The high temperature composites have been studied for applications to secondary structures due to their light weight and thermal resistance. Relatively few studies have been conducted to consider them for primary structural load bearing capabilities. These studies focused on titanium matrix composites to characterize their material behavior [1], unidirectional [2], simple loading conditions [3, 4] in a laboratory environment or unrealistic structural geometry [5]. The purpose of this work was to study fatigue damage and determine fatigue life in titanium matrix composite panels at unloaded fastener holes subjected to thermo-mechanical fatigue loads with variable amplitudes and temperature ranges. The test panels were machined from a prefabricated structural component with pre-drilled fastener holes. The test material was a 32 ply, quasi-isotropic, approximately .224 inch thick titanium matrix laminated composite with SCS-6 fibers and Ti-15-3 metal matrix. The material was HIP consolidated followed by slow cool to room temperature. The thermal zone area was 2 inches long along specimen length and 1.875 inch wide with a .3125 inch fastener hole at the center of the thermal zone. All specimens were machined using a 3-D water jet cutter. The test system consisted of a closed loop servo-hydraulic 30 Kip test system equipped with an MTS model 458 control system, a 486 PC containing a Keithley Metrabyte DAS 1601 computer card. The specimens were gripped using MTS model 647 side load hydraulic wedge grips equipped with surfalloy grip surface. The thermal loads were provided by an Ameritherm 5 kilowatt induction power supply and a total temperature instrumentation model MC-125 temperature controller. The temperature controller was equipped with analog set point and recorder output of temperatures with both set for 1–5 volt signal levels for 0 to 1832 F. The computer generated the temperature and load profiles and monitored error band for temperature. The computer system was set to null pace the temperature and loads if the temperature exceeded a 18 degree F variation. In effect all processes would hold until the temperature error returned inside the error band. This temperature error control was accomplished by comparing the command signal to the temperature controller to the process temperature signal from the temperature controller. The nominal uniform temperature zone was one inch long centered at the specimen geometric center and maintained required temperatures within 10 degrees. The variations in temperatures along the crack line were controlled to with in 5 degrees. Cooling blocks were attached to the test samples at the end of uniform sections near the fillet blend. These blocks were cooled with water passages and compressed air was passed through holes in the blocks and impinged on the samples to provide additional cooling at the end of the thermal ramp during cool -down. The air was turned on by the computer at about 400 degrees F during each block. On all notched test samples, an extensometer was mounted across the center flaw to obtain load-deflection data (COD). The optical crack lenth measurements were made using a 20 X Gaertner traveling microscope. The load versus crack mouth opening displacement readings were taken to compare with the optical measurements of the crack length. The thermomechanical load spectrum was developed from the distribution and frequency of loading that the airframe will experience based on the design service life and typical design usage. The loads and environmental spectra are used to develop design flight by flight stress environment spectra. The data and failure surfaces were analyzed to study the high stress and low stress failure, environmental degradations, surface cracks in matrix and the effect of notch on crack initiation failure mechanism. During this investigation it was observed that the most difficult task in thermomechanical fatigue testing is to control the cooling rate as required by the thermal profile. The results show that the fatigue life depend on the applied maximum stress, increased temperatures and hold levels of both the loads and the temperatures. The variation in experimental fatigue life is with in the order of magnitude typical of fatigue data considering the complexity of the test and loading conditions. The SEM photographs and micrographs showed that in titanium matrix composite, the mode of cracking is under partial bridging of fibers at the matrix crack. The COD data was of little use for totally automated measurements when comparing with the crack sizes measured.

This content is only available via PDF.
You do not currently have access to this content.