US20080027693A1 - Method of designing blowout preventer seal using finite element analysis - Google Patents

Abstract

A method of manufacturing, certifying, and optimizing a seal for a blowout preventer. The method includes generating a finite element analysis seal model, smoothing the finite element analysis seal model, and analyzing a strain plot of the smoothed finite element analysis seal model based upon a displacement condition.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS
• This application claims the benefit of the following provisional applications under 35 U.S.C. 119(e): U.S. Provisional Patent Application Ser. No. 60/820,723 filed on Jul. 28, 2006; U.S. Provisional Patent Application Ser. No. 60/847,760 filed on Sep. 28, 2006; U.S. Provisional Patent Application Ser. No. 60/862,392 filed on Oct. 20, 2006; and U.S. Provisional Patent Application Ser. No. 60/912,809 filed on Apr. 19, 2007, all of which are incorporated by reference in their entirety herein.
BACKGROUND OF INVENTION
• 1. Field of the Invention
• Embodiments disclosed herein generally relate to blowout preventers used in the oil and gas industry. Specifically, embodiments selected relate to methods of designing and manufacturing seals for use in blowout preventers, in which the seals may include elastomer and a rigid material.
• 2. Background Art
• Well control is an important aspect of oil and gas exploration. When drilling a well, for example, safety devices must be put in place to prevent injury to personnel and damage to equipment resulting from unexpected events associated with the drilling activities.
• Drilling wells involves penetrating a variety of subsurface geologic structures, or “layers.” Occasionally, a wellbore will penetrate a layer having a formation pressure substantially higher than the pressure maintained in the wellbore. When this occurs, the well is said to have “taken a kick.” The pressure increase associated with a kick is generally produced by an influx of formation fluids (which may be a liquid, a gas, or a combination thereof) into the wellbore. The relatively high-pressure kick tends to propagate from a point of entry in the wellbore uphole (from a high-pressure region to a low-pressure region). If the kick is allowed to reach the surface, drilling fluid, well tools, and other drilling structures may be blown out of the wellbore. Such “blowouts” may result in catastrophic destruction of the drilling equipment (including, for example, the drilling rig) and substantial injury or death of rig personnel.
• Because of the risk of blowouts, devices known as blowout preventers are installed above the wellhead at the surface or on the sea floor in deep water drilling arrangements to effectively seal a wellbore until active measures can be taken to control the kick. Blowout preventers may be activated so that kicks are adequately controlled and “circulated out” of the system. There are several types of blowout preventers, the most common of which are annular blowout preventers (including spherical blowout preventers) and ram blowout preventers. Each of these types of blowout preventers will be discussed in more detail.
• Annular blowout preventers typically use large annular, rubber or elastomeric seals having metal inserts, which are referred to as “packing units.” The packing units may be activated within a blowout preventer to encapsulate drillpipe and well tools to completely seal an “annulus” between the pipe or tool and a wellbore. In situations where no drillpipe or well tools are present within the bore of the packing unit, the packing unit may be compressed such that its bore is entirely closed. As such, a completely closed packing unit of an annular blowout preventer acts like a shutoff valve. Typically, packing units seal about a drillpipe, in which the packing unit may be quickly compressed, either manually or by machine, to affect a seal thereabout to prevent well pressure from causing a blowout.
• An example of an annular blowout preventer having a packing unit is disclosed in U.S. Pat. No. 2,609,836, issued to Knox, assigned to the assignee of the present invention, and incorporated herein by reference in its entirety. The packing unit of Knox includes a plurality of metal inserts embedded in an elastomeric body, in which the metal inserts are completely bonded with the elastomeric body. The metal inserts are spaced apart in radial planes in a generally circular fashion extending from a central axis of the packing unit and the wellbore. The inserts provide structural support for the elastomeric body when the packing unit is radially compressed to seal against the well pressure. Upon compression of the packing unit about a drillpipe or upon itself, the elastomeric body is squeezed radially inward, causing the metal inserts to move radially inward as well.
• Referring now toFIG. 1, anannular blowout preventer101 including ahousing102 is shown.Annular blowout preventer101 has abore120 extending therethrough corresponding with awellbore103. Apacking unit105 is then disposed withinannular blowout preventer101 aboutbore120 andwellbore103.Packing unit105 includes an elastomericannular body107 and a plurality ofmetal inserts109.Metal inserts109 are disposed within elastomericannular body107 ofpacking unit105, which are distributed in a generally circular fashion and spaced apart in radial planes extending fromwellbore103. Further,packing unit105 includes a bore111 concentric withbore120 ofblowout preventer101.
• Annular blowout preventer101 is actuated by fluid pumped into opening113 of apiston chamber112. The fluid applies pressure to apiston117, which movespiston117 upward. Aspiston117 moves upward,piston117 translates force to packingunit105 through awedge face118. The force translated to packingunit105 fromwedge face118 is directed upward toward aremovable head119 ofannular blowout preventer101, and inward toward a central axis ofwellbore103 ofannular blowout preventer101. Becausepacking unit105 is retained againstremovable head119 ofannular blowout preventer101,packing unit105 does not displace upward from the force translated to packingunit105 frompiston117. However,packing unit105 does displace inward from the translated force, which compressespacking unit105 toward central axis ofwellbore103 of theannular blowout preventer101. In the event drillpipe is located withinbore120, with sufficient radial compression,packing unit105 will seal about the drillpipe into a “closed position.” The closed position is shown inFIG. 5. In the event a drillpipe is not present,packing unit105, with sufficient radial compression, will completely seal bore111.
• Annular blowout preventer101 goes through an analogous reverse movement when fluid is pumped into opening115 ofpiston chamber112, instead of opening113. The fluid translates downward force topiston117, such thatwedge face118 ofpiston117 allows thepacking unit105 to radially expand to an “open position.” The open position is shown inFIG. 4. Further,removable head119 ofannular blowout preventer101 enables access topacking unit105, such thatpacking unit105 may be serviced or changed if necessary.
• Referring now toFIGS. 2, 3A, and3B together,packing unit105 andmetal inserts109 used inannular blowout preventer101 are shown in more detail. InFIG. 2,packing unit105 includes an elastomericannular body107 and a plurality ofmetal inserts109.Metal inserts109 are distributed in a generally circular fashion and spaced apart in radial planes within elastomericannular body107 ofpacking unit105.FIGS. 3A and 3B show examples ofmetal inserts109 that may be disposed and embedded within elastomericannular body107 ofpacking unit105. Typically,metal inserts109 are embedded and completely bonded to elastomericannular body107 to provide a structural support forpacking unit105. The bond betweenannular body107 andmetal inserts109 restricts relative movement betweenannular body107 andinserts109, movement which is seen to cause failure of the elastomer within the elastomericannular body107. More discussion of the bonds between elastomeric bodies and metal inserts within a packing unit may be found in U.S. Pat. No. 5,851,013, issued to Simons, assigned to the assignee of the present invention, and incorporated herein by reference in its entirety.
• Referring now toFIGS. 4 and 5, an example of packingunit105 in the open position (FIG. 4) and closed position (FIG. 5) is shown. When in the open position, packingunit105 is relaxed and not compressed to seal aboutdrillpipe151 such that a gap is formed therebetween, allowing fluids to pass through the annulus. As shown inFIG. 5, when in the closed position, packingunit105 is compressed to seal aboutdrillpipe151, such that fluids are not allowed to pass through the annulus. Therefore, the blowout preventer may close thepacking unit105 to seal against wellbore pressure from the blowout originating below.
• Similarly, spherical blowout preventers use large, semi-spherical, elastomeric seals having metal inserts as packing units. Referring toFIG. 6, an example of aspherical blowout preventer301 disposed about awellbore axis103 is shown.FIG. 6 is taken from U.S. Pat. No. 3,667,721 (issued to Vujasinovic and incorporated by reference in its entirety). As such,spherical blowout preventer301 includes alower housing303 and anupper housing304 releasably fastened together with a plurality of bolts311, whereinhousing members303,304 may have a curved, spherical inner surface. Apacking unit305 is disposed withinspherical blowout preventer301 and typically includes a curved, elastomericannular body307 and a plurality of curved metal inserts309 corresponding to the curved, spherical inner surface ofhousing members303,304. Metal inserts309 are thus disposed withinannular body307 in a generally circular fashion and spaced apart in radial planes extending from a central axis ofwellbore103.
• Additionally, ram blowout preventers may also include elastomeric seals having metal inserts. The large seals are typically disposed on top of ram blocks or on a leading edge of ram blocks to provide a seal therebetween. Referring now toFIG. 7, aram blowout preventer701 including ahousing703, aram block705, and atop seal711 is shown. With respect toFIG. 7, only oneram block705 is shown; typically, though, two corresponding ram blocks705 are located on opposite sides of a wellbore103 from each other (shown inFIG. 8).Ram blowout preventer701 includes abore720 extending therethrough,bonnets707 secured tohousing703 and piston actuatedrods709, and is disposed about central axis of awellbore103.Rods709 are connected to ramblocks705 and may be actuated to displace inwards towardswellbore103. Rams blocks705 may either be pipe rams or variable bore rams, shear rams, or blind rams. Pipe and variable bore rams, when activated, move to engage and surround drillpipe and/or well tools to seal the wellbore. In contrast, shear rams engage and physically shear any wireline, drillpipe, and/or well tools inwellbore103, whereas blind ramsclose wellbore103 when no obstructions are present. More discussion of ram blowout preventers may be found in U.S. Pat. No. 6,554,247, issued to Berckenhoff, assigned to the assignee of the present invention, and incorporated herein by reference in its entirety.
• Referring now toFIG. 8, ram blocks705A,705B andtop seals711A,711B used inram blowout preventer701 are shown in more detail. As shown,top seals711A,711B are disposed withingrooves713 of ram blocks705A,705B, respectively, and seal between the top of ram blocks705 and housing703 (shown inFIG. 7). As depicted,ram block705A is an upper shear ram block havingtop seal705A, and ram block705B is a lower shear ram block having top seal705B. When activated, ram blocks705A,705B move to engage, in which shears715A engage above shears715B to physically sheardrillpipe151. As ram blocks705A,705B move,top seals705A,705B seal againsthousing703 to prevent any pressure or flow leaking betweenhousing703 andram blocks705A,705B.
• Referring now toFIGS. 9A and 9B,top seals711A,711B are shown in more detail. As shown particularly inFIG. 9A,top seals711A,711B comprise anelastomeric band751,elastomeric segments753 attached at each end ofelastomeric band751, and ametal insert755 disposed within eachelastomeric segment753.Top seal705A forram block705A (i.e., the upper shear ram block) may also include asupport structure757 connected betweenelastomeric segments753. As shown in a cross-sectional view inFIG. 9B,metal insert755 disposed withinelastomeric segment753 has an H-shaped cross-section. The H-shaped cross-section ofmetal insert755 provides support and optimal stiffness toelastomeric segment753. Furthermore, it should be understood thattop seals711A,711B may be used with either pipe rams, blind rams, or shear rams (shown inFIG. 8).
• Referring now toFIG. 10, aram block705A with a top seal and aram packer717A used in ram blowout preventer (e.g.,701 ofFIG. 7) are shown.FIG. 10 is taken from U.S. Publication No. US 2004/0066003 A1 (issued to Griffen et al. and incorporated herein by reference in its entirety). Instead of a shear rams (shown inFIGS. 7 and 8),FIG. 10 depicts a pipe ram assembly having a variablebore ram packer717A comprised of elastomer and metal. As shown, variablebore ram packer717A comprises anelastomeric body761 of a semi-elliptical shape having metal packer inserts763 molded inelastomeric body761. Metal packer inserts763 are arranged around abore765 ofelastomeric body761. As mentioned above with respect to pipe rams or variable bore rams, when activated, rampacker717A (along with a corresponding ram packer oppositely located fromram packer717A) moves to engage and surround drillpipe and/or well tools located inbore765 to seal the wellbore.
• For any seal mechanism comprising elastomers and metal in blowout preventers (e.g., packing units in the annular and spherical blowout preventers and top seals and ram packers in the ram blowout preventer), loads may be applied to contain pressures between various elements of the blowout preventers. For example, with respect to the annular blowout preventer shown inFIG. 1, as the fluid force is translated frompiston117 andwedge face118 to packingunit105 to close packingunit105 towards central axis ofwellbore103, the fluid force generates stress and strain withinpacking unit105 at areas and volumes thereof contacting sealing surfaces (e.g.,wedge face117 and drillpipe151) to seal against wellbore pressure from below. The stress occurring in packingunit105 is approximately proportional to the fluid force translated to packingunit105.
• As stress is incurred by blowout preventer seals, the material of the seals will strain to accommodate the stress and provide sealing engagement. The amount of strain occurring in the material of the seal is dependent on a modulus of elasticity of the material. The modulus of elasticity is a measure of the ratio between stress and strain and may be described as a material's tendency to deform when force or pressure is applied thereto. For example, a material with a high modulus of elasticity will undergo less strain than a material with a low modulus of elasticity for any given stress. Of the materials used in blowout preventer seals, the metal inserts have substantially larger moduli of elasticity than the elastomeric portions. For example, the modulus of elasticity for steel (typically about 30,000,000 psi; 200 GPa) is approximately 20,000-30,000 times larger than the moduli of elasticity for most elastomers (typically about 1,500 psi; 0.01 GPa).
• Historically, when examining, designing, and manufacturing seals for blowout preventers, such as packing units for blowout preventers, the locations and amounts of stress and/or strain (i.e., stress concentrations, strain concentrations) occurring within the seal have been the largest concern and received the most attention and analysis. As the seal is subject to loads (e.g., repetitive and cyclic closures of a packing unit of an annular blowout preventer about a drillpipe or about itself), the magnitude and directions of the stresses and strains occurring across the seal are evaluated to determine the performance of the seal. A common technique used for this evaluation is finite element analysis (“FEA”). Specifically, the FEA may be used to simulate and evaluate the stress and/or strain concentrations which occur across the seal under given displacement conditions.
• Traditionally with FEA, seals for blowout preventers are modeled with finite elements to determine the performance of the seal under various displacement conditions. For example, using FEA modeling, the packing unit of an annular blowout preventer may be simulated with a displacement condition to move into the closed position around a drillpipe, in which the packing unit would be compressed between the piston and the removable head from the annular blowout preventer and the drillpipe. The FEA model may be used to produce a strain plot of the seal (packing unit in this example) to display the strain concentrations within the seal under that specific displacement condition.
• However, this evaluation of the strain concentrations may not result in the most accurate prediction and representation of the performance of the seals used in blowout preventers. Typically, the seals used in blowout preventers experience extremely high amounts of strain from the stresses that may be incurred. For example, when a packing unit is compressed into the closed position to seal about a section of drillpipe, an elastomeric body of the packing unit may experience strains in excess of 300% in the areas of the strain concentrations. Further, in a case where no drillpipe is present, the packing unit may begin experiencing strains of about 400-450% in seating about itself These elevated strains, especially when repetitively and cyclically performed upon the packing seal, usually lead to the ultimate failure of the seal.
• Furthermore, as described above, the metal and elastomers used for seals in blowout preventers typically have large differences in their moduli of elasticity. Because of this difference between the moduli of elasticity, when bonded together, the metal will tend to control the “flow” and deformation of the elastomers in the seals when compressed in the blowout preventers. With the large amounts of strain, especially the strain resulting from repetitive and cyclic displacements, coupled with the significant difference between the moduli of elasticity of the seal's materials, FEA evaluating strain concentrations may not accurately represent the capabilities of the seals.
• In FEA applications, the seal comprising a rigid material and elastomer may be represented by a geometrically similar representation consisting of many finite elements (i.e. discrete regions), commonly referred to as a mesh. The finite elements interact with one another to model the seal and provide simulated data and results for various displacement conditions. However, the finite elements within areas of high stress and/or strain (i.e., stress and/or concentrations) with substantial differences between materials' moduli of elasticity may improperly deform. Common improper deformations of the finite elements that may occur include the elements collapsing upon themselves, distorting without bound, or sustaining losses in stress, strain, and/or energy. These, in addition to other improper deformations of the finite elements may produce inaccurate results for the stress and strain occurring across the model.
• Historically, when the FEA is producing erroneous results, the number of finite elements of the mesh is increased for better resolution in at least some selected locations (e.g., areas of high stress or strain concentration). Thus, it is common for areas with stress and/or strain concentrations to receive more localized “attention” when modeling in FEA than other areas. However, this process may allow the analysis to become inherently localized on the areas of the seal models with the stress and/or strain concentrations, leading to solutions that may be narrowly constructed and/or inaccurate. For example, it is common PEA practice to increase the number of elements of (and thus further complicate) the seal model in the areas of these concentrations to increase the accuracy of the simulated stress and strain within the concentration regions. The same may also be done for a seal model in the areas of strain concentrations. However, it should be understood that by increasing the number of elements, or decreasing the mesh size, the solution time and the amount of computing power required may be increased. This may lead to solution stalling (due to computational error) and/or the generation of inaccurate results.
• Referring now toFIG. 11, a graph displaying strain (y-axis) versus number of iterations (x-axis) within FEA is shown. Specifically, the simulated strain displayed on the y-axis may be a magnitude of a principal strain occurring in a specific direction simulated across a finite element of a seal model for a given displacement condition. For example, those having ordinary skill in the art will recognize that the y-axis of the graph may display the magnitude of a principal strain (e.g., strain occurring in the direction of the z-axis; shear strain occurring in the plane of the y-axis and the z-axis) occurring within a finite element when the seal model is simulated with a displacement condition (e.g., closing of a packing unit about a drillpipe). Further, the number of iterations displayed on the x-axis refers to the amount of simulations of FEA used when modeling the seal. As such, each “iteration” refers to a single execution of the FEA process to simulate a displacement of the seal for the blowout preventer, thus determining the magnitude of strain of the finite element of the seal model.
• In this approach, the resolution of the finite elements in the mesh (seal model) is increased with each iteration. Specifically, as mentioned above, it is ordinary practice to increase the resolution of the finite elements of the mesh in regions that experience large amounts of stress and/or strain. However, because of the characteristics of metal reinforced elastomer seals, such localized analysis may result in an FEA stress and/or strain output that fails to correlate to an experimentally observed solution. Furthermore, because of the complexity, the FEA stress and/or strain output may not even be capable of converging to a solution at all.
• As shown, theoretical strain of the finite element occurring in the direction of the simulated principal strain from the y-axis inFIG. 11 is determined and shown for a seal of a blowout preventer under the displacement condition. As the number of iterations increases for the FEA model, the simulated strain solution produced (i.e., a trend line of strain points found from each iteration produced using FEA) thereby may not correspond and converge with the theoretical strain under a comparable displacement condition. A tolerance band of ± about 1% of the theoretical strain is shown to indicate a range that may be acceptable for the simulated strain solution to converge within. This concept of convergence of FEA stress and/or strain output may be understood as when the simulated stress/strain solution reaches a solution within the tolerance band, the simulated stress/strain solution continues to stay within the tolerance band as further iterations of the solution are continued.
• Therefore, as shown, when designing and manufacturing high strain elastomeric seals containing rigid inserts, there may be a significant discrepancy between the theoretical stress and strain predicted by FEA and actual stress and strain. Thus, current modeling and analysis techniques for blowout preventer seals may not provide adequate information to improve their design and manufacture.
SUMMARY OF INVENTION
• In one aspect, embodiments disclosed herein relate to a method of manufacturing a seal of a blowout preventer. The method comprises selecting a seal design, generating a first finite element analysis seal model from the selected seal design, smoothing the first finite element analysis seal model, analyzing a strain plot of the smoothed first finite element analysis seal model based on a displacement condition, and manufacturing a seal.
• In another aspect, embodiments disclosed herein relate to a method to certify a seal of a blowout preventer. The method comprises generating a first finite element analysis seal model, smoothing the first finite element analysis seal model, analyzing a strain plot of the smoothed first finite element analysis seal model based upon a displacement condition, and comparing the strain plot of the smoothed first finite element analysis seal model against at least one specified criteria.
• Further, in another aspect, embodiments disclosed herein relate to a method of optimizing a seal of a blowout preventer. The method comprises smoothing a first finite element analysis seal model, analyzing a strain plot of the smoothed first finite element analysis seal model based upon a displacement condition, generating a second finite element analysis seal model based on the analyzed strain plot of the smoothed first finite element analysis seal model, smoothing the second finite element analysis seal model, analyzing a strain plot of the second smoothed finite element analysis seal model based upon a displacement condition, and repeating the analyzing and generating of smoothed finite element analysis seal models until an optimized seal model is reached.
• Other aspects and advantages of the embodiments disclosed herein will be apparent from the following description and the appended claims.
BRIEF DESCRIPTION OF DRAWINGS
• FIG. 1 is a cross-sectional view of an annular blowout preventer.
• FIG. 2 is a cross-sectional view of a packing unit for an annular blowout preventer.
• FIG. 3A is a perspective view of a metal insert for a packing unit for an annular blowout preventer.
• FIG. 3B is a side view of an alternative metal insert for a packing unit for an annular blowout preventer.
• FIG. 4 is a cross-sectional view of a prior art packing unit for an annular blowout preventer shown in a relaxed position.
• FIG. 5 is a cross-sectional view of a packing unit for an annular blowout preventer in a closed position.
• FIG. 6 is a cross-sectional view of a spherical blowout preventer.
• FIG. 7 is a cross-sectional view of a ram blowout preventer.
• FIG. 8 is a perspective view of ram shears for a ram blowout preventer.
• FIG. 9A is a perspective view of a top seal for ram blocks of a ram blowout preventer.
• FIG. 9B is a cross-sectional view of a top seal for ram blocks of a ram blowout preventer.
• FIG. 10 is a perspective view of a variable bore ram packer for a ram block of a ram blowout preventer.
• FIG. 11 is a graphical representation of strain versus the number of FEA iterations.
• FIG. 12 is a flow chart depicting a method of manufacturing a seal for a blowout preventer in accordance with embodiments disclosed herein.
• FIG. 13 is a cross-sectional, axial profile of an annular packing unit in a two-dimensional plot (using x and z axes) in accordance with embodiments disclosed herein.
• FIG. 14 is a cross-sectional, radial profile of an annular packing unit in a two-dimensional plot (using x and y axes) in accordance with embodiments disclosed herein.
• FIG. 15 is a portion of a seal model of an annular packing unit in a three-dimensional plot (using x, y, and z axes) in accordance with embodiments disclosed herein.
• FIG. 16 is a portion of a seal mesh of an annular packing unit in a three-dimensional plot (using x, y, and z axes) in accordance with embodiments disclosed herein.
• FIG. 17A is an end view of a metal insert for a packing unit for an annular blowout preventer.
• FIG. 17B is an end view of a metal insert for a packing unit for an annular blowout preventer in accordance with embodiments disclosed herein.
• FIG. 18A is a top view of a metal insert for a packing unit for an annular blowout preventer.
• FIG. 18B is a top view of a metal insert for a packing unit for an annular blowout preventer.
• FIG. 19A is a strain plot of a seal model of an annular packing unit in accordance with embodiments disclosed herein.
• FIG. 19B is a strain plot of a seal model of an annular packing unit in accordance with embodiments disclosed herein.
• FIG. 20A is a strain plot of a seal model of an annular packing unit in accordance with embodiments disclosed herein.
• FIG. 20B is a strain plot of a seal model of an annular packing unit in accordance with embodiments disclosed herein.
• FIG. 21A is a strain plot of a seal model of an annular packing unit in accordance with embodiments disclosed herein.
• FIG. 21B is a strain plot of a seal model of an annular packing unit in accordance with embodiments disclosed herein.
• FIG. 22 is a graphical representation of strain versus number of FEA iterations in accordance with embodiments disclosed herein.
• FIG. 23A is a strain plot of a seal model of an annular packing unit with selective de-bonding in accordance with embodiments disclosed herein.
• FIG. 23B is a strain plot of a seal model of an annular packing unit with selective de-bonding in accordance with embodiments disclosed herein.
• FIG. 24A is a strain plot of a seal model of an annular packing unit with selective de-bonding in accordance with embodiments disclosed herein.
• FIG. 24B is a strain plot of a seal model of an annular packing unit with selective de-bonding in accordance with embodiments disclosed herein.
• FIG. 25A is a strain plot of a seal model of an annular packing unit with selective de-bonding in accordance with embodiments disclosed herein.
• FIG. 25B is a strain plot of a seal model of an annular packing unit with selective de-bonding in accordance with embodiments disclosed herein.
• FIG. 26 depicts a computer system used to design seals for blowout preventers in accordance with embodiments disclosed herein.
• FIG. 27A is a strain plot of a seal model of an annular packing unit in accordance with embodiments disclosed herein.
• FIG. 27B is a strain plot of a seal model of an annular packing unit in accordance with embodiments disclosed herein.
• FIG. 28 is a seal model of an annular packing unit in accordance with embodiment disclosed herein.
DETAILED DESCRIPTION
• In one aspect, embodiments disclosed herein relate to a method of manufacturing a seal for a blowout preventer. In another aspect, embodiments disclosed herein relate to a method of optimizing a seal for a blowout preventer that incorporates using a strain plot in the method. In another aspect, embodiments disclosed herein relate to a method of certifying a seal model for a blowout preventer using FEA to produce a strain plot after the model has been smoothed and bulk analyzed in response to a displacement condition.
• As used herein, a “rigid material” refers to any material that may provide structure to a seal of a blowout preventer, both metal and non-metal. Examples for a rigid material may include, but are not limited to, steel, bronze, and high strength composites (e.g., carbon composites, epoxy composites, thermoplastics). Further, as used herein, a “seal” refers to a device that is capable of separating zones of high pressure from zones of low pressure. Examples of blowout preventer seals include, but are not limited to, annular packing units, top seals, and variable bore rams.
• As mentioned above, techniques and models historically used to design and manufacture seals having elastomer and rigid materials for blowout preventers may not provide accurate information to improve the performance of the seal's design. Therefore, in designing, manufacturing, and certifying a seal for a blowout preventer in accordance with embodiments disclosed herein, a method including FEA of bulk strain and generating a strain plot may be used to yield more accurate convergent results under a given displacement condition. This FEA method, in addition to certain techniques for generating and modifying the seal models, may more accurately calculate the strain in the seal because it is tailored to accommodate the large amounts of stress and strain experienced by blowout preventer seals. Suitable software to perform such FEA includes, but is not limited to, ABAQUS (available from ABAQUS, Inc.), MARC (available from MSC Software Corporation), and ANSYS (available from ANSYS, Inc.).
• Specifically, embodiments and methods disclosed herein may advantageously provide techniques for generating and analyzing seal models within FEA to determine the seal's response under displacement conditions characterized by large amounts of strain. Methods disclosed herein may use a simplified seal design and/or model of a seal to assist in the analysis of the seal. For example, methods disclosed herein may avoid analyzing stress and strain concentrations of a complex seal design by “smoothing” that design.
• As used herein, the term “smoothing” refers to various techniques to simplify a complex geometry of a seal design for use with FEA. These techniques may allow the analysis of a smoothed model (i.e., a FEA model constructed from a smoothed design) to correlate with experimentally observed conditions and to converge to a definitive result when analysis of a non-smoothed model may not. As such, a model constructed from a smoothed design may be analyzed within FEA to determine an overall, or “bulk”, strain condition. By analyzing this bulk (i.e., non-localized) strain, the performance, and/or possibly failure, of a seal under various displacement conditions may be predicted with more accuracy. Following the analysis of the smoothed model for the bulk strain condition, knowledge obtained therefrom may be incorporated into a (non-smoothed) seal design that is to be manufactured.
• Referring now toFIG. 12, a flow chart depicting a method of manufacturing a seal including an elastomer and a rigid material is shown. As afirst step1210, properties of the seal's materials (e.g., the elastomers and the rigid materials) are determined. The material properties may either be determined through empirical testing or, in the alternative, may be provided from commercially available material properties data. Next, a three-dimensional seal model (i.e., a mesh) for the seal is generated1220. As such, generating aseal model1220 may also comprise importing a seal design1221 and subsequently smoothing the importedseal design1222 to simplify FEA analysis.
• Next, displacement conditions are simulated in FEA using the smoothedseal model1230. Preferably, these simulated displacement conditions reflect the forces, load states, or strains that the seal may expect to experience in operation. Further, after simulating displacement conditions, a strain plot showing the strain and deformation occurring in the seal model is generated and analyzed1240. Ideally, the strain plot shows the location and amount of strain occurring in the seal model in response to the simulated displacement conditions. The strain plot may be analyzed and reviewed1240 to determine the performance characteristics of the seal model. If the seal model requires improvement, the method may loop back to1210 to determine material properties of another material for the seal, or alternatively may loop back to1220 for generation and analysis of another seal model. This loop allows the seal model to be further simulated in FEA to determine its performance after further modifications or models. Otherwise, if the seal model is considered acceptable and meets a specified criteria, the seal model may be used to manufacture a seal for a blowout preventer1250.
• Ininitial step1210, the properties of the seal's materials are determined. Of the materials, the elastomeric materials will have lower moduli of elasticity than the rigid materials. Thus, when the seal is subjected to large amounts of stress, the elastomeric portion of the seal will strain more than the rigid material portions. For example, when the packing unit in an annular blowout preventer is stressed in the closed position, the elastomeric body of the packing unit will strain significantly more than the metal inserts. Because elastomers strain significantly more than the rigid materials for any given stress input, it may be especially important to determine the material properties of an elastomer used in the seal, specifically the relationship between stress and strain across the elastomer.
• In viscoelastic materials under constant stress, the strain may increase with time (i.e., creep). Conversely, under a constant level of strain, the stress within viscoelastic materials decreases over time (i.e., relaxation). Furthermore, higher levels of strain and lower temperatures may lead to an increase in the moduli of elasticity for viscoelastic materials. Elongation of a material refers to the percentage change in length of a material. The maximum amount of tensile strain to which a material may be subjected, or elongated to, before failure is referred to as the elongation at break. A material may have a high or low modulus of elasticity, but may exhibit a low elongation at break such that the material will fail without experiencing much strain. Further, the tensile strength of a material is the maximum amount of stress (in tension) a material may be subjected to before failure. As stress is exerted upon the material, the material strains to accommodate the stress. Once the stress is too much for the material, it will no longer be able to strain, and the material fails. The failure point of the material is known as the ultimate tensile strength.
• Furthermore, if cyclic displacements are applied to an elastomeric material, hysteresis (phase lag) may occur, leading to a dissipation of mechanical energy within the elastomeric material. Hysteresis may occur when there is softening induced by stress. This may be described as an instantaneous and irreversible softening for a material that occurs when an applied displacement increases beyond any prior maximum value, resulting in a shift of the stress-strain curve of the material. This induced softening, which may also be referred to as the Mullin's effect, is thought to be at least partially attributed to the microscopic breakage of links in a elastomeric material. This weakens the elastomeric material during an initial deformation so that the material is, in turn, weaker in subsequent deformations of the material.
• Thus, in one embodiment of the present disclosure, to determine at least one of the material properties of the elastomer for the seal of the blowout preventer, as described above, empirical testing of the elastomer may be used. Specifically, tests may be performed to determine the properties of the elastomeric material. Examples of tests that may be performed include, but are not limited to, a uniaxial tension test, a uniaxial compression test, a lap shear test, and a biaxial tension test. A uniaxial tension test applies tensile load in one direction to a material and measures the corresponding strain induced in the material. A uniaxial compression test applies compressive load in one direction to a material and measures the corresponding strain induced in the material. A lap shear test applies shear loads to a material and measures the corresponding shear strain of the material. Further, a biaxial tension test applies tensile loads in two directions to a material and measures the corresponding strain of the material. The use of these tests, in addition to other tests commonly known in the art, may assist in analyzing and determining the material properties of the elastomer. Furthermore, it should be understood by one of ordinary skill in the art, that as material properties of most materials vary by temperature, the performance of multiple tests at differing temperatures may be prudent to establish certain material properties.
• Instep1220, a model (i.e., a mesh) for the seal is generated. When generating the model of the seal, design features of the seal are chosen and applied to the model. For example, for a packing unit for an annular blowout preventer, the number of inserts used, the width of the rigid material inserts, and the specific material used for the rigid material inserts may be chosen when generating the seal model. The seal models may be created in a computer aided design (“CAD”) software package (e.g., AutoCAD available from Autodesk, Inc., and Pro/Engineer available from Parametric Technology Corporation) and imported into the FEA software package or, in the alternative, may be generated within the FEA packages (e.g., ABAQUS and PATRAN) themselves.
• Referring now toFIGS. 13-16, a method of generating a seal model in accordance with embodiments disclosed herein is shown. Specifically, as shown, a model ofpacking unit105 of an annular blowout preventer may be generated from a seal design created using CAD software. As shown inFIG. 13, cross-sectional,axial profiles1301 of a seal design may be generated ofannular packing unit105 in a two-dimensional plot (using x and z axes).Packing unit105 includeselastomeric body107 and rigid (e.g., metal)material insert109 with bore111. Multiple radial and axial cross-sectional profiles may be generated to represent different sections of the seal. For example, profiles may be generated of the sections of apacking unit105 that do or do not have metal inserts109.
• From here, as shown inFIG. 14, in addition to generating cross-sectional,axial profiles1301, cross-sectional,radial profiles1401 of the seal design may be generated to represent different radial sections of the seal in a two-dimensional plot (using x and y axes). Because of the symmetry ofpacking unit105, only a radial portion of cross-sectional,radial profiles1401, as shown, may need to be generated. Then, as shown inFIG. 15, by combining axial andradial profiles1301,1401, a three-dimensional seal design1501 may be generated to represent at least a portion ofpacking unit105 in a three-dimensional plot (using corresponding x, y, and z axes fromFIGS. 13 and 14). In three-dimensional seal design1501, metal inserts109 andelastomeric body107 are generated as separate bodies which may interact with one another. Depending on the complexity of the design of the seal (i.e., packing unit in this case),more profiles1301,1401 of the seal may be generated for more detail inseal design1501.
• Further, as shown,seal design1501 and model or mesh1601 (discussed below) may only represent a radial portion ofpacking unit105. However, the remainder of packingunit105 may be easily generated by taking advantage of the symmetrical geometry ofpacking unit105. Those having ordinary skill in the art will appreciate that in the case of radially symmetric models, symmetric portions and profiles may be used and replicated to simplify the generation of the model.
• Referring now toFIG. 16,seal design1501 created using CAD software may be imported into FEA software to generate a model ormesh1601 of numerousfinite elements1603.Finite elements1603 ofmesh1601 work together to simulate a seal and a packing unit when stresses and forces are applied.Finite elements1603 ofelastomeric body107 of packingunit105 will simulate and respond to stress and forces (i.e., they will exhibit strain) corresponding to the material properties of the elastomeric material.
• Similarly,finite elements1603 of metal inserts109 of packingunit105 will simulate and respond to stress and forces corresponding to the material properties of the metal inserts. Thus,finite elements1603 deform and strain to simulate the response of the different materials (e.g., elastomers and rigid materials) of the seal in accordance with their material properties. Whilefinite elements1603 are shown as eight-noded elements (i.e., brick elements), finite elements of any shape known in the art may be used.
• Further, while generating aseal model1220, a number of smoothing techniques may be used on theseal design1222. In many circumstances, as mentioned above, analyzing the actual manufactured geometry of the seal using FEA may lead to complications when large amounts of stress and strain are simulated. Particularly, as manufactured, the geometry of metal seal components include radiused corners and other stress-concentration reducing features to more evenly distribute stress across the component as it is loaded. However, it has been discovered that these techniques may adversely affect FEA models in FEA in that they increase the complexity of the model and may prevent the FEA from producing accurate results. Therefore, a seal model generated from a smoothed design may include removing as-manufactured stress concentration features in an effort to improve the results of FEA.
• In one embodiment, the seal design's rigid material may be modified (i.e., smoothed) to reduce their complexity. Referring now toFIG. 17A, an end view of ametal insert1701 includingflanges1703 connected by aweb1705 is shown.Metal insert1701 typically includes radiusedinternal corners1707 and squaredexternal corners1709. However, in one embodiment of smoothing a design, the corners of the metal insert may be modified. For example, referring now toFIG. 17B, an end view of ametal insert1711design including flanges1713 connected by aweb1715 in accordance with embodiments disclosed herein is shown. In smoothing the design, internal corners1717 may be modified to reduce or eliminate their radii (as shown) in an attempt to simplify a subsequently constructed model. Further, in smoothing the seal design,external corners1719 may be modified to add or increase their radii (also shown) in an attempt to simplify a subsequently constructed model. A seal model constructed in this manner may be analyzed for bulk strains such that the FEA may produce more accurate and definitive results than would be possible using the former, more “localized” approach.
• Furthermore, in another embodiment, instead of smoothing the design by modifying internal and external corners of the rigid material insert, the smoothing may include modifying the shape of the rigid material insert and its position within the elastomeric body. Referring now toFIG. 18A, a top view of ametal insert1801 disposed within a portion of anelastomeric body1802 of an annular packing unit is shown.Flange1803 and web1805 (outline shown) ofmetal insert1801 shown has a rectangular outline, in which flange ends1804A,1804B offlange1803 and web ends1806A,1806B ofweb1805 are defined by straight edges.Ends1804A,1806A are radially closer tocentral axis103 than ends1804B,1806B.
• However, referring toFIG. 18B, the shape and orientation of the metal insert may be smoothed for bulk strain analysis. InFIG. 18B, a top view of ametal insert1811 disposed within a portion of anelastomeric body1802 of an annular packing unit in accordance with embodiments disclosed herein is shown. As shown, flange1813 and web1815 (outline shown) ofmetal insert1811 have arcuate ends to define a radial outline centered aboutcentral axis103. Specifically, sides1814C,1814D of flange1813 may follow alongradial lines1817 extending radially out fromcentral axis103. Sides1816C,1816D ofweb1815 may similarly follow along radial lines (not shown). With this, flange ends1814A,1814B disposed between flange sides1814C,1814D and web ends1816A,1816B disposed between web sides1816C,1816D may then follow an arcuate path to have an arc, bow, or bend, as shown. Preferably, arcuate ends1814A,1814B,1816A,1816B followradial paths1818 defined aboutcentral axis103. Thus, as shown, a width of flange1813 andweb1815 increases when following along their sides1814C,1814D,1816C,1816D from ends1814A,1816A to ends1814B,1816B. As such, a seal model constructed in this manner may be able to more accurately simulate strain during FEA to produce more accurate and definitive results.
• Further still, the elastomeric body of the seal design may be smoothed as well. Referring again toFIG. 15,elastomeric body107 includes acompression face108 corresponding to wedgeface118 of piston (117 ofFIG. 1). Whenpiston117 is activated, wedge face118 contacts and compresses packingunit105 to seal the well. In one technique, the seal design may be smoothed by modifying the compression face to have approximately the same angle as the wedge face of the piston. Alternatively, the wedge and compression faces may be modified to increase a contact region therebetween. By modifying the compression face, the wedge face, or both, a seal model constructed therefrom may be able to more accurately simulate strain for the strain plots during FEA. As the compressive face of the elastomeric body would otherwise have a different angle than the wedge face of the piston, the output of the FEA may be simplified to produce more accurate or definitive results when displaced.
• Those having ordinary skill in the art will appreciate that, in addition to these described smoothing techniques and modifications, other techniques may be used as well in addition. For example, in another embodiment, the web of the rigid material insert may be modified, such as hollowing the web of the insert, as long as the rigid material insert provides sufficient structural support for the seal to sustain the forces applied thereto when under any and all displacement conditions.
• Preferably, when generating the seal models instep1220, especially when smoothing theseal design1222 of the seal model, the volume of the elastomeric body and the rigid material inserts of the seal model remains substantially constant. If the volume does not remain constant, the results and simulated strain from the strain plots created by the FEA may not be accurate or consistent. For example, when applying a force to an element, the force upon the element will stress the element, causing the element to strain to accommodate the stress. The stress applied to the element, though, is directly proportional to the force applied to the element and inversely proportional to the area or volume of the element. Thus, if the force applied to the element increases and/or the volume of the element decreases, the stress will correspondingly increase in the element.
• Using this concept, the respective volumes of the elastomeric body and the rigid material inserts preferably remain substantially constant to provide accurate results. For example, if the volume of the overall seal model has substantially changed from the actual seal, the strain plots of the seal model may show an increase in strain in the elastomeric body with corresponding displacement conditions. Further, if the volume of the seal model changes from the smoothing techniques applied to the seal design of the seal model, such as increasing the volume of the elastomeric body of the seal model during the smoothing process, the strain plots of the smoothed model may show a decrease in simulated strain with corresponding displacement conditions. Thus, if the volume of the elastomeric body and the rigid material insert of the model of the seal increases or decreases, the simulated strain in the model would inherently change, independent if the seal model was modified for any improvements. Furthermore, if the overall volume of the seal remains consistent between non-smoothed and smoothed models but the relative volumes of the elastomeric body and the rigid inserts change, the strain plots may be similarly compromised.
• Continuing now withstep1230, displacement conditions are simulated upon a seal for a blowout preventer in FEA using the generated seal model. Preferably, the simulated displacement conditions are loads and strains the seal may expect to experience in service. For example, a model of a packing unit of an annular blowout preventer may require a simulated displacement condition correlating to compressing into a closed position to seal about a section of drillpipe. Further, if no drillpipe is present, the model may experience a simulated displacement condition correlating to compressing to close about itself to seal the bore.
• Instep1240, a strain plot, showing strain and deformation occurring in the seal model in response to displacement conditions may be analyzed and reviewed to determine the performance of the modeled seal. Referring now toFIGS. 19-21, cross-sectional strain plots of a seal model in accordance with embodiments disclosed herein are shown. Specifically, the seal model is of a packing unit for an annular blowout preventer, in which packing unit model is initially simulated with a displacement condition as closed about adrillpipe151. Then, the packing unit is shown in an original condition before the packing unit is simulated with the displacement condition, but the strain from the simulated displacement condition is superimposed across the non-displaced packing unit. This technique may be performed by calculating the strain from each element of the seal model with the displacement condition and showing the strain upon each corresponding element of the seal model in the original condition. This may allow the strain occurring in the packing unit under the simulated displacement condition to be “mapped” back to its original location in the packing unit.
• Referring now toFIG. 19A, a strain plot of the packing unit model shows the maximum principal log strain occurring in the seal model with a simulated displacement condition of closing the packing unit aboutdrillpipe151. InFIG. 19B, a strain plot of the seal model shows the packing unit originally before the displacement condition is simulated across the seal model inFIG. 19A, but the maximum principal log strain plot fromFIG. 19A is superimposed across the undistorted seal model. Specifically, the strain of each element in the seal model in the displacement condition inFIG. 19A is added to each element in the undistorted seal model inFIG. 19B. This allows the strain plot to show where the strain concentrations will be located when in an undisplaced condition.
• Similarly, referring toFIG. 20A, a strain plot of the packing unit model shows the axial log strain occurring in the seal model with a simulated displacement condition of closing the packing unit aboutdrillpipe151. InFIG. 20B, a strain plot of the seal model shows the packing unit originally before the displacement condition is simulated across the seal model inFIG. 20A, but the axial log strain plot fromFIG. 20A is superimposed across the undistorted seal model.
• Similarly still, referring toFIG. 21A, a strain plot of the packing unit model shows the shear log strain occurring in the seal model with a simulated displacement condition of closing the packing unit aboutdrillpipe151. InFIG. 21B, a strain plot of the seal model shows the packing unit originally before the displacement condition is simulated across the seal model inFIG. 21A, but the shear log strain plot fromFIG. 21A is superimposed across the undistorted seal model.
• As shown inFIGS. 19-21, the packing unit experiences large amounts of strain to accommodate the closed position simulated displacement condition simulated with the seal model. Because of these large strains, the finite elements of the model or mesh may not deform properly to converge to an accurate or definitive result. However, by analyzing a bulk strain plot of a smoothed model instep1240, a definitive result may be found. FEA focusing on the evaluation of bulk strain may be used to produce more accurate results.
• Referring now toFIG. 22, a graph displaying strain (y-axis) versus number of iterations (x-axis) within FEA in is shown. The simulated strain on the y-axis is a magnitude of the principal strain in a specific direction simulated across a finite element of the seal model for a given displacement condition. Further, the number of iterations on the x-axis refers to the amount of simulations of PEA used when modeling the seal. However, in contrast to the FEA iterations ofFIG. 11 whereby the model is iteratively made more localized (i.e., complex), each iteration ofFIG. 22 may incrementally smooth the analyzed model (while maintaining consistent volume) to make such analysis less complex in nature. As such, as the analysis progresses from a more localized strain analysis (i.e., the left side of the x-axis) to a bulk strain analysis (i.e., the right portion of the x-axis), the solution converges and is contained within a tolerance band of about ±1%. Specifically, the FEA solution may be seen to converge inFIG. 11 because when the simulated strain solution reaches a solution within the tolerance band, the solution continues to stay within the tolerance band even as more iterations are continued. Desirably, the simulated strain of the seal model may converge within a tolerance of at least about 0.5% of the theoretical strain.
• As such, in contrast to what one of ordinary skill in the art would intuitively believe, a simplified, smoothed model may produce a more convergent and accurate FEA solution than more complex, detailed models. As shown in this embodiment, the simulated strain produced using FEA correlates with experimentally observed solutions and converges to a definitive and correct result about the theoretical strain and within the tolerance band limitations. As the number of iterations increases (and as the model is further smoothed), the simulated strain solution produced by the FEA corresponds to the strain found in the seal through empirical testing. With these results, bulk strain FEA may provide useful results for simulation of seals for blowout preventers to further improve their designs.
• For example, referring now toFIGS. 27A, 27B, and28, strain that a seal model will sustain when simulated with a displacement condition may be shown on a strain plot when still in an undisplaced condition. This technique allows strains to be determined within areas and elements of the seal model while still in the undisplaced condition. InFIG. 27A, an enlarged view of a strain plot of a packing unit model shows the maximum principal log strain occurring in the seal model with a simulated displacement condition of closing the packing unit aboutdrillpipe151. Threefinite elements2711,2713,2715 experiencing strain when simulated with the closed displacement condition have been marked and identified. InFIG. 27B, an enlarged view of a strain plot of the seal model shows the packing unit originally before the displacement condition is simulated across the seal model inFIG. 27A, but the maximum principal log strain occurring in the seal model from the displacement condition inFIG. 27A is superimposed across the seal model. Aselements2711,2713,2715 were marked when in the displacement condition inFIG. 27A,elements2711,2713,2715 may be followed back inFIG. 27B to determine their original location within the seal model to graphically represent the magnitude and direction of the strains they experience.FIG. 28 also shows the packing unit seal model and mesh fromFIGS. 27A, 27B withelements2711,2713,2715. Using this and similar techniques, the areas of the seal model with the strain concentrations may be more easily determined to further improve the design of the seal model as necessary.
• Further, when analyzing the strain plot instep1240, the strain plots may be used to certify the seal model for use in a blowout preventer. Specifically, the strain plots may be compared against one or more specified criteria to determine if the performance of the seal model meets necessary requirements. Specified criteria, for example, may include performance requirements, customer's requirements, or even industry requirements for seals. Furthermore, such criteria may be compared against the strain plots of an analyzed seal model to determine if a seal manufactured in accordance with the model would be in compliance with such requirements. For example, a customer may require packing units of annular blowout preventers to be capable of experiencing strains in excess of 300%. A strain plot of the seal model packing unit in a closed position displacement conditions may then be compared against the specified criteria to determine if the seal model is capable of satisfying such requirements.
• In another example, industry requirements, such as API 16A/ISO 13533:2001, may be used as specified criteria to compare and certify a seal model. In particular, API 16A, Section 5.7.2 references a “closure test” for ram-type blowout preventers, while API 16A, Section 5.7.3 references a closure test for annular-type blowout preventers. Under API 16A/ISO 13533:2001, a packing unit may be required to undergo six closures about the drill pipe and, on a seventh closure, be capable of effectively sealing against pressure of about 200-300 psi (1.4-2.1 MPa). Thus, displacement conditions from industry requirements may be used in conjunction with a simulation to determine if a seal is capable of satisfying such requirements. Using methods and embodiments disclosed herein, the seal model may then be certified by comparing the strain plots of the seal model against these specified criteria.
• If the seal model generated instep1220 and analyzed instep1240 may be improved further (e.g., if the model does not meet the specified criteria), the method may loop back to step1210 to determine material properties for another material of the seal, or the method may loop back to step1220 to have the seal model regenerated or modified as necessary. This loop of generating theseal model1220 and analyzing theseal model1240 may be repeated several times until an “optimized” seal model is reached.
• In one embodiment, selected portions of an elastomeric body of a packing unit may be de-bonded from the rigid material inserts when looping back and re-generating theseal model1220 to reduce to reduce the amount and location of strain. Typically, the elastomeric body is completely bonded to metallic inserts for the packing unit to maintain maximum rigidity, as discussed above with respect to the prior art. However, if selected portions of the elastomeric body are not bonded to the rigid material inserts, this may reduce strain in the elastomer of the packing unit when the packing unit is modeled in FEA to show the strain plots.
• Referring now toFIGS. 23-25, strain plots of a smoothed seal model having such selective de-bonding are shown. Specifically, the seal model is of a packing unit for an annular blowout preventer, in which packing unit model is initially simulated with a displacement condition as closed about adrillpipe151. Then, the packing unit is shown in an original condition before the packing unit is simulated with the displacement condition, but the strain from the simulated displacement condition is superimposed across packing unit. This technique is similar toFIGS. 19-21 from above. However, the elastomeric body of the seal model inFIGS. 23-25 is additionally de-bonded from aback surface109B behind ahead109A ofmetal insert109.
• Referring now toFIG. 23A, a strain plot of the packing unit model with such a “selectively de-bonded” elastomeric body shows the maximum principal log strain occurring in the seal model with a simulated displacement condition of closing the packing unit aboutdrillpipe151. InFIG. 23B, a strain plot of the seal model shows the selectively de-bonded packing unit model originally before the displacement condition is simulated across the seal model inFIG. 23A, but the maximum principal log strain plot fromFIG. 23A is superimposed across the undistorted seal model. This allows the strain plot to show where the strain concentrations will be located when in an undisplaced condition.
• Similarly, referring toFIG. 24A, a strain plot of the packing unit model with a selectively de-bonded elastomeric body shows the axial log strain occurring in the seal model with a simulated displacement condition of closing the packing unit aboutdrillpipe151. InFIG. 24B, a strain plot of the seal model shows the selectively de-bonded packing unit model originally before the displacement condition is simulated across the seal model inFIG. 24A, but the axial log strain plot fromFIG. 24A is superimposed across the undistorted seal model.
• Similarly still, referring toFIG. 25A, a strain plot of the packing unit model with a selectively de-bonded elastomeric body shows the shear log strain occurring in the seal model with a simulated displacement condition of closing the packing unit aboutdrillpipe151. InFIG. 25B, a strain plot of the seal model shows the selectively de-bonded packing unit model originally before the displacement condition is simulated across the seal model inFIG. 25A, but the shear log strain plot fromFIG. 25A is superimposed across the undistorted seal model
• Each of the strain plots of the packing unit model with a selectively de-bonded elastomeric body (i.e.,FIGS. 23-25) indicates less strain than the strain plots of the packing unit model without selective de-bonding of the elastomeric body (i.e.,FIGS. 19-21). Specifically, the volume of the elastomeric body adjacent to the back surface of the head of the rigid material insert indicates less strain in the strain plots of the seal model when the elastomeric body is de-bonded from the rigid material insert. Thus, as shown with the selectively de-bonded packing unit, the seal model may be modified and regenerated to produce an optimized seal model that reduces the location and amount of strain occurring in the seal model.
• Similar to above with respect to generating a seal model instep1220, when simulating displacement conditions across theseal models1230, it is preferable for the volumes of the seal model and its components to remain substantially constant. If the volumes do not remain constant, the results of the strain plots and simulated strain in FEA may not correlate with experimentally observed results or with one another, thereby providing inaccurate results. For example, if the volume of the seal models of the packing units shown in the strain plots ofFIGS. 19-21 changes from the volume of the seal models of the packing units shown in the strain plots ofFIGS. 23-25, it would be difficult to compare the strain plots because of the added factor of the changing volume. As the volume of the seal model of the packing unit increases or decreases, the simulated strain in the packing unit inherently changes, independent if the seal model was modified for any improvements.
• In step1250, after being generated, analyzed, and possibly re-generated (if necessary), the seal model may be used to manufacture a seal for a blowout preventer1250. Specifically, using techniques known in the art, a seal based upon the three-dimensional seal model may be manufactured for use in a blowout preventer, such as a packing unit for an annular blowout preventer or a top seal or variable bore ram packer a ram blowout preventer. For example, the seal model of the packing unit for the annular blowout preventer having selective de-bonding, as discussed above and shown inFIGS. 23-25, may be manufactured for use in the industry. The selective de-bonding packing unit generated in FEA reduced the strain concentrations in the packing unit when in the closed position, as compared to the packing unit shown inFIGS. 19-21. This selective de-bonding seal model may then be manufactured for use or testing within a blowout preventer because of its improved performance over the other packing unit shown from the FEA.
• Aspects of embodiments disclosed herein, such as generating and analyzing a seal model of a seal for a blowout preventer using FEA, may be implemented on any type of computer regardless of the platform being used. For example, as shown inFIG. 26, anetworked computer system3060 that may be used in accordance with an embodiment disclosed herein includes aprocessor3062, associatedmemory3064, astorage device3066, and numerous other elements and functionalities typical of today's computers (not shown).Networked computer3060 may also include input means, such as akeyboard3068 and a mouse3070, and output means, such as amonitor3072.Networked computer system3060 is connected to a local area network (LAN) or a wide area network (e.g., the Internet) (not shown) via a network interface connection (not shown). Those skilled in the art will appreciate that these input and output means may take many other forms. Additionally, the computer system may not be connected to a network. Further, those skilled in the art will appreciate that one or more elements ofaforementioned computer3060 may be located at a remote location and connected to the other elements over a network.
• Advantageously, methods and embodiments disclosed herein may provide improved and more accurate results when using FEA. Methods and embodiments disclosed herein use strain within FEA to determine the performance characteristics of seals for blowout preventers under simulated displacement conditions. This allows the finite elements within the seal model to displace when accommodating large amounts of strain.
• Further, methods and embodiments disclosed herein may provide techniques for analyzing, smoothing, simplifying, and modifying seal models for use in FEA. Using these techniques, the accuracy of the results of the strain plots created using FEA may be improved. Additionally, using these techniques, the seal model may be modified to reduce the amount and location of strain (e.g., strain concentrations) occurring in the seal model from the simulated strain plots.
• Furthermore, methods and embodiments disclosed herein may provide for a seal for a blowout preventer with an increased working lifespan. For example, the packing unit may be modeled with simulated displacement conditions of repeated closures (i.e., repeatably closing the seal about a drillpipe or itself) to determine design features that may extend the working lifespan (i.e., number of closures) of the packing unit.
• While the invention has been described with respect to a limited number of embodiments, those skilled in the art, having benefit of this disclosure, will appreciate that other embodiments may be devised which do not depart from the scope of the invention as disclosed herein. Accordingly, the scope of the invention should be limited only by the attached claims.

Claims (31)

1. A method of manufacturing a seal of a blowout preventer, the method comprising:

selecting a seal design;

generating a first finite element analysis seal model from the selected seal design;

smoothing the first finite element analysis seal model;

analyzing a strain plot of the smoothed first finite element analysis seal model based on a displacement condition; and

manufacturing a seal.

2. (canceled)

3. The method ofclaim 1, wherein the smoothing comprises modifying at least one of an internal corner and an external corner of a rigid material insert of the first finite element analysis seal model.

4. (canceled)

5. The method ofclaim 1, wherein the smoothing comprises modifying a compression face of an elastomeric body of the first finite element analysis seal model.

6. The method ofclaim 1, wherein the smoothing comprises modifying an end of a flange of a rigid material insert of the first finite element analysis seal model.

7. The method ofclaim 1, wherein the smoothing comprises modifying an end of a web of a rigid material insert of the first finite element analysis seal model.

8. The method ofclaim 1, wherein the smoothing comprises modifying a side of a flange of a rigid material insert of the first finite element analysis seal model.

9. The method ofclaim 1, wherein the smoothing comprises modifying a side of a web of a rigid material insert of the first finite element analysis seal model.

10. The method ofclaim 1, further comprising:

generating a second finite element analysis seal model based on the analyzed strain plot of the smoothed first finite element analysis seal model; and

analyzing a strain plot of the second finite element analysis seal model based on the displacement condition.

11. (canceled)

12. The method ofclaim 10, wherein at least one of the first finite element analysis seal model and the second finite element analysis seal model converges within a tolerance of about 1%.

13. The method ofclaim 10, wherein at least one of the first finite element analysis seal model and the second finite element analysis seal model converges within a tolerance of about 0.5%.

14. The method ofclaim 10, wherein a volume of an elastomeric body of the second finite element analysis seal model is maintained substantially constant with the volume of the smoothed finite element analysis seal model.

15. The method ofclaim 1, wherein a volume of an elastomeric body of the first finite element analysis seal model is maintained substantially constant during smoothing.

16. The method ofclaim 1, wherein the seal comprises an elastomer and a rigid material.

17-20. (canceled)

21. The method ofclaim 1, wherein the displacement condition comprises strain of at least about 300%.

22. (canceled)

23. The method ofclaim 1, wherein the strain plot comprises one of maximum principal strain, axial strain, and shear strain.

24. The method ofclaim 1, wherein the strain plot comprises a cross-sectional view of the first finite element analysis seal model.

25. A method to certify a seal of a blowout preventer, the method comprising:

generating a first finite element analysis seal model;

smoothing the first finite element analysis seal model;

analyzing a strain plot of the smoothed first finite element analysis seal model based upon a displacement condition; and

comparing the strain plot of the smoothed first finite element analysis seal model against at least one specified criteria.

26. The method ofclaim 25, further comprising:

generating a second finite element analysis seal model based on the analyzed strain plot;

analyzing a strain plot of the second finite element analysis seal model based on the displacement condition; and

comparing the strain plot of the second finite element analysis seal model against the at least one specified criteria.

27. The method ofclaim 26, further comprising smoothing the second finite element analysis seal model.

28. The method ofclaim 25, wherein the seal comprises an elastomer and a rigid material.

29. (canceled)

30. The method ofclaim 29, wherein the industry requirements comprise API 16A/ISO 13533:2001.

31. A method of optimizing a seal of a blowout preventer, the method comprising:

smoothing a first finite element analysis seal model;

analyzing a strain plot of the smoothed first finite element analysis seal model based upon a displacement condition;

generating a second finite element analysis seal model based on the analyzed strain plot of the smoothed first finite element analysis seal model;

smoothing the second finite element analysis seal model;

analyzing a strain plot of the second smoothed finite element analysis seal model based upon a displacement condition; and

repeating the analyzing and generating of smoothed finite element analysis seal models until an optimized seal model is reached.

32. The method ofclaim 31, wherein the seal comprises an elastomer and a rigid material.

33. The method ofclaim 31, wherein a volume of the first finite element analysis seal model and a volume of the second finite element analysis seal model are substantially the same.

34. The method ofclaim 31, wherein the optimized seal model is compared against at least one specified criteria.

Images