GoTo HomePageHarwood Engineering Company, Inc.

Some Safety Problems Associated With High Pressure Equipment — Case 38142

by O.L. Anderson

Available information is assembled on the safety features involved in the operation of some high pressure equipment.
This memorandum discusses the construction and erection of barriers, the problem of barricading equipment, protection from burns, the design and operation of high pressure vessels, and hazards connected with coupled hydraulic systems. The summary of recommendations is given on pages 14, 15, 16.

Introduction

In spite of accelerated activities in high pressure* experimentation, the available information on safety procedures in the use of such equipment is meager. Since the installation of high pressure equipment is contemplated by Bell Telephone Laboratories, it is deemed necessary to assemble available information on pertinent safety devices and procedures. Some of the information discussed here must be considered tentative and subject to change. Some items discussed here are published and may be considered authoritative. Much of it is lore, passed on by word-of-mouth from those few who have experience. A few items are original with the author. We shall attempt to discuss safety devices in terms of basic physics, whenever possible.

It must be understood that high pressure equipment is inherently dangerous, and no amount of gadgeteering will eliminate all hazards. It is difficult to attain pressure levels in high pressure equipment that are useful for experimental purposes, and yet maintain the high factor of safety deemed desirable in engineering practice.

Safety equipment, such as check valves, rupture disks and relief valves should be engineered into a high pressure hydraulic system whenever possible. However, there is another kind of risk when a hydraulic system becomes complicated by automatic safety devices. This risk is the ignorance or indifference of an operator who has come to depend upon such safety devices so much that he has forgotten them. For this reason high pressure systems should be simple.

There are three types of hazards connected with the operation of high pressure gear:

  1. Shock from pressure blast
  2. Fragments and debris
  3. Burns
The first type is especially important in experiments using a gas as the pressure medium, or whenever certain chemical reactions are likely. It has been shown1 that more than half the energy released by an explosion is radiated outward by the shock wave. Low shock pressures (2 to 15 psi overpressure) may be sufficient to damage most man-made structures.2 The energy released to create such overpressures may result from combustion, or may be due to the release of elastic energy stored in a compressible fluid. If non-reacting fluids are used, the combustion hazard is eliminated. If liquids rather than gases are used as pressure media, the stored elastic energy is not sufficient to create a dangerous shock wave.

The flying fragment hazard is present in all types of high pressure work. The safety problem here is centered around a design which will protect personnel and equipment from flying fragments in the event of an explosion. There are two types of design problems. The first is the creation of pressure equipment which will operate safely under the desired pressure. The second is the installation of barriers which will stop the fragments in case the pressure equipment fails. Pressure equipment includes valves, piping, fittings, rupture disks, intensifiers and pumps, as well as pressure vessels.

The third type of hazard is perhaps the most dangerous because it is the least appreciated. For example, there is the possibility that compressed fluids will flash or geyser to a vapor state (or simply become hot) if they suddenly pass through an orifice. In this case, elastic energy is converted to heat. This possibility is especially dangerous when an operator is removing the closure of a pressure vessel while a residual pressure remains inside the vessel.**

The Design of Shelters

The safest way in which to conduct high pressure experiments is to have no personnel in the vicinity of the pressure equipment. This is done by isolating the equipment in some kind of shelter. There are three factors governing the penetration of a projectile into a resisting mass:
  1. The velocity of the projectile at impact.
  2. The weight and cross-sectional area of the projectile.
  3. The characteristics of the resisting material comprising the shelter.
These factors have been combined in several empirical expressions. The one used by the Department of Navy3 is the so-called Petry formula in which the depth of penetration (in feet) in an infinite medium is given by:
Equation 1
where V', the velocity factor, is given by:
Equation 2
where the velocity, V, is in ft/sec., Ap is called the sectional pressure and is equal to the weight of the projectile divided by its maximum cross-sectional area, and K is the coefficient of penetration, a property of the shelter.

The coefficient of penetration has been experimentally determined4 for some materials which are listed in Table I.
 
 
Table I
Material
K x 103
ft3 lb-1
Notes
Alloy Steel
0.26
Reference 3, Page 14
Steel
0.40
Reference 3, Page 14
Limestone
5.38
Concrete
7.99
2,206 psi crushing strength
Reinforced Concrete
4.76
3,300 psi crushing strength
Special Reinforced Concrete
2.82
5,700 psi crushing strength
Stone Masonry
11.72
Brickwork
20.48
Sandy Soil
36.7
Soil with Vegetation
48.2
Soft Soil
73.2
Equation 1 is applicable where the thickness of the shelter is several times the depth of penetration. Experiments reported by the Navy5 show that the minimum thickness of the barrier must be three times the penetration. That is to say, if the thickness of the shelter is made less then 3D, the actual penetration is more than the D calculated by the Petry formula (see Figure 1).

Examination of Table I shows that a concrete shelter should be 18 times as thick as a steel one, and an earth barricade should be 4 or 5 times as thick as a reinforced concrete structure.

An efficient structure used for outdoor shelters consists of earth surrounding a crib made of interlaced logs. The coefficient of penetration of this structure has not been measured, but it must be less than for soil. This type of shelter was used as a bunker with success by the Japanese Army in its Pacific operations in the last war. Such a structure is also used for testing by the Harwood Engineering Company which manufactures high pressure equipment exclusively. This type of shelter is relatively inexpensive since its construction involves only excavation and carpentry, and both operations must precede the erection of a reinforced concrete building. An illustration of the shelter used by Harwood Engineering Company is given in Figure 2. It is always safer to conduct high pressure experiments in an outside shelter rather than inside a laboratory. When the experiments involve (a) gases as pressure media, (b) testing to failure, (c) testing under unknown conditions (for example, at high temperatures), an outside shelter is essential.

In using the Petry formula to compute the penetration, it is necessary to calculate the velocity and the sectional pressure.

Barricading Elements of a High Pressure System

Locating Steel Barriers
After determining the thickness of a barrier, there still remains the task of placing it in the safest position. It is, of course, obvious that the barrier must be anchored against a solid mass such as a table or column, so that it cannot be blown over. It is just as important to fasten the barrier in such a way that any bolts or metal fasteners are not subjected to sudden tensile stress resulting from an explosion. In the latter event, boltheads may be torn loose and become lethal fragments (see Figure 3).

Whenever possible, metal fasteners which might be exposed to sudden tensile stresses should be eliminated. Sometimes they can be replaced by welding or by using extruded tubing. Barriers must always be placed so as to reduce the hazards of ricocheting fragments.

The Principle of the Ballistic Pendulum
Many pressure laboratories and high pressure equipment manufacturers design their shielding so as to use the principle of the ballistic pendulum. That is to say, they do not fasten the barrier rigidly but allow some degree of motion so that part of the kinetic energy of the colliding fragment is converted into kinetic energy of the barrier. Several applications of this principle are illustrated in Figure 4. Dr. H. C. Miller, director of the high pressure laboratory of the Armour Research Foundation, has told the author that in his experience rigidly attached armor plate glass which shatters when struck, may only crack when it is allowed to swing with the fragment upon collision.

Only a small fraction of the kinetic energy of a projectile will be converted to kinetic energy of the barrier. The proportion so converted depends upon the velocity, size, and aerodynamic properties of the projectile. The lower the velocity and the more irregular the shape, the larger this proportion. Although more quantitative results need to be obtained by actual experiment, there is probably a tangible safety feature in the use of the ballistic pendulum (see Appendix B).

For design purposes, it is probably safest to build a barrier as though it were to be fastened rigidly, and then mount it, incorporating the principle of the ballistic pendulum whenever practicable. This principle is advocated by Mr. D.H. Newhall, manager of Harwood Engineering Company.

Shielding Pipes and Tubing
High pressure tubing is manufactured today which is designed to withstand pressures in excess of 200,000 psi. For obvious reasons, tubing should be kept as short as possible and should be shielded. One practical and economical shielding device is to run the high pressure tubing through heavy steel pipe. Personnel should never be exposed to the tubing. When it is necessary to run tubing across the floor, the staff at Armour Research Foundation cover it with heavy channel iron.

In cases of rupture, the tubing is most apt to spring a leak without throwing a fragment. There are, however, other hazards. A jet of high velocity fluid can easily penetrate flesh. There is also the possibility of burns resulting from exposure to hot vapors and steam. If sufficient space is left between the barrier and the tubing, a fluid jet is easily deflected, and hot vapors condensed.

Shielding Pressure Indicating Instruments
At least two safety devices are available for indicating pressure. Foxboro*** manufactures a pneumatic system which transmits pressure recordings by a 3-15 psi air pressure system. By using the principle of the electrical strain gauge, pressure cells have been developed which allow a practical (although expensive) electronic measurement of high pressures up to 200,000 psi****. If one or the other of these devices is used, an operator need not be exposed while taking pressure readings.
Shielding Valves
Occasionally a hydraulic system must be controlled by a valve. The hazard here is the possibility that while the operator is turning the valve, the stem will be projected violently outward. The obvious answer is to turn valves by indirect methods. High pressure equipment manufacturers have met this problem in two ways. Some valves are operated pneumatically, and some by a secondary low pressure system. Both of these methods are expensive and subject to mechanical defect. A much cheaper and more effective way has been designed by the author. This method uses a mechanical coupling by gears, see Figure 5. In this design, the handle of the high pressure valve is replaced by gear 1. This gear meshes with gear 2 which is mounted on an axle passing through the barrier. The thread of the axle of gear 2 is made identical with the thread of the stem in the valve, so that both gears advance uniformly. In case of a blow-out the stem and gear 1 pass over gear 2 and are stopped by the barrier. This design has the following advantages:
  1. The cost of this coupling is a few dollars compared with several hundred using an indirect hydraulic or pneumatic system.
  2. Disassembly of the high pressure stem from the barrier is easily permitted by disengaging the gears.
  3. The frictional forces on the threads of the valve stem can be overcome by making appropriate gear ratios.
  4. Arbitrary metering of the valve is possible.
  5. Gear 2 can be operated by an Allen wrench which reduces the possibility that unauthorized personnel will accidentally (or intentionally) turn the valve.
  6. Micro switches can be mounted which will detect the position of gear 1 and operate a signal.
This design has the approval of the Harwood Engineering Company.
Shielding Pressure Vessels
Pressure vessels are loaded in two ways. The pressure is often produced by a hydraulic press. Otherwise the pressure is produced by an intensifier supplied by a hand pump or automatic pump. In the first case, the pressure vessel is mounted between the platens of a press. In the second case, the pressure vessel can be placed in an arbitrary position. When a hydraulic press is used, the platens and stanchions of the press themselves as barricades. The mass of the press can be used to buttress the barriers. There are many possible schemes for shielding personnel from the pressure vessel. One scheme in use in existing pressure laboratories***** is shown in Figure 6a. Another possibility is shown in Figure 6b.

When the pressure vessel is activated by an intensifier, the problem of shielding takes on a different aspect. A convenient way to shield vessels which are also heated, is to armor-plate the inside of the oven (or water jacket). This can be done quite economically by using standard annealing ovens (for example, see the annealing ovens manufacturing by Gruenberg Electrical Co.) in which the steel is on the inside and the insulation on the outside. If the oven is turned upside down, and is not fastened, the weight of the oven acts as a ballistic pendulum in the event of an explosion, see Figure 4b.

If the pressure equipment is conducted at room temperature, the annealing oven can be replaced by a section of extruded steel tubing (see for example, the "Steel Catalog" of J.T. Ryerson & Sons, Inc.). This tubing is a safe shielding since it has no seams, and still the most economical since fabrication costs are low. A plate welded on the top of the cylinder provides protection from above. Again such a steel tube can be made into a ballistic pendulum.

We can say a few words about the dangers involved in the use of masses as ballistic pendulums. Unless the weight of the pendulum is large, it may become a dangerous projectile itself. In Appendix B it is shown that the height to which a ballistic pendulum will rise varies directly proportional with the pressure, and inversely with the weight of the pendulum. To the author's knowledge, there are no published data on the proportionality constant. This constant should be small, otherwise the penetration of the projectile into the mass is too small. The author has assurances, which must be regarded as lore, from two persons (H.C. Miller and D.H. Newhall) with extensive experience in the high pressure field, that protective bodies which weigh several hundred pounds will not translate more than one or two inches assuming the worst case of conditions of impact from a flying projectile. This information must be checked by experiment. Before such experiments can be performed, it is advisable to rely on the advice of people who have had experience. This advice is: it is safer to use the principle of dynamic pendulum than to secure barriers.

Shielding Fittings
Elbows, tees, collars, couplings, and other hydraulic fittings are no less dangerous than other elements of a high pressure system. Each fitting must be barricaded. A useful scheme is to fasten all fittings and tubing on the bottom of the plate on which the pressure vessel is mounted (see Figure 13). If this plate is loosely attached to the top of a table, the weight of the plate, pressure vessel, and oven, all act against a possible explosion from the plumbing. When practicable, elbows should be replaced by bent tubing.
Shielding Intensifiers
An intensifier may be described as a pressure vessel which separates high pressure fluid from low pressure fluid by means of a piston. An operator pumps up the low pressure side and is able to produce very high pressures on the high pressure side.

In case the intensifier is mounted in a fixed position, this apparatus can be shielded in the same way as the pressure vessel. However, if there are a number of independent pressure vessels in a pressure system, the cost of installing an intensifier for each pressure vessel may be prohibitive. In this case a single intensifier is used. Two hazards must be faced. The first hazard arises because the intensifier may be required to be portable, and the second hazard arises because a coupling must be made in the high pressure line between the pressure vessels and the intensifier. The first hazard can be reduced by mounting the intensifier and pump onto a cart which has shielding built around the intensifier (see Figure 4c). The second hazard can be reduced by enclosing the coupling in a piece of heavy steel pipe.

Shielding Against Burns
Here the problem is to protect personnel from physical contact with hot vapors or liquids which unexpectedly gush out of the pressure system. Ordinarily, the same barriers which protect personnel from flying fragments, will protect them from burns. The danger arises when the operator thinks the pressure is completely reduced and goes behind the barriers to work on the equipment. The safety rule which must be followed to avoid burns is: always have some kind of solid barrier between the operator and the equipment until there is absolute certainty that the pressure within the hydraulic system is reduced to atmospheric. Sheets of plastic, cardboard, or plyboard are sufficient to deflect liquid jets, or hold vapors until they condense.

There appears to be no standard procedure in the industry to avoid these types of hazards. The cartoon appearing as Figure 7 shows one possible procedure.

Shielding Against Shock Waves
Because the change in volume of liquids is so much smaller than in gases, it is common practice not to shield high pressure equipment against shock waves when there is no danger of combustion, and when the pressure fluid is a liquid rather than a gas. For example, Bridgman6 shows that all liquids seem to approach a limit in reduction of volume to about 75% of the original, while gases decrease to several thousandths of their original volume at high pressures (of the order 150,000 psi).

Calculations of the pressure front in the event of a gas explosion can be made.7 For information on the design of structures to withstand shock waves of such pressures see References 1 and 3.

The Design of High Pressure Vessels

Operating Pressures
In the last few years, experiments have been reported at pressures up t 1,000,000 psi, and also at very high temperatures (L. Koes of the Norton Co. recently reported manufacturing a new form of silica at 30,000 atmospheres and 1,000°C). These enormous pressures are a result of Bridgman's development of his so-called "unsupported area" seal. Using this seal, the closure of a vessel is effectively stronger than the vessel itself. This seal must be used at pressures above 50,000 psi. As long as the seal is functioning properly, the only limit to an experiment is the strength limit of the metal of the pressure vessel.

Extensive experiments by Bridgman have shown that cylinders subjected to internal pressures will withstand pressures of much higher magnitude than predicted by popular theories of strength. For example, he found8 that tool steel withstood internal pressures four times the tensile strength.

Pressure vessels are designed today by an empirical rule, which is recommended by manufacturers. The following expression has been found experimentally to be satisfactory for a ductile steel cylinder free of imperfections:

P = T ln(ro/ri)
where P is the pressure at rupture, T is the tensile strength and (ro/ri) is the wall thickness ratio. (This equation was recently derived9 by a theory of flow using the von Mises yield condition). Aircraft quality SAE 4340 steel is often used, which has a tensile strength (with appropriate heat treatment) of 180,000 psi (see for example the J.T. Ryerson and Son, Inc., Steel Data Booklet10). For a wall ratio of 8, such a pressure vessel is designed for failure at 360,000 psi.

It is significant that manufacturers will seldom guarantee extended performance at such pressures, although they do guarantee pressure vessels to be tested at these levels.

For pressures below 50,000 psi, closures which use Amagats' principle of packing may be used. According to this principle the packing is entirely surrounded by metal walls. Such closures leak when the pressure rises above the stress in the packing. Vessels with such closures are manufactured relatively cheaply.11

Type of Fracture
There are two types of rupture common to pressure vessels. These depend upon whether the steel is brittle or ductile. In the case of brittle steel, the cylinder fails along a radial plane. In the case of ductile steel, the vessel fails along an equiangular spiral, which is the plane of maximum shear, (see Figure 8a). The danger of flying fragments exists in both cases of rupture. In the first case, the rupture may suddenly change directions and proceed along a shear plane inclined at 90° to the first. In this even, a triangular prism may be detached and projected through the rupture crack with explosive violence. In the latter case, a fragment may be broken off the plane of rupture, and projected outward. (see Figure 8b). More often, however, the pressure vessel opens up and the pressure is slowly relieved, without expelling fragments.

Rupture is more likely to occur if the steel is brittle. This is due to the fact that in a ductile steel cylinder under internal pressure, an incipient rupture, once started may be checked by local flow.12

It is evident that the steel for pressure vessels must be carefully selected and prepared. It must be ductile, have a high tensile strength, and no pains must be spared to insure that it is free from imperfections. In this regard, the reputation of the supplier is important.

The Effect of Temperature and Time

It is known that the tensile strength of steel decreases as the temperature increases. In the design of experiments involving high temperatures, the effective decrease of tensile strength with temperature should always be checked. Curves showing strength vs. temperature for SAE 4340 steel do not appear to be published. However, there are curves for another chromium-nickel-molybdenum steel13 (see Figure 9). From these curves it appears that it is safe to operate up to about 800°F. However, more exact information on the SAE 4340 steel is needed.

The reduction in strength with time at high temperatures is due to metallurgical changes such as embrittlement, spheroidization, graphitization and migration of carbides. These effects are pronounced13 at 1000°F and at times of about 10,000 hours (see Figure 9). Further information is needed for experiments under these conditions.

The Closure in Pressure Vessels
So long as the "unsupported area" closure is functioning properly, it is as strong as other regions of the pressure vessel. An illustration of the closure is shown in Figure 10. The packing of the closure includes three rings, A and C of steel, and B of lead. The unsupported area is the entire back surface of C. High pressure pushes ring C out into the space E, which crowds it against the containing walls in such a way that the contact pressure is greater than the fluid pressure. This seal is not operative until the steel ring C begins to flow. The lead ring B provides tightness until the steel ring reaches a flow stress.

For proper functioning14, the following conditions must be maintained: (a) the rings must be of softer steel than the vessel, (b) the walls around E must be smooth, (c) the rings must be tightly fitted so that the lead ring B receives an initial compression, and (d) the steel rings must be replaced when extensive deformation has taken place. Although these conditions are easy to attain and are essential to avoid accidents which are due to blowouts of the closures. For this reason, strict supervision in the preparation of packing is required. Bridgman uses the same packing many times. Some pressure laboratories throw the packing away each time the closure is removed. We intend to use new packings each time.

Special Hazards in a Combined Hydraulic System

Requirements
There are some hazards which arise because hydraulic systems are coupled together. The need for coupled hydraulic systems arises when a high pressure system must have concurrent rather than subsequent tests in operation. For example, let us suppose that it is required to measure the relaxation time of some phenomenon under several sets of pressure and temperature. If the relaxation time is long — say one month — it becomes necessary to have concurrent different tests. In such a case, it is wasteful to have an individual pressure source for each pressure vessel, since a pressure source often costs more than one thousand dollars. The obvious answer is to have cone pressure source for several vessels, and this means coupling hydraulic systems.

A high pressure system should be versatile as well as economical. It should be possible to drain or activate any pressure vessel without disturbing the others. It should also be possible to withdraw any pressure vessel and its associated plumping from the system without disturbing the remainder.

Large reduction costs is possible whenever pressure levels can be reduced.

Protection from Hazards
It is necessary to design the following safety features: (a) adequate protection of personnel from the remaining pressure vessels while any one is being worked on, (b) prevention of fluid in a high pressure element of the system from suddenly surging to a lower pressure element, and (c) prevention of hot fluid in one part of the system from suddenly surging to elements at lower temperatures.

It is possible to discuss each of these safety devices in general terms. However, it is probably more illustrative to discuss the particular safety features incorporated in the high pressure system proposed by Dept. 1141.

Supporting Structure
In this proposed system there are six pressure vessels each of which is designed to withstand pressures and temperatures up to limits given by the following scheme.
 
Temperature
Pressure
25°C
100°C
300°C
10,000 psi
C
D
E
50,000 psi
A
 
B
150,000 psi
 
 
F
In order to prevent accidents due to the sudden surges from the one pressure level to the next, vessel F has its own pressure source, vessels C, D, and E have their own pressure source, and vessels A and B have their own pressure source.

In order to make each pressure vessel safely accessible, individual shielding was provided. This was accomplished by mounting each of the vessels in a fabricated steel table divided into six bays (see Figure 11). The pressure vessel is mounted on a plate with the plumbing on the lower side of the plate. The plate is lowered into a recess in the bay by a traveling hoist. The shielding consists of extruded cylindrical tubing or cylindrical annealing ovens which are lowered over the pressure vessels. The control valves are mounted on a skirt which is fastened to the front of the bay. The pressure sources (intensifiers) are mounted on the second shelf of the table.

Pressure Vessel and Associated System
Each pressure vessel with its associated plumbing back to the coupling (with the pressure source), is a standard and to some extent interchangeable unit. The schematic diagram of this sub-system is shown in Figure 12. The actual relationship of pressure vessel to oven, plumbing and plate is shown in Figure 13. Here we note that adequate space is provided between the oven and the pressure vessel in order to condense hot vapors which might accidentally escape. We note also that the pressure vessel which shields personnel from fragments of these parts flying upward in the event of an explosion. Rupture disks are designed to blow out at pressures about 20% above that designed for the sub-system. In case of a blowout, the liquid passes to an open drain.

In order to prevent accidents due to a surge of hot fluid from, say, vessel B to A, two valves are mounted between the pressure source and the sub-system compromising B and A, (see Figure 14). It is conceivable that someone may turn the wrong valve. However, in order to accidentally open a line between B and A would require the erroneous operation of four valves and sometimes six, which is hardly a likely occurrence.

Coupling
Because of the dangers of burns arising from flashes of hot vapor, it is not deemed advisable to make the intensifier portable. In this case, a coupling must be made in a high pressure line before and after a pressure vessel is activated. Alternatively, in the scheme shown in Figure 14, the coupling between the intensifier and the pressure vessel is fixed in a protected position. The operator is exposed to the coupling only when the system is being assembled or disassembled.
Pressure Liquid
An ideal pressure liquid has low compressibility, low thermal conductivity, a high flashpoint and boiling point, low viscosity, and good chemical stability. Apparently the best liquid meeting these requirements is a silicone oil. Although expensive, it should prove satisfactory for long-time operation in the vicinity of 300°C.

A pressure vessel should always be brought up to temperature before it is brought up to pressure.

Summary and Conclusions

The following safety devices, procedures, and designs concerning the operation of high pressure equipment are recommended:
  1. The Petry formula should be used for calculation of the thickness of barriers.
  2. If placed inside laboratories the equipment should not be used to conduct experiments involving (a) gases as pressure media, (b) testing to failure, or (c) testing under certain conditions.
  3. Barriers should be fastened in such a way that bolts cannot be subjected to sudden tensile stresses in the event of an explosion.
  4. The principle of the ballistic pendulum should be used in securing barriers.
  5. Piping and tubing should be shielded by outside piping or channel iron.
  6. Pressure indicating instruments should be indirect, using either pneumatic or strain gauge transmission.
  7. Valves should be turned by an indirect (preferably mechanical) method.
  8. Pressure vessels should be individually shielded by steel tubing, armored ovens, armored cooling jackets, or by the stanchions and platens of a press.
  9. Hydraulic fittings should be individually shielded just as pressure vessels.
  10. Adequate protection must be taken against burns resulting from vapor leaking from high pressure equipment (see Figure 7).
  11. Adequate protection must be taken against damage due to shock waves when the pressure media is a gas, or when there is danger of combustion.
  12. The wall thickness of a pressure vessel should be designed by using the formula on page 10.
  13. The steel used in pressure equipment must be ductile, have a high tensile strength, and be free from imperfections.
  14. Insufficient data are available on the advisability of operating pressure vessels above about 350°C.
  15. Insufficient data are available on the advisability of operating pressure vessels for longer than about 5,000 hours.
  16. Close supervision is required in the preparation of the packing in the closures of pressure vessels.
  17. New packings are required for each run of the pressure vessel.
  18. In combined pressure systems, it is necessary to assure adequate protection of personnel from remaining pressure vessels, while any one is being worked on and take: (a) measures to prevent fluid from surging from a high pressure region to a low pressure region, (b) measures to prevent hot fluid from surging to low temperature regions, and (c) measures to reduce the hazards of coupling systems.
  19. A pressure fluid should have low thermal conductivity, a high flashpoint and boiling point, low viscosity, and good chemical stability.
  20. A pressure vessel should always be brought up to temperature before it is brought up to pressure.

Footnotes

* By high pressure we mean pressures above levels generally found in ordinary hydraulic systems (say 3000 psi).
** Such an accident occurred at the Harwood Engineering Company in April, 1954. When the closure was removed, the operator was scalded by steam. Extensive hospitalization was required before the burns healed. The residual pressure was estimated at only 20 psi.
*** See Foxboro Bulletin #415
**** See the Harwood Engineering Company Catalog
***** The pressure laboratory of the Armour Research Foundation

Appendix A: Maximum Velocity of a Fragment

Assume that the compressed liquid is acting like a compressed spring. If a rupture of the vessel occurs, let us suppose that a fragment is suddenly projected outward with a force pa, where p is the pressure and a is the area of the fragment over which the pressure acts. The distance through which the force acts upon the fragment will be taken equal to the linear recovery of the spring (i.e., the liquid). Taking the relative change of volume at 0.25, and with a capacity of 10 cubic inches, the fluid will suddenly expand by 2.5 cu. in. Assuming the cross-sectional area of the fluid while expanding is one sq. in., the fluid will recover about 2.5 in. The average force acting over this distance is .

The velocity of a body starting from rest after traveling a distance d under the force pa is

where W is the weight. Taking pa»200,000 lb., and W=0.1 lb., we have
V»3,670 ft./sec.
We note that the velocity decreases as W-½. However, there is a limit to the velocity even for very small fragments. This limit arises because a fragment cannot have a velocity greater than the maximum velocity of the fluid leaving the fissure (neglecting shock waves). The pertinent equation for the velocity at an orifice is
where r is the density.

The pressure drops rapidly as soon as it is relieved. Assuming that the effective pressure is about half its instantaneous value we calculate for p = 200,000 / 2 psi, and r = 60 lb./cu.ft.,

V»3,920 ft./sec.
Because of the rapid pressure drop, a heavy fragment carried by the escaping fluid will be accelerated less than light fragments and will be thrown with smaller velocity.

It can be shown by calculating the drag of a fragment thrown through the air that its velocity will not be significantly reduced within a range of 100 ft. For calculations of barrier thickness, a velocity once attained by a fragment is maintained.

Appendix B: The Ballistic Pendulum

The ballistic pendulum was invented by Robins in 1740 to measure the velocity of projectiles. It consisted of a heavy wooden bob suspended from a tripod, arranged so as to receive the impact of a projectile and measure the resulting swing; from the velocity imparted to the bob, the striking velocity of the projectile was measured.

The height that the pendulum rises against gravity upon impact with a projectile can be calculated using the law of conservation of energy, provided one knows the ratio of the energy absorbed by the penetration of the projectile. Let a be the ratio of penetration energy to the total energy of the system, then:

,
where h is the height, w/W is the ratio of the weight of projectile to mass, and V is the velocity of the projectile.

In Appendix A, V² was calculated, which reduces the expression for the height to:

where the weight of the projectile has canceled out. For pa about 200,000 lbs., d about 2 inches, and W about 1 lb., it is necessary for a to be a very large number if h is to be less than one inch.

References

1. "The Effects of Atomic Weapons," U.S. Atomic Energy Commission, p. 87.
2. Ibid., p. 45.
3. "Design of Protective Structures," Navy Docks P-51, Bureau of Yards and Docks, Dept. of Navy, Washington, D.C., Aug., 1950.
4. "Civil Protection," Samuel, F.J. and Hamann, C.W., The Architectural Press, London, 1939.
5. Design of Protective Studies, ibid, p. 6 and Figure 4
6. Bridgman, P.W., "The Physics of High Pressures," G. Bell & Sons, Ltd., London (1949), Chapter V.
7. Brown, F.W., "Theoretical Calculations of Explosives, II Explosion Pressures." Technical Paper 643, Bureau of Mines, U.S. Government Printing Office, 1942.
8. Bridgman, ibid, pp. 81-82.
9. Faupel, J.H. and Furbeck, A.R., Trans. A.S.M.E., 75 (1953) p. 352.
10. J.T. Ryerson and Son, Inc., Steel Data Booklet.
11. See, for example, the "Superpressure" catalog of the American Instrument Co.
12. Bridgman, ibid., p. 83 and Figure 22.
13. "Steels for Elevated Temperature Service," adv-18566 (S), U.S. Steel pamphlet, 1949.
14. Bridgman, ibid, p. 33.

Afterword from "The Peril in Pressurized Liquids" by Lloyd M. Polentz of Richland, Washington

"High pressure testing with gases is dangerous. Energy stored in the compressed gas can turn a vessel into a bomb if the wall or fittings rupture. The hazards of high-pressure testing can be lessened if the gas is replaced by a liquid. But risk is not eliminated — proof testing with water or oil can also be dangerous. The energy contained in a liquid under pressure, due to compression of the liquid and the expansion of the vessel, can create extensive damage if the device should rupture. Here's how to calculate the energy contained in a liquid-pressurized system..."
Harwood Engineering Company, Incorporated – © July, 1999
455 South Street, Walpole, Massachusetts 02081
phone: 508-668-3600
fax: 508-660-2276
http://www.ultranet.com/~harwood/
email: harwood@ma.ultranet.com