A High Integrity, Simple Action Pipe Joint
THIS PAGE CONTAINS PROPRIETARY INFORMATION WHICH MUST NOT BE USED FOR COMPETITIVE PURPOSES OR IN ANY WAY DETRIMENTAL TO BETE FOG NOZZLE INC. ©BETE FOG NOZZLE, INC.
E.J. Nobles, BSc, MSc, PhD, CEng, MIMechE
Department of Mechanical Engineering,
University of Manchester Institute of Science and Technology
Flanged joints are the most common type of remarkable pipe joint and, given the correct gasket and bolt loading, their application is usually successful. However, there are certain circumstances arising out of operational considerations where an alternative jointing method is needed. External corrosion can make bolt removal difficult, gaskets can be over or under tightened by poor fitting practice, and the mass of the flange may be structurally undesirable. This paper and long term integrity are not sensitive to the skill of the maintenance personnel. The design concepts are first given leading to thte description of the coupling itself. To verify the structural integrity of the coupling, an experimental stress analysis of the device has been performed under combined loading: pressure, bending and shear force. An account of field experience concludes the paper.
Remakable pipe joints are widely recognised as a significant source of maintenance expenditure, particularly when pipelines have to be regularly broken to permit cleaning or inspection (1,2). Flanged joints are the most common type of remakeable pipe joint and, given the correct bolt loading, their application is usually successful. The make and break time for a flange largely consists of the time taken to tighten and release the bolts plus the time taken to clean up the flange faces ready for a new gasket. If there is extensive external corrosion then bolted connections can be very difficult to release and maintenance times can become excessive. Even without a corrosive environment flanged joints can give rise to difficulties. Gaskets can be over- or undertightened, both cases giving rise to leakage.
There are a wide range of remakeable pipe joints commercially available, ranging from “V” clamp designs to quick release couplings which are commonly used in hydraulic and pneumatic systems. The types that have found more common application in process plant applications use bolted connections (axial or tangential) to provide a force to grip the plane outside diameter of a pipe. Often, one finds resistance to the use of jointing methods which rely on a friction grip rather than on a positive location, especially if significant axial forces and/or bending moments are applied to a pipe since the seal may be affected.
Therefore, there is a need for a pipe coupling which will meet the following requirements: the coupling should have a simple action suitable for use in an internally and ,or externally corrosive environment, and should provide a high integrity joint under complex loading. This paper discusses the design of a pipe joint to meet these requirements, and presents the results of a test of the device under combined loading: pressure, bending and shear force.
2. DESCRIPTION OF THE PIPE COUPLING
2.1 Design principles
The Dur O Lok coupling, shown in figure 2, was originally designed by Hitz as a compact high pressure coupling for pipelines and pressure vessel closures (5). It uses a self-energizing seal with the result that it is not necessary to provide a high contact force between coupling halves to make the seal effective. This eliminates the greater part of the high compressive force which is a feature of flange design. Indeed, many problems experienced in flange usage can be traced to incorrect bolt loading due to bad fitting practice.
Consider the case of a 3 in class 1500 ANSI flange, see Fig. 3. The bolt load is applied at a moment am of 61mm (2.5”) relative to the mid thickness of the pipe wall. In order to minimise the distortion of the flange faces and to keep stresses to a safe level, a large flange thickness is specified. However, such large material sections create metallurgical, thermal transient and handling problems. Excessive stresses can be created in the bolts, flanges or gasket due to a combination of any of the following: initial bolt load, hydrostatic end load, thermal loading (6). A common case of flange failure is permanent distortion of the flange due to uneven bolt loading around the joint circumference which causes leakage. Correct bolt loading is therefore a pre-requisite to successful flange usage, and heavy sitff components are a feature of flange design.
In contrast, consider the design of a 3” schedule 80 Dur O Lok coupling which meets the same specification as the flanged joint given above. The axial location is provided by the interlocking of the split coupler with the multiple grooves, see fig. 2 The moment arm of the axial forces about the mid thickness of the pipe wall is 6mm (0.25”), see Fig. 4, which is one tenth that of the bolted flange. Thus, combined with complete circumferential distribution of the forces, distortion of the sealing faces of the coupling is not a problem and a smaller lighter coupling results. Figures 3 and 4 are drawn on the same scale for comparison.
The self-energized seal must be located in a dimensionally stable cavity if it is to be effective. The axial length between the loading faces either side of the seal is 10mm (0.4”). For a rating of 170 bar (2500 lbs/in2) at 340°C (644°F) the expansion of this dimension is 0.05mm (0.002”). Note that under the same conditions the bolt stretch in the equivalent ANSI flange is 0.5mm (0.02”).
The net effect of the design approach is a reduction in the normal flange diameter of 267mm (10.5”) to a coupling diameter of 124mm (4.9”) and a corresponding reduction in length from 254mm (10”) to 168mm (6.6”). The weight of the connection assembly for the 1500 ANSI flange is 544 N (122lbf) as compared to 55N (12.4lbf) for the steel Dur O Lok couplings.
2.2 Detail Design.
The detail design is illustrated in Fig. 2 as an exploded view, the assembled components are shown in Fig. 4. The hub section progressively thickens from the pipe wall thickness to its maximum to eliminate stress risers. A specific tooth and groove form has been adopted from that used in joining threads for oil well drilling work. The front angle of the tooth is designed below the angle of friction for all the metal involved. Thus, separating forces will not tend to separate the couplers even if the retainer ring is removed. The back of the tooth has been designed at an angel that assures a perfect tooth fit without possible wobble or lost motion in a situation of vibration. Note that the coupler sections are designed with a gap between the ends to further assure complete seating of the special tooth in the hub grooves. Tests of the engagement by blueing methods demonstrate over 90% of total area contact which is created by machining tolerances of 0.025mm (0.001”) to 0.05mm (0.002”).
The retainer ring has a small angle which forces the coupler sections into full engagement of tooth and groove. The effect of the two angles of the retainer ring and hub tooth working in unison provides a mechanical advantage of 150 to 1 for driving parts into total engagement. Anchoring set screws in the retainer ring engage a reverse taper. They are made of a non-corrosive stainless to insure against any possible back away of the retainer ring or thread jamming. The set screws are designed to be flush with the outside of the ring. It proves correct assembly of parts and full engagement of holding grooves to the craftsman and supervisor.
The seal is located in a groove machined into one face of the coupling. At lower temperature services a standard elastomer O-ring seal is used, whilst at higher temperatures a metal to metal (“omega”) seal is called for. For extremely high temperature service conditions an “over-centre” seal is specified.
2.3 Maintenance procedure.
Each half of the coupling, welding into the pipeline, is brought together approximately into the required position and the retaining ring slipped into the required position and the retaining ring slipped onto one half. The seal cavity and mating surface are checked for cleanliness, and the appropriate seal inserted. The fitted seal stands just proud of its cavity to ensure adequate compression for an initial leak free assembly. The hubs are then brought together into reasonable alignment. Manual or hydraulic tools can be used if necessary, location for which is provided by the two ledges. The split couplers are placed in position, they may be tapped into place. Finally, the retaining ring is slipped over the split couplers and lightly driven up flush with the coupling ledge using a soft hammer and the locking screws run up. This is not a difficult task. If the split couplers are incorrectly positioned then the retaining ring will not pass over them.
The above steps are all sequentially self -checking. In practice the coupling is simple to make and break, and tends to give one confidence that a secure joint has been made.
3. COUPLING INTEGRITY
3.1 Failure criterion
It is necessary to establish a failure criterion for the coupling before intiating a test procedure. A pipe coupling may be deemed to have failed when leakage begins. However, a significant leak in one application may not give cause for concern in another. For example, a leak rate of “parts per million” in an inert gas test would be inappropriate for a general purpose steam line. In this study, a significant leak was defined as one which could be detected visibly using water as the pressurising medium. This test was considered to be generally suitable for process plant applications.
Leaks may be caused by changes in dimensions as the coupling strains, or by the structural failure of a component. It is not possible to determine changes in the dimensions of the assembled coupling parts therefore “nearness to failure” in this respect cannot be identified, only the failure itself. However, the surface stress levels in the coupling components can be measured using electrical resistance strain gauges. Thus the maximum shear stress may be used to estimate how close components are to yielding (von Mises yield criterion).
3.2 Test Procedure
It was known that the coupling could withstand a simple hydrostatic test to 1600 bar (23500 lb/in2) without leaking (7). Leak detection was as described above. No components were observed to have suffered permanent deformation in that test, but no stress levels were measured.
Rarely do couplings suffer conditions of simple hydrostatic pressure in plant use. It is necessary to know how a coupling will behave when subjected to bending and shear forces. Typically these may arise if the coupling is subjected to an overhung load during maintenance (say if a section of pipe is broken), or when a pipeline shifts under service loading (temperature, pressure).
A test rig was constructed which could apply a bending moment and shear force to a pressurised coupling. Two lengths of 75 mm (3 in.) diameter steel pipe were joined by a Dur O Lok coupling, the ends were closed by ANSI 150 flanges through which the pipe was pressurised and vented. One half of this assembly was held and a point load applied near the opposite end, see fig. 5. Thus one length of pipe acted as a cantilever with an end load, and the Dur O Lok coupling was subjected to a bending moment and a shear force in addition to internal pressure.
Strain gauge rosettes were mounted at intervals around the circumference of the coupling: on the hub, on the ledge, and around the ring. An initial test was carried out in which the internal pressure was progressively increased to 24 bar (350 lb/in2) with no external loads applied. Strains were recorded as the pressure increased. The test assembly was then pressurised to 24 bar and the transverse end load progressively increased to 10 KN (2 ton). The 10 KN load gave an in-plane bending moment of 6.4 KNm (9400 lb ft.) at the centre of the coupling. The shear force is of course 10KN. Strains were recorded as the transverse load increased. All tests were carried out at room temperature.
The pressure test to 24 bar produced very little strain in the coupling (< 70 µ on all gauges) and no strain plot departed from linearity. No leakage was observed.
The combined loading of 24 bar internal pressure, 6.4 KNm applied moment and 10 KN shear force caused no leakage from the coupling. All parts of the coupling were re-usable, and no leakage was observed from repeat tests under these conditions even using the same elastomer seal. No departure from linearity was found in any strain plot. The strain induced in the coupling resulted from the external loading primarily. The largest stress was found on the compression side of the hub: the minimum principal stress was -214 MN/m2 (31,130 lb/in2) giving a maximum shear stress of 181 MN/m2 (26,000 lb/in2). The strain levels recorded on the ledge were an order of magnitude smaller, and those on the ring smaller still. The maximum shear stress found is clearly insufficient to cause yield.
As a comparison, the calculated maximum stress in a 75 mm (3 in.) pipe due to a bending moment of 6.4 KNm and an internal pressure of 24 bar is 408 MN/m2 (59400 lb/in2). The hoop stress due to internal pressure is 16 MN/m2, (2300 lb/in2). The lower stress levels found experimentally are due to the increased thickness of the coupling.
Under the combined loading of 24 bar internal pressure, 6.4 KNm in-plane bending moment and a 10 KN transverse load, the coupling proved leak tight and suffered no permanent deformation.
4. FIELD EXPERIENCE
The coupling has found application in a wide range of industries. Within the first two years of marketing, over 300 units have been placed in service and applications include the following.
In the refining industry, the smooth bore and close alignment is especially valued in soldis transport lines. Maintenance problems have been reduced in the chemical and other process industries, especially where joints have to be made and broken regularly for cleaning and inspection. Research and development companies have found in the coupling unique flexibility to solve many problems. Dur O Lok pipe couplings and closures are being increasingly specified by engineering contractors. The list of users in these categories includes many well known international companies.
To date, there has not been a single mechanical failure reported in any installation. Only a few problems hve surfaced with regard to seals but they have been solved as they appeared. This has led to improved operability and better definition of seal materials. Upon leakage of a seal, the leak has never been in the form of a jet and there has been no destruction of the mechanical structure. Also, experience has shown that adjustment is not required as a result of operating pressure or temeperature flutuations.
During the initial marketing period several improvements have been incorporated into the design, including additional advantages such as:
Development of an interchangeable elastomer and “metal to metal” omega seal. This simplifies dealer and maintenance stocking as only a single type of hub is needed.
Development of a design with interchangeable orifice and blank plate.
Development of a “clean out” fitting for refinery use in fired heater or serpentine coils.
Development of tell-tale device on the hub blocking the retainer ring to meet code requirements for lethal gases and liquids.
The Du O Lok coupling is a simple action high integrity pipe coupling which is suitable for use in a wide range of piping systems including hazardous duties. The coupling may be made and broken simply which reduces maintenance costs. When compared to a flanged joint for the same duty, it is easier to maintain, forms a good seal easily which is less sensitive to pressure-temperature fluctuations, is lighter and more compact, and is competitively priced.
Under a combined loading test of an internal pressure of 24 bar (350 lb/in2), an in-plane bending moment of 6.4 KNm (9400 lb ft.) coupling did not leak and did not suffer any permanent deformation. When subject to a simple hydrostatic test, the coupling will withstand a pressure of 1600 bar (23500 lb/in2) without leaking.
Report No. MD20767, Mond Division (Eng. Dept.), Proc. Of University/Industry Information Exchange, Runcorn, England July (1979).
SEWARD, D.N. Anomalies of boltings. Petroleum Review pp. 35-37 (1976).
COTTRELL, B. the split nut. Chartered Mechanical Engineer pp.73-76 Nov. (1978)
LEE, F.O.Y. A design study to reduce the maintenance costs of pipe joints, MSc Dissertation, UMIST (1980)
VAN TASSEL, D. H. and HITZ, G. L. Design of compact, high pressure couplings and closures for pressure vessels and piping. Pressure Vessel and Piping Conference, ASME, Orlando, Florida, June (1982)
ASME Boiler and Pressure Vessel Code. Section VIII, Division 1, (1980).
Test report No. 82-29763. Midstates Analytical Laboratories Inc. Tulsa, Oklahoma, U.S.A. (1982).
THIS PAGE CONTAINS PROPRIETARY INFORMATION WHICH MUST NOT BE USED FOR COMPETITIVE PURPOSES OR IN ANY WAY DETRIMENTAL TO BETE FOG NOZZLE INC. ©BETE FOG NOZZLE, INC.