1999-2000 OTRC Project: Interdisciplinary Design for Composite Coiled Tubulars
OBJECTIVES AND APPROACH:
Comprehensive lamina and laminate characterization is essential for robust design tools that reflect the actual service conditions for composite structures. Fundamental issues of processing, manufacturing in relation to design, and analysis of hybrid composite coiled tubulars will be investigated through three different perspectives; manufacturing and service environment, testing rationale, and computational simulations of time dependent nonlinear material and geometric response. Advantages associated with this approach include greater accessibility to design parameters, reduced computational demands, and the ability to customize and diversify. The strength of the proposed collaborative program (Drs. Ochoa, Schapery, Whitcomb) is based upon the integration of time and rate dependent deformation and damage mechanisms and hybridization to provide and verify a design with the least susceptibility to damage.
The major issues addressed at laminate, lamina and constituent scales will be ; (i) the hybrid layup and fiber orientation to develop a robust and economical reinforcement architecture, (ii) response to combined loads, and (iii) simulation of nonlinear material behavior induced by processing and service loads. At the lamina and micro scales prediction of damage initiation and growth can require time dependent modeling and nonlinear material and geometry considerations. The lamina and laminate characteristics for generating realistic design allowables may require non-standard tests in addition to conventional ones at multiple scales. Furthermore, the advanced computational tools that are necessary for capturing the complex behavior need to be robust.
(i) Hybrid Layup and Fiber Orientation
Two significantly different hybridization options will be undertaken to study the resistance to damage initiation and progression;namely, in-plane and through-the-thickness placement of glass and carbon fibers. The volume fraction of each fiber and their geometric placement will be taken as the initial metrics for deciding on the preliminary tube wall geometry. The selection will be consistent with typical industrial sizes for tube radius to tube wall thickness in the range of 3<(r/t)<20 with spooling strains controlled by the ratio of the tube to spool radius. Optimization techniques will be utilized in the design of coiled hybrid tubes, followed by similitude studies for proper testing of composite prototypes, and finally, detailed stress analysis will be performed for a variety of loading situations to determine damage initiation in the tubes to gather insight to design the fixtures and specimens for the experiments. The commercial finite element packages MARC/MENTAT and ABAQUS, as well as in-house codes will be used to evaluate the possibility of preventing and/or delaying certain failure modes by optimal placement of glass and carbon fibers with filament winding.(ii) Combined Loads
Realistic experiments will be designed at multiple scales to understand the structural and material response during service and at proof testing at the manufacturing site. The focus will be on developing innovative short-term laboratory tests that allow for accurate, reliable prediction of long-term performance of glass-carbon fiber reinforced composite coiled tubulars subjected to the following loads.
a. internal pressure and bending: FAT (Factory Acceptance Test) where the tube is pressurized while it is sitting on the spool. For 1.5" tube, the pressure is about 15-20 ksi.
b. bending and torsion: The anticipated source of torque is from a turn table where the coiled tubular spool may be installed during drilling. The bending may be as high as 35 degrees per 100ft .
c. internal pressure and torsion: If the torsion is induced by rotation, the internal pressure expected is 5 ksi where as with a pressurized drill bit, it may be as high as 20 ksi
d. compression and torsion: In downhole operations in contrast to flow line applications, the presence of casing and the tolerance between the coiled tubular and casing wall may lead to buckling
(iii) Computational Models for Nonlinear Material Behavior
The combination of robust finite element theory and increasingly powerful computers has greatly improved prediction of structural response. However, accurate modeling of composites structures with thick sections remains a major challenge. Three of the characteristics that contribute to this challenge are
a. The large thickness, which prevents ply-by-ply modeling.
b. Damage is diffuse, so discrete modeling of each damage site is impractical.
c. The polymer matrix exhibits viscoelastic behavior. It can also behave nonlinearly.
In recent years, we have made considerable progress in global/local strategies that permit progressive failure predictions for thick laminates. Some of the techniques were used in developing a constitutive module for ABAQUS that was used to analyze the NIST riser joint. However, this constitutive module does not include viscoelastic effects that may be important in the spooling of tubulars. The amount of microcracking is potentially controllable by selective heating of the tubular during spooling and unspooling.
Accordingly, the goal of the proposed effort will be to develop robust strategies for including viscoelastic effects in structural analysis of thick composite structures. Analyses will be developed for determining effective viscoelastic properties at both the fiber/matrix and sublaminate scale. These properties are a critical part of the global/local strategy. The adequacy of the correspondence principle combined with approximate Laplace transform inversion will be evaluated by comparison with direct transient simulation. A global/local viscoelastic analysis will be developed based on parametric studies with these micro- and macro-mechanics analyses.
The response of a viscoelastic material can be characterized using relaxation moduli or creep compliances. Although the relaxation moduli formulation is perhaps more directly compatible with finite elements, the experimental measurement of relaxation moduli is more complex than for creep compliances. One goal of this effort will be to assess the merits of each formulation for the particular types of material systems of interest to the offshore oil industry.
MILESTONES:
YEAR 1
YEAR 2
YEAR 3
PRINCIPAL INVESTIGATORS: O.O. Ochoa and J.J. Whitcomb