3%.

5. A method according to any preceding claim, wherein the step of releasing the applied strain from the internal chamber (14) of the body (12) ofthe gas cylinder (10) is carried out after completion of the forming of a surface layer (20) on the internal chamber (14).

6. A method according to any preceding claim, wherein the step of forming a surface layer (20) on the internal chamber (14) comprises applying a coating material (20) to the internal chamber (14) of the gas cylinder (10).

7. A method according to any preceding claim and further comprising the step of rotating the gas cylinder (10) at a defined speed.

8. A method according to claim 7, and further comprising the step of introducing an abrasive slurry into the internal chamber (14) while rotating the gas cylinder (10) at the defined speed.

9. A method according to any preceding claim, wherein the strain applied during the step of applying a strain to the internal chamber (14) of the body (12) of the gas cylinder (10) is less than 80% of the elastic strain limit of a material from which the body (12) of the gas cylinder (10) is formed.

10. A method according to claim 9, wherein the strain applied during the step of applying a strain to the internal chamber (14) of the body (12) of the gas cylinder (10) is less than 60% of the elastic strain limit of a material from which the body (12) of the gas cylinder (10) is formed.

Images