Engineering topics
Publications and source records attributed to Walker, James L., II.
This viewgraph presentation describes the nondestructive evaluation methods that are suitable for detecting bond line defects in Space Shuttle High Temperature Reusable Surface Insulation Tiles.
A probability of detection study was performed for the detection of impact damage using flash heating infrared thermography on a full scale honeycomb composite structure. The honeycomb structure was an intertank structure from a previous NASA technology demonstration program. The intertank was fabricated from IM7/8552 carbon fiber/epoxy facesheets and aluminum honeycomb core. The intertank was impacted in multiple locations with a range of impact energies utilizing a spherical indenter. In a single blind study, the intertank was inspected with thermography before and after impact damage was incurred. Following thermographic inspection several impact sites were sectioned from the intertank and cross-sectioned for microscopic comparisons of NDE detection and actual damage incurred. The study concluded that thermographic inspection was a good method of detecting delamination damage incurred by impact. The 90/95 confidence level on the probability of detection was close to the impact energy that delaminations were first observed through cross-sectional analysis.
Acoustic emission data were collected during the hydroburst testing of eleven 15 inch diameter filament wound composite overwrapped pressure vessels. A neural network burst pressure prediction was generated from the resulting AE amplitude data. The bottles shared commonality of graphite fiber, epoxy resin, and cure time. Individual bottles varied by cure mode (rotisserie versus static oven curing), types of inflicted damage, temperature of the pressurant, and pressurization scheme. Three categorical variables were selected to represent undamaged bottles, impact damaged bottles, and bottles with lacerated hoop fibers. This categorization along with the removal of the AE data from the disbonding noise between the aluminum liner and the composite overwrap allowed the prediction of burst pressures in all three sets of bottles using a single backpropagation neural network. Here the worst case error was 3.38 percent.
A full scale thin walled composite tank, designed and fabricated for the storage of pressurized RP- I rocket fuel, was fully inspected with digital infrared thermography (IR) during assembly and prior to proof testing. The tank featured a "pill capsule" design with the equatorial bondline being overwrapped on both the inner and outer surfaces. A composite skirt was bonded to the aft dome of the tank to serve as a structural support when the tank was stood on end in service. Numerous anomalies were detected and mapped prior to proof testing, some along bondlines and some scattered throughout the acreage. After the tank was intentionally burst, coupons were cut from the regions including thermographic anomalies. These coupons were again inspected thermographically to document the growth of any indications due to proof testing. Ultrasonic inspections (UT) were also performed on the coupons for comparison to thermography. Several coupons were dissected and micrographed. Relationships between IR and UT indications and the physical nature of the dissected material are presented.
The methods utilized to perform an infrared thermography inspection of the bondline of structural reinforcing caps on X-33 liquid hydrogen (LH2) tank during the manufacture are described in this presentation. A total of thirty-six caps were bonded to the tank using a wet cobond process with vacuum bagging and low temperature curing. The thermal map of each bondline was acquired through the vacuum bag by applying heat to the cap and then viewing it during cool down. Poorly bonded regions, or voids, were characterized by localized hot spots in the thermal image. These regions were marked on the tank by referring to a live thermal image, then hand massaged in an attempt to remove the entrapped air. The cap was then reinspected and, if it was determined that the entrapped air could not be removed, the cap was removed and the adhesive cleaned away before it cured. Measures were then taken to better match the surface of the cap to the tank and the process repeated. After the bondline had cured sufficiently to allow vacuum bag removal, a "flash heating" thermographic Investigation was performed to verify that the cap had bonded properly to the tank. Any regions that had opened up after the preliminary inspection were marked and filled by drilling small holes in the cap and injecting resin. A final inspection was then performed to ensure the quality of the bondline.
Acoustic emission (AE) data were taken during hydroproof for three sets of ASTM standard 5.75 inch diameter filament wound graphite/epoxy bottles. All three sets of bottles had the same design and were wound from the same graphite fiber; the only difference was in the epoxies used. Two of the epoxies had similar mechanical properties, and because the acoustic properties of materials are a function of their stiffnesses, it was thought that the AE data from the two sets might also be similar; however, this was not the case. Therefore, the three resin types were categorized using dummy variables, which allowed the prediction of burst pressures all three sets of bottles using a single neural network. Three bottles from each set were used to train the network. The resin category, the AE amplitude distribution data taken up to 25 % of the expected burst pressure, and the actual burst pressures were used as inputs. Architecturally, the network consisted of a forty-three neuron input layer (a single categorical variable defining the resin type plus forty-two continuous variables for the AE amplitude frequencies), a fifteen neuron hidden layer for mapping, and a single output neuron for burst pressure prediction. The network trained on all three bottle sets was able to predict burst pressures in the remaining bottles with a worst case error of + 6.59%, slightly greater than the desired goal of + 5%. This larger than desired error was due to poor resolution in the amplitude data for the third bottle set. When the third set of bottles was eliminated from consideration, only four hidden layer neurons were necessary to generate a worst case prediction error of - 3.43%, well within the desired goal.