Test Methods for Hoop Tensile Strength of Ceramic Composite Tubes for Light Water Nuclear Reactor Applications
No abstract available
SEARCH · Engineering Papers
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
No abstract available
No abstract available
This contract reviewed the available literature on mechanisms of low velocity impact damage in filament wound rocket motor cases, MDE methods to quantify damage, critical coupon level test methods, manufacturing and material process variables and empirical and analytical modeling off impact damage. The critical design properties for rocket motor cases are biaxial hoop and axial tensile strength. Low velocity impact damage is insidious because it can create serious nonvisible damage at very low impact velocities. In thick rocket motor cases the prevalent low velocity impact damage is fiber fracture and matrix cracking adjacent to the front face. In contrast, low velocity loading of thin wall cylinders induces flexure, depending on span length and the flexure induces delamination and tensile cracking on the back face wall opposed to impact occurs due to flexural stresses imposed by impact loading. Important NDE methods for rocket motor cases are non-contacting methods that allow inspection from one side. Among these are vibrothermography, and pulse-echo methods based on acoustic-ultrasonic methods. High resolution techniques such as x-ray computed tomography appear to have merit for accurate geometrical characterization of local damage to support development of analytical models of micromechanics. The challenge of coupon level testing is to reproduce the biaxial stress state that the full scale article experiences, and to determine how to scale the composite structure to model full sized behavior. Biaxial tensile testing has been performed by uniaxially tensile loading internally pressurized cylinders. This is experimentally difficult due to gripping problems and pressure containment. Much prior work focused on uniaxial tensile testing of model filament wound cylinders. Interpretation of the results of some studies is complicated by the fact that the fabrication process did not duplicate full scale manufacturing. It is difficult to scale results from testing subscale cylinders since there are significant differences in out time of the resins relative to full scale cylinder fabrication, differences in hoop fiber tensioning and unsatisfactory coupon configurations. It appears that development of a new test method for subscale cylinders is merited. Damage tolerance may be improved by material optimization that uses fiber treatments and matrix modifications to control the fiber matrix interface bonding. It is difficult to develop process optimization in subscale cylinders without also modeling the longer out times resins experience in full scale testing. A major breakthrough in characterizing the effect of impact damage on residual strength, and understanding how to scale results of subscale evaluations, will be a sound micromechanical model that described progressive failure of the composite. Such models will utilize a three dimensional stress analysis due to the complex nature of low velocity impact stresses in thick composites. When these models are coupled with non-contact NDE methods that geometrically characterize the damage and acoustic methods that characterize the effective local elastic properties, accurate assessment of residual strength from impact damage may be possible. Directions for further development are suggested.
For angleply composites, lamina tension and compression strengths are commonly determined by applying classical lamination theory to test data obtained from testing of angleply composite specimens. For textile composites such as 2D triaxial braids, analysis is more complex and standard test methods do not always yield reliable strength measurements. This paper describes recent research focused on development of more reliable tensile test methods for braided composites and presents preliminary data for various approaches. The materials investigated in this work have 0deg+/-60 2D triaxial braid architecture with nearly equal fiber volume fraction in each of the three fiber directions. Flat composite panels are fabricated by resin transfer molding (RTM) using six layers of the braided preform aligned along the 0deg fiber direction. Various epoxy resins are used as matrix materials. Single layer panels are also fabricated in order to examine local variations in deformation related to the braid architecture. Specimens are cut from these panels in the shape of standard straight-sided coupons, an alternative bowtie geometry, and an alternative notched geometry. Axial tensile properties are measured using specimens loaded along the 0deg fiber direction. Transverse tensile properties are measured using specimens loaded perpendicular to the 0deg fibers. Composite tubes are also fabricated by RTM. These tubes are tested by internal pressurization using a soft rubbery material sealed between the inside diameter of the tube and the load fixtures. The ends of the tube are unconstrained, so the primary load is in the hoop direction. Tubes are fabricated with the 0deg fibers aligned along the tube axis by overbraiding the preform on a mandrel. Since the loading is in the hoop direction, testing of the overbraided tube provides a measure of transverse tensile strength. Previous work has indicated that straight-sided coupons yield a transverse tensile strength that is much lower than the expected material strength because of premature edge-initiated failure. Full-field strain measured during transverse tensile tests clearly showed accumulation of edge damage prior to failure. In the current work, high speed video and testing of single layer specimens are used to investigate potential failure mechanisms in more detail. High speed video clearly shows the edge initiation in six layer transverse tensile test coupons. Specimens with the bowtie geometry and the notched geometry minimize this edge effect and yield significantly higher transverse tensile strength values compared to the straight-sided coupons. However, bowtie and notched specimens geometries are not ideal because of the non-uniform stress and strain fields in the region of failure. Testing of tubes using internal pressurization eliminates edge-initiated failure and provides a more uniform state of stress and strain. Preliminary results indicate that bowtie, notched, and tube specimens yield comparable values for transverse tensile strength and that these values are much higher than the strength measured using a straight-sided coupon.
A curved beam 'C-coupon' was used to assess fracture behavior in a Sylramic(tm)/melt infiltration (MI) SiC matrix composite. Failure stresses and fracture mechanisms, as determined by optical and scanning electron microstructural analysis, are compared with finite element stress calculations to analyze failure modes. Material microstructure was found to have a strong influence on mechanical behavior. Fracture occurs in interlaminar tension (ILT), provided that the ratio of ILT to tensile strength for the material is less than the ratio of radial to hoop stresses for the C-coupon geometry. Utilization of 3D architectures to improve interlaminar strength requires significant development efforts to incorporate through thickness fibers in regions with high curvatures while maintaining uniform thickness, radius, and microstructure.
SiC/SiC composites are excellent material candidates for high temperature applications where the performance requirements are high strength, high creep-rupture resistance, high environmental durability, and high thermal conductivity. In the past, the NASA UEET program has demonstrated fabrication of high-performance SiC/SiC flat panels reinforced by Sylramic-iBN SiC fibers. Currently NASA UEET is scaling up this SiC/SiC system by fabrication of more complex shaped components using the same fiber type. This paper reports the effects of various fiber architectures on the processing, mechanical, and durability behavior of small-diameter 0.5" ID SiC/SiC tubes, which are potential sub-elements for leading edges and cooling channels in turbine vanes and blades. Nine different fiber architectures were utilized for construction of seamless tube preforms, from simple 2D jelly-rolling to complex braiding, pin-weaving, filament-winding and 3D orthogonal weaving with approximately 5% fibers in the thru-thickness direction. Using the BN interphase and Sic matrix processing steps established for the flat panels, SiC/SiC tubes were fabricated with wall thicknesses of approximately 60 mils and total fiber fractions of approximately 35%. The "D" split ring tests for hoop tensile properties, micro-structural examinations for relationship between fiber architecture formation and matrix infiltration, and the low-pressure burner rig tests for the high temperature durability under thru-thickness thermal gradient were conducted. The better matrix infiltration and higher hoop strength were achieved using the tri-axial braided and the three-float pin woven SiC/SiC tubes. In general, it needs not only higher hoop direction fibers but also axial direction fibers for the higher hoop strength and the better infiltration, respectively. These results are analyzed to offer general guidelines for selecting fiber pre-form architectures and SiC/SiC processes that maximize tube hoop strength, thru-thickness thermal conductivity, and burner-rig durability under a high thermal gradient.
The elastic and yield response of extruded thin-walled high-density polyethylene tubes with a density in the range of 0.961 to 0.964 gm/cu cm was investigated. Material properties in the axial and hoop directions were measured, and the tubes were found to be mildly transversely isotropic. The yield response was pressure sensitive, and was well predicted using the pressure-modified Hill criterion using a compressive to tensile yield strength ratio of 1:12.
A diametral compression test was used to measure the transverse strength of SCS-6 SiC fibers before and after heat treatment. Subjecting fibers to diametral compression successfully produced transverse tensile failure in the form of fiber cracking along the same diametral plane in which the compressive load was applied. An analysis of the hoop stress along the diametral plane, in which the effects of the C core were included, showed that there is a large tensile hoop stress concentration in the SiC sheath at the interface between the C core and the SiC sheath, where the stress is 6.3 times greater than the stress present in a solid SiC fiber under identical loading. This high tensile hoop stress concentration promotes crack initiation near the core and significantly limits the capability of these fibers to withstand transverse compressive loading. The maximum tensile hoop stresses, located at the interface between the C core and SiC sheath, at the measured failure loads were 850 MPa for the as-received SCS-6 fiber and 1210 MPa for fibers exposed to a 1-h heat treatment at 1850°C in 138 MPa of Ar.
SiC/SiC engine components, high-modulus Sylramic-iBN SiC fiber tows were used to form nine different tubular architectural preforms with 13 mm (0.5 in.) inner diameter and lengths of approx. 75 and 230 mm (approx. 3 and approx, 9 in.). The thin-walled preforms were then coated with a BN interphase and densified with a hybrid SiC matrix using nearly the same process steps previously established for slurry-cast melt-infiltrated Sylramic-iBN/BN/SiC flat panels. The as-fabricated CMC tubes were microstructurally evaluated and tested for tensile hoop and flexural behavior, and some of the tubes were also tested in a low-pressure burner rig test with a high thru-thickness thermal gradient. To date, four general tube scale-up issues have been identified: greater CVI deposits on outer wall than inner wall; increased ply thickness and reduced fiber fraction; poor test standards for accurately determining the hoop strength of a small-diameter tube; and poor hoop strength for architectures with seams or ply ends. The underlying mechanisms and possible methods for their minimization are discussed.
The initial edition of the Mil Handbook 17, Vol. 5 on ceramic matrix composites (CMCs) was recently published. Part C of this CMC volume describes procedures for characterization of the thermal, physical, and mechanical properties of CMCs and subsequent data analysis procedures. As is the case for entire CMC volume, this chapter is a "work in progress" with sections on recommended test plans and matrices, data reduction, and test methods. This presentation will describe the current status in development of this section of the CMC volume and will highlight examples of critical issues that will be addressed in this volume. CMC systems are a different class of material than metals or polymer matrix composite with different mechanical response and failure modes, design rules, test methods, and characterization techniques. The CMC test protocols need to include guidance on how to generate enough reliable data in order to estimate, with reasonable confidence, the lower bounds of performance properties, such as strength and life. Strength data sets on two material systems, C/SiC and Nicalon/SiNC, and a creep rupture data set on C/SiC, will be presented to highlight this issue. The structural properties of CMC parts can sometimes be significantly different than those obtained from flat panels because of changes in fiber architecture often required to make shapes and spatial and lot-to-lot variations in processing conditions. To illustrate an example, the hoop tensile properties of SiC/SiC cylinders will be compared to properties obtained from flat panels fabricated from the same composites. These CMC testing-design issues are being defined and documented in the Mil Handbook 17 CMC volume.
The application of polyimide resin as a matrix for glass filament-wound thin metal-lined pressure vessels was studied over a temperature range of (minus) 320 to 600 F. Keramid 601 polyimide was found to perform quite well over the entire range of temperature. Hoop stress values of 425 ksi were determined at 75 F which is equivalent to epoxy resin in similar structures. At -320 and 600 F, 125 and 80% of this strength was retained. Thermal ageing at 500 F for up to 50 hours was studied with severe reduction in strength, but there is evidence that this reduction could be improved. Another polyimide resin studied was P10PA which was found to have processing characteristics inappropriate for filament-winding. NOL ring tensile and shear data was determined from both resins with S-glass. Pressure vessel design, fabrication and test procedures are described in detail.
The increasing use of advanced composite materials in the wide range of applications including Space Structures is a great impetus to the development of smart materials. Incorporating these FBG sensors for monitoring the integrity of structures during their life cycle will provide valuable information about viability of the usage of such material. The use of these sensors by surface bonding or embedding in this composite will measure internal strain and temperature, and hence the integrity of the assembled engineering structures. This paper focuses on such a structure, called a composite wound pressure vessel. This vessel was fabricated from the composite material: TRH50 (a Mitsubishi carbon fiber with a 710-ksi tensile strength and a 37 Msi modulus) impregnated with an epoxy resin from NEWPORT composites (WDE-3D-1). This epoxy resin in water dispersed system without any solvents and it cures in the 240-310 degrees F range. This is a toughened resin system specifically designed for pressure applications. These materials are a natural fit for fiber sensors since the polyimide outer buffer coating of fiber can be integrated into the polymer matrix of the composite material with negligible residual stress. The tank was wound with two helical patterns and 4 hoop wraps. The order of winding is: two hoops, two helical and two hoops. The wall thickness of the composite should be about 80 mil or less. The tank should burst near 3,000 psi or less. We can measure the actual wall thickness by ultrasonic or we can burst the tank and measure the pieces. Figure 1 shows a cylinder fabricated out of carbon-epoxy composite material. The strain in different directions is measured with a surface bonded fiber Bragg gratings and with embedded fiber Bragg gratings as the cylinder is pressurized to burst pressures. Figure 2 shows the strain as a function of pressure of carbon-epoxy cylinder as it is pressurized with water. Strain is measured in different directions by multiple gratings oriented in both axial and hoops directions.
Flight 5 aft segments with nozzles were exposed to -17 F temperatures while awaiting shipment to KSC in February, 1989. No records were found which show that any previous nozzles were exposed to air temperatures as low as those seen by the Flight 5 nozzles. Thermal analysis shows that the temperature of the fixed housing, and forward and aft exit cone components dropped as low as -10 F. Structural analysis of the nozzles at these low temperatures show the forward and aft exit cone adhesive bonds to have a positive margin of safety, based on a 2.0 safety factor. These analyses show the normal and shear stresses in the fixed housing bond as low values. However, the hoop and meridinal stresses were predicted to be in the 4000 psi range; the failure stress allowable of EA913NA adhesive at -7 F. If the bonds did break in directions perpendicular to the surfaces, called bond crazing, no normal bond strength would be lost. Testing was conducted in two phases, showing that no degradation to the adhesive bonds occurred while the Flight 5 nozzles were subjected to subzero temperatures. The results of these tests are documented. Phase 1 testing cooled a full-scale RSRM insulated fixed housing to -13 F, with extensive bondline inspections. Phase 2 testing cooled the witness panel adhesive tensile buttions to -13 F, with failure strengths recorded before, during, and after the cooldown.