Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
Research Project Overviews
Federal Aviation Administration Joint Advanced Materials and Structures (JAMS) CoE & Materials & Structures Branch
Technical Review Meeting Hosted by CECAM
May 24 – May 26, 2005
Table of Contents
Presentation Day Pages Tuesday, May 24, 2005 …………………………………………………………….2‐20 Wednesday, May 25, 2005 ……………………………………………………….21‐64 Thursday, May 26, 2005 ………………………………………………………….65‐77
2
Tuesday, May 24, 2005
Structural Health Monitoring for Life Management of Aircraft Principle Investigator(s): J. D. Achenbach, S. Krishnaswamy, and I. M. Daniel Background Research and development is proposed for structural health monitoring of aircraft components made of composite materials, specifically, graphite‐epoxy laminates. The proposed work has three primary components: (1) investigate the evolution of fatigue damage and select a damage parameter; (2) select and employ suitable sensor technology to monitor fatigue damage; and (3) use the measured damage parameters in probabilistic failure analysis. The failure sequence in composite laminates consists of matrix cracking, local delaminations, and, finally, fiber breakage. In combination with the experimental work, appropriate modeling will be used to define an appropriate damage parameter and a scalar damage function. Sensor technology suitable for permanent installation in or on composite components will provide the selected damage parameter as a function of the number of fatigue cycles. The measured damage, the probability of detection, the stress level and the damage growth characteristics will be incorporated in the probabilistic damage model, to calculate the probability of critical damage accumulation. This is a proposal for an initiation project that is expected to lead to a full‐fledged application of structural health monitoring of composite aircraft components. At Northwestern University we have significant experience with work related to the SHM of metal aircraft structures, including damage detection of a measurable damage parameter and the use of this parameter for growth prediction by a probabilistic fatigue damage procedure (Achenbach and Krishnaswamy), and with the study of degradation mechanisms and failure modes and their detection in advanced composites (Daniel). The proposed work will bring together the experience gained in these two areas for the development of structural health monitoring of composite aircraft structures. Based on the experience of the present investigators, existing problems are expected to be surmountable for composite structures. The experience gained with metal structures will be invaluable in the proposed developments for composite structures. The main differences between the two developments are the character of failure mechanisms and the placement of sensors, primarily on the surface for metal structures and possibly embedded for composites. This proposal will, therefore, be concerned with the following points of the request for work stated in the Solicitation: Objectives • Study and develop understanding of material degradation mechanisms and failure modes; • Study and develop onboard sensing technologies;
3
• Study and develop sensor optimization and integration; • Study and develop material prognostics; • Validate and demonstrate health monitoring instrumentation and life extension
methodologies; • Train the workforce on health monitoring instrumentation. This will be done by our Senior
Research Engineer, Igor Komsky, who has extensive experience in dealing with aircraft maintenance personnel.
Specifically, this is a proposal for a relatively small initiation project that should be seen as a precursor for a full‐fledged project who’s “GRAND PLAN” has the following components:
• permanently installed microsensors • continuous monitoring in real time with known POD • wireless transmission to a central station • instantaneous interpretation of sensor data • detection of unacceptable material damage at critical high‐stress locations • monitoring of evolution of material damage into critical size • growth prediction by a probabilistic fatigue damage procedure • adjustments for the actual damage state at prescribed intervals • probabilistic forecast of lifetime.
Technical Approach An intelligent health monitoring network should include a network of sensors to monitor several critical parameters that affect structural integrity. The sensors will need to function in an autonomous fashion and conform to stringent restrictions of size, weight, and power consumption. It is also desirable that the sensors be integrated with wireless telemetry for data uplink to a central processing unit. Where possible, the sensors should be either passive, or powered remotely. A significant amount of work has been done in industry, national labs, and universities in the area of smart structural health monitoring systems. This work includes the use of fiber optic sensors, remote monitoring of electrical continuity of thin crack wires, embedded microsensors, embedded piezoelectric sensors, and wireless condition monitoring systems. The principal investigators are familiar with and have contributed to this body of earlier work [1‐4]. The proposed work builds on these earlier efforts, and has three components:
1. Sensor Technology and Measurement Techniques 2. Monitoring of Evolution of Fatigue Damage in Composites and Selection of a Damage
Parameter 3. Use of the Damage Parameter in Probabilistic Failure Analysis
Sensor Technology and Measurement Techniques Sensor development: In this work, we propose to explore the use of microsensors and fiber‐optic sensors for structural health monitoring of aircraft components. Microsensors are typically solid‐state devices built on inexpensive silicon chips, or “coupons.” These sensors can be distributed over critical regions of a structure. Such a coupon may contain a variety of microsensors for detection and cross‐configuration of multiple defect signatures. The microsensors that will be investigated in this work include: wireless SAW sensors, and MEMS‐based accelerometers and acoustic emission sensors. Wireless SAW sensors can be configured as temperature sensors, strain sensor, and with appropriate modification, as flaw‐sizing sensors and acoustic emission sensors. Fiber‐optic sensors that will be considered include Bragg‐grating sensors for ultrasound and acoustic emission monitoring. The emphasis in this work will be on identifying useful
4
applications of such microsensors for condition monitoring of aircraft components, and to select appropriate sensors and systems of sensors for both in‐flight and maintenance facility condition monitoring. A significant part of the project will be concerned with testing of sensors for detection of microscopic damage and quantification of damage and fatigue cracks that evolve out of damaged zones. Acoustic emission sensors will be used to passively monitor damage formation and fatigue crack initiation and growth. These sensors will be complemented by SAW sensors for active interrogation of the damaged zone for quantitative measurement purposes. Figure 1 shows a typical configuration of SAW sensors for sizing of damage and surface breaking cracks. Both SAW sensor arrays can serve as generators and receivers, and the sensor response can be tailored to be sensitive to the flaw size. Figure 1: SAW sensors for flaw sizing Placement of sensors: A related issue that needs to be addressed is the determination of the optimal location for placement of sensors. Sensors can be placed on the exterior of structures, or for composites, they can be embedded. Sensors should be distributed at structurally critical locations and optimized for sufficient sensitivity to monitor incipient failure. For this, the critical areas of flaw initiation need to be identified, and appropriate acoustic emission and SAW sensors should be installed in such locations. At the same time, it is essential that sensors do not adversely affect structural integrity. Interpretation of sensor data. Measurements must be interpreted correctly and accurately to characterize damage in complex structures. Sensor data obtained will be used as input to the probabilistic failure models discussed in section 6. Monitoring of Evolution of Fatigue Damage in Composites and Selection of a Damage Parameter The objective of this task is to characterize and monitor the damage mechanisms and damage evolution in composite laminates and select a damage parameter relevant to the fatigue life of the part. Under quasi‐static loading the failure sequence in composite laminates consists of matrix cracking in the transversely loaded plies, matrix cracking in the axial (longitudinal) plies, local delaminations at the intersections of these sets of cracks, and finally fiber breakage at ultimate failure. Use of Damage Parameter in Probabilistic Failure Analysis Probabilistic failure methods will be applied in conjunction with structural health monitoring so that damaged components can be identified and repaired or replaced. Quantified measures of reliability (provided by probabilistic methods) allow maximization of inspection benefits through optimization of the health monitoring technique. A systematic approach to reliability assessment for a structural component containing damage is illustrated in Fig. 7. As can be seen from the figure, the underlying concept in developing accept/reject criteria for a component is based on detecting and characterizing damage and
S1 S2
5
evaluating it in terms of failure mechanics and a damage growth law. The aim is to determine whether damage in a structure will be sufficiently small that failure can be precluded with a high degree of certainty within a preset time interval.
Fig. 7: Flow diagram of a structural health monitoring system
Plan of Work To demonstrate proof‐of‐concept in the first phase of this project, we propose the following plan of work.
• select a material and geometry: Laminated Graphite Epoxy • set up fatigue tests • instrument the specimen with sensors • define damage parameter to be measured • collect sensor data on‐line • verify damage off‐line • define damage functions • apply probabilistic fatigue procedure • probabilistic forecast of lifetime • verify result
The understanding gained in this work will set the stage for health condition monitoring combining fatigue reliability assessment and inspection methodologies for aircraft components, and will provide important guidelines on computational efficiency requirements for the reliability assessment in relation to sensor efficacy and the processing of sensor data. In a second phase of the project, the focus will be directed toward actual aircraft components. These components will be selected in consultation with engineers from the aircraft industry. The general ideas outlined in this study, suitably adjusted, will be applied to specific components. Interaction with Industry The PIs have ongoing interactions with personnel at the Phantom Works Group of Boeing, St. Louis, in the area of Structural Health Monitoring. Preliminary discussions with that group have provided valuable input into this proposal. Boeing has indicated that they will be willing to provide guidance and assistance (including providing composite specimens) for this project.
Structural Health Monitoring System
Failure Model (Damage vs Time or cycles)
Damage Growth Characteristics
Stress History Current Extent of Damage
Measured Extent of Damage
Probability of Detection of Damage
Failure probability within preset interval
Inspection and Repair at a Maintenance Facility
high
low
6
Expected Outcomes The outcomes of this feasibility study are expected to be: • A better understanding of emerging sensor technologies that are key to the development of in‐flight health monitoring systems. • Improved sensor reliability and transition of the technology to implementation transportation systems. • Demonstration of applicability of sensors to fatigue damage monitoring. • Direct and real‐time use of data in structural reliability assessment. • Integration of sensor data into probabilistic fatigue damage models for assessment of residual life • Training of graduate students and other engineers in a challenging technology that is vital to the future of US aviation efforts. The initiation phase of this project (first 12 months) will be considered successful upon the demonstration of the feasibility of using reliable sensors for acoustic measurements on sample specimens, data interpretation and the use of data in life‐time prediction. The goal for the second phase of this project (2 and 3 years) will be demonstration on real aircraft components, and the transition of lab technology to full‐scale implementation and commercialization, with subsequent application of the technique to achieve improved aircraft safety.
7
(The following two presentations are recognized in the agenda under the heading “Shear Characterization Of Composite Laminates And Structural Adhesives) Development and Evaluation of A V‐Notched Rail Shear Test For Composite Laminates Principle Investigator(s): Dr. Daniel O. Adams, University of Utah Background A V‐notched rail shear test was developed for measuring the in‐plane shear properties of a variety of composite laminates. This test method incorporate S attractive features from both the existing ASTM D 5379 V‐notch beam (Iosipescu) shear test [1] and the ASTM D 4255 two‐rail shear test [2]. V‐notched rail shear test provides a larger gage section than the existing Iosipescu specimen and enhanced loading capability compared to either existing test method. One primary objective that led to the development of this new shear test method was the desire to determine the in‐plane shear modulus and shear strength of multidirectional composite laminates. For composite laminates with angle plies (other than 0° and 90°), the shear strength increases dramatically, making load introduction into the specimen problematic. Both the Iosipsecu and the two‐rail shear test experience loading/gripping problems in their current form that limits their use for high shear strength composite laminates. The Iosipescu shear test method, shown in Figure 1, is capable of measuring both in‐plane and interlaminar shear properties of a unidirectional composite. The relatively small gage section provides limitations for some textile composites and edge loading of the specimen limits the load that may be applied to the specimen without producing localized failures. For unidirectional composites, these limitations are not problematic, since both the in‐plane and the interlaminar shear strengths are relatively low. For multidirectional composite laminates, however, much higher shear strengths are possible, and thus a much higher loading capability is required than possible with the existing test method.
Figure 1. Iosipescu shear test fixture and specimen (ASTM D 5379). The two‐rail shear test fixture configuration, shown in Figure 2, uses a relatively large 3 in. x 6 in. rectangular specimen. Six holes must be machined in the specimen for the bolts used to attach the rails. The determination of shear strength using this fixture is somewhat questionable due to stress concentrations produced in the specimen at the rails. Additionally, slipping between the specimen and the rails has been a limitation with this fixture.
3.0 in.
0.75 in.
90°
8
Hussain and Adams [3] modified the gripping method of the two‐rail shear fixture using roughened rails that are clamped onto the specimen. Using a C‐clamping arrangement, the need
Figure 2. Two‐rail shear test fixture and specimen (ASTM D 4255).
for clearance holes in the specimen was eliminated. With this modification to improve specimen gripping, the two‐rail shear test became a promising test method for high shear strength composite laminates.
Figure 3 (below) Modified two‐rail shear fixture of Hussain and Adams [3].
Figure 4 (above) Modified two‐rail shear fixture and specimen developed for initial investigation.
Initial testing was performed using AS4/3501‐6 carbon/epoxy panels. An initial investigation was performed using a 4.5 in. long by 2.75 in. wide rectangular specimen as shown in Figure 4. This
3 in.
6 in.
4.5 in.
2.75 in.
0.75 in.
9
specimen configuration provided for a 1.0 in. wide gripping area on either side of a 0.75 in. wide by 4.5 in. tall gage section. Excellent gripping performance was obtained for a variety of laminates, including the highest shear strength configuration tested a 16 ply [±45]4S laminate. However, specimen failures were commonly observed adjacent to the fixture rails, indicating the presence of stress concentrations. Three‐dimensional finite element modeling was performed to identify improved specimen configurations. Initial finite element analysis focused on identifying desirable specimen configurations, particularly for testing high shear strength laminates. Two alternate specimen configurations were identified: a tabbed rectangular specimen and a 90° V‐notched specimen. The tabbed specimen utilized the same rectangular geometry of the previous untabbed specimen. The V‐notched specimen utilized the same geometry of the Iosipescu shear specimen, especially the ratio of the notch depth to specimen height ratio. Finite element results were used to investigate the uniformity of the shear stress and the presence of normal stresses in the gage sections of prospective specimen designs. Results showed that the high shear stress concentrations predicted in the untabbed rectangular specimen near the rails were reduced significantly by the addition of bonded tabs. The V‐notched specimen also showed a significant reduction in the shear stress concentration. Additionally this V‐notched rail specimen displayed a much more uniform shear stress distribution and reduced magnitudes of normal stresses throughout the central region of the gage section. Further testing was performed to evaluate the candidate specimen configurations identified through finite element modeling. A total of five 16‐ply AS4/3501‐6 carbon/epoxy laminates were tested with differing percentages of ±45° layers: [08]S, and [(0/90)4]S (0%); [(0/90)2/±45/0/90]S (25%); [0/±45/90]2S (50%); [±45/90/±45/0/±45]S (75%); and [±45]4S (100%). From each laminate, specimens were prepared in three configurations: rectangular, tabbed, and V‐notched. A comparison of the shear strengths obtained from the three specimen configurations for each laminate tested showed that the rectangular specimen consistently produced the lowest shear strengths. Significant increases in shear strength were observed for both the tapered tab and V‐notched configurations. Both the tabbed and V‐notched specimen configurations produced comparable shear strengths for the four laminates with ±45 layers present. However, the highest average shear strength was obtained for both the [08]s and [(0/90)4]s laminates using the V‐notched specimen configuration. Additionally the preparation of a notched specimen was preferred over a specimen with adhesively bonded tapered tabs. Thus, the use of a notched specimen was selected over a tabbed configuration. Further testing and finite element analysis was performed to determine whether the V‐notch specimen configuration shown in Figure 5b was optimal. Several geometric variables were investigated, including the specimen dimensions, notch shape (V‐notch, U notch, and slot notch), notch depth, notch angle, and gage section width. Both finite element analysis and mechanical testing were performed. In summary, the V‐notched configuration shown in Figure 5b was determined to be optimal for a range of laminates. With an optimal V‐notched specimen configuration selected, a modified test fixture was designed to accommodate the shorter specimen height. The new fixture is shown in Figure 6. Limited mechanical test results obtained using this fixture have been compared to those obtained from ASTM D 5379 (Iosipescu) and ASTM D 4255 (two rail shear). These results suggest that this shear
10
test method is suitable for obtaining the in‐plane shear properties of a variety of composite laminates. A draft ASTM standard has been prepared for initial review by members of the Shear Subcommittee of the ASTM D‐30 committee. Emphasis in the third year of this investigation will focus on addressing remaining aspects of the test method required for ASTM standardization.
Figure 6. New test fixture developed for V‐notched shear specimen.
Following the development and evaluation efforts performed during the first two years of this investigation, the V‐notched rail shear test is headed towards ASTM standardization. Two sets of comparison testing were also recently completed, featuring the ASTM D 5379 V‐notch beam (Iosipescu) shear test [1] and the ASTM D 4255 two‐rail shear test [2]. These tests were the first to be performed with the most recent V‐notched rail shear test fixture (Figure 6). As a result of these recent activities, several issues were identified that had not been addressed in the development and evaluation activities to date. These issues requiring further investigation will be addressed in this phase of the research investigation. Each issue is listed and described briefly below. Objective
1. Establishment of recommended bolt torques – An investigation is required using the new, reduced sized V‐notched rail shear test fixture (Figure 6), which utilizes a smaller number of larger diameter clamping bolts than the previous fixture. An in‐depth investigation will be performed to investigate the acceptable range of bolt torques for different materials, laminate thicknesses, stacking sequences (ie. percent ± 45 degree plies), and for different test temperatures. 2. Development/implementation of a specimen alignment/mounting fixture – A fixture/jig is needed for assuring properly positioning the specimen and fixture halves during bolt tightening to ensure proper test fixture/specimen alignment. A positioning fixture will be designed and fabricated for initial evaluation during the bolt torque investigation described above. Any modifications to this positioning fixture will be implemented and further evaluated in subsequent tests. This positioning fixture will be incorporated into the draft ASTM standard. 3. Determination of recommended strain gage shapes/sizes – Further research is required to investigate the best suited strain gage sizes and shapes for modulus determinations on a variety of different composite laminates, and for textile composites with different unit cell
11
sizes. A small, centrally located strain gage is the most economical option, but a thin, elongated shear strain gage is useful for recording the average strain between the V notches. For textile composites, a wider strain gage may be necessary to obtain the reduce variability due to strain variations within the unit cell structure. Finite element analysis will be performed to investigate the effect of strain gage size and shape on measured shear modulus for a variety of composite laminates. Based on these analyses, an experimental investigation will be performed to evaluate the best suited strain gage patterns for at least three different laminates and for at least two different unit cell sizes of woven composites. Recommended strain gage sizes and shapes determined from this task will be incorporated into the draft ASTM standard. 4. Evaluation of the elevated temperature/low temperature performance of the flame‐sprayed gripping surfaces – Based on other experiences with flame‐sprayed gripping surfaces at elevated/low temperatures, further investigation is required to ensure the suitability of the V‐notched rail shear test at elevated and low temperatures. Condensation on the gripping surfaces following testing at low temperatures and residue retained in the gripping surfaces at elevated temperature may require a modified fixture preparation/post test cleaning procedure than normally followed. Testing is proposed at elevated temperature/wet and low temperature conditions to evaluate the specimen gripping surfaces. Recommendations based on these tests will be incorporated into the draft ASTM standard.
12
Investigating the Thin‐Film versus Bulk Material Properties of Structural Adhesives Principle Investigator(s): Dr. Daniel O. Adams, University of Utah Currently, there appears to be considerable confusion and a lack of consensus on whether mechanical properties obtained from the testing of bulk adhesive specimens may be used in the design and analysis of thin film adhesive joints. At the source of this confusion is the question of whether the mechanical properties of an adhesive are different when used in a relatively thin bondline, or “in‐situ” versus a relatively thick bondline. The proposed research is intended to provide a definitive answer to this question using a combined experimental and computational approach. Background Certain adhesive test methods, such as the lap joint test, are known to produce nonuniform states of stress in the adhesive bondline and thus do not provide a simple or straightforward measure of the stiffness or strength properties of the adhesive. However, these mechanical properties of the adhesive commonly are required for design and analysis purposes when the adhesive layer is modeled. Test methods such as the[ lap joint test consider the adhesive in its thin film or ʺin‐situʺ form. Another approach to determine the mechanical properties of an adhesive is through ʺbulkʺ adhesive tests, where an entire specimen is cast or machined from the adhesive material. Although no ASTM standard tests exist for bulk adhesive testing, many of the standards included in ASTM Volume 8 (Sections 1 through 4) for plastics or Volume 9 (Sections 1 and 2) for rubbers may be adapted to test the properties of bulk adhesives. Tensile testing of bulk adhesives is relatively straight‐forward and may be performed using either cast or machined tensile specimens. Shear strength and shear modulus determinations of the bulk adhesive may be accomplished using several test methods, including solid rod torsion testing or using the V‐notched Iosipescu shear test method (ASTM D 5379) or the recently developed V‐notched rail shear test method. In a recent research investigation, the author successfully characterized both the tensile and shear stress‐strain response of structural adhesives using bulk adhesive specimens for use in modeling tabbed composite specimens [1]. Although hundreds if not thousands of adhesives have been characterized in thin film or “in‐situ” form and many test laboratories and researchers (including the author) have performed bulk adhesive testing, there have been surprisingly few investigations that have addressed the thin‐film versus bulk material properties of structural adhesives. The author became aware of this lack of attention recently when asked to provide a state‐of‐the‐art assessment of thin film versus bulk adhesive testing as part of an invited book chapter on adhesive test methods [2]. A review of the open literature revealed that among the limited studies that have been published, there is considerable confusion and a lack of consensus on whether mechanical properties obtained from the testing of bulk adhesive specimens may be used in the design and analysis of thin film adhesive joints. Dolev and Ishai [3] conducted bulk and in‐situ adhesive tests to compare mechanical properties under different states of stress. Good correlation between in‐situ and bulk shear yield strength and elastic modulus was obtained. The authors concluded that elastic and strength properties of an in‐situ adhesive may be determined by bulk adhesive testing. In contrast, Peretz [4] concluded that the in‐situ adhesive shear modulus increased with increasing adhesive thickness up to the bulk material shear modulus. The shear strengths obtained from thin adhesive layers were similar to those obtained from bulk testing. Lilleheden [5] performed a detailed experimental investigation of modulus variations in adhesives for
13
differing adhesive thicknesses using a modified lap adherend specimen and found no difference in the measured moduli of the adhesive between the thin‐film and bulk forms. Following this review of the open literature, the author concluded in his book chapter [2] that there was no clear consensus on the equivalence of thin film versus bulk adhesive testing. One explanation that has been offered for the existence of differences in mechanical properties of in‐situ versus bulk adhesive is the presence of a diffuse region or “interphase” at the boundary between the adhesive and adherend [6]. Others, however, have attributed differences in mechanical properties to factors such as variability in adhesive casting and curing conditions, lack of a well‐defined state of stress, and inadequate methods of strain measurement [5]. Clearly, a complex state of stress is produced by the geometric discontinuities in many in‐situ test configurations and by the drastically different material properties of the adhesive and adherends. Thus, it is not clear whether differences in material properties are due to material‐related differences or test/measurement‐related differences. The goal of the proposed research program is to determine conclusively whether the mechanical properties of structural adhesives differ when in thin film (in‐situ) versus bulk forms. As a result, this research will address whether bulk adhesive properties are suited for use in the design and analysis of adhesively bonded structures. Objective A combined experimental and computational approach is proposed to evaluate the thin film versus bulk mechanical properties of structural adhesives. Both bulk and in‐situ adhesive testing will be performed using two or three different structural adhesives. Selection of adhesives will be made following consultation with FAA personnel. Emphasis will be placed on selecting adhesives that are commonly utilized, that show different material responses, and if possible, have been at least partially characterized in the past. Both paste and film adhesives will be considered. Bulk adhesive testing will focus primarily on the stiffness and strength properties under shear loading. Based on the authors past success, it is proposed to use either the V‐notched Iosipescu shear test method (ASTM D 5379) or the recently developed V‐notched rail shear test [7] for characterizing the shear response of the bulk adhesives. Note that the Iosipescu shear test may be preferred due to the fragile nature of the bulk adhesive specimens. V‐notched bulk adhesive specimens will either be machined from a flat adhesive plaque or cast into a machined mold to produce the final specimen configuration. One of the in‐situ shear test methods to be used is a V‐notched shear test method, either the Iosipescu shear test method (ASTM D 5379) or the recently developed V‐notched rail shear test. To explore the thin film shear properties, V‐notched specimens made from either metallic, plastic, or composite, will be cut through the central notched section and adhesively bonded using the adhesives proposed for investigation. The thickness of the adhesive bondline will be varied, but the midplane of the adhesive bondline will be centered between the notches. The use of the V‐notched shear test should provide a definitive answer to the question of whether shear modulus and shear strength properties vary as the adhesive bondline increases to the limiting case of a bulk adhesive test. Although the use of a V‐notched shear specimen produces a highly uniform state of pure shear stress in the gage section between the notches, the use of an adhesively bonded specimen with metallic, composite, or plastic adherends will likely produce stress concentrations. To understand the state of strain present in the in‐situ V‐notched specimens, Moiré interferometry will be utilized. This full‐field, high‐sensitivity experimental technique is well suited for such determinations, and the author has successfully used Moiré interferometry
14
for a number of comparable investigations, including an investigation of the state of strain in adhesive layers in tabbed composite specimens. The use of Moiré interferometry will ensure that proper shear strain measurements will be used in the calculation of shear modulus. Additionally, Moiré interferometry will help to determine whether failures in the in‐situ specimens, especially those with relatively thin adhesive bondlines, are resulting from uniform states of shear stress. The proposed detailed experimental investigation, coupled with finite element analyses of the test configurations being pursued, will be used to distinguish whether observed differences in mechanical properties of adhesives can be attributed to the state of the adhesive ‐ thin film versus bulk, or a combined state of stress produced in the commonly used in‐situ test configurations. Thus, the proposed research will determine whether differences in adhesive material properties are due to material‐related differences or test/measurement‐related differences and thus whether the mechanical properties of structural adhesives differ when in thin film (in‐situ) versus bulk forms.
15
Effect of Repair Procedures Applied to Composite Airframe Structures Principle Investigator(s): Dr. John Tomblin, Executive Director, NIAR; Lamia Salah, Manager, Fatigue & Fracture Lab; Mike Borgman, The Boeing Company With the increasing use of composite materials in aircraft structural components, it has become essential to answer not only the fundamental questions related to the proper repair methods/systems to restore the aircraft part structural integrity but also the question of how long the repair will last under the specified design conditions and what are the most critical factors affecting the static performance and the long‐term durability of the repair. The lack of fatigue data to assess the durability of repairs, added to the lack of confidence in bonded repairs especially when dealing with large damaged areas, has led to the use of fasteners to reinforce the adhesively bonded areas in some cases. The ultimate goal of a bonded repair is to achieve a good level of confidence in bond strength as well as the ability to avoid long‐time service failures of these joints. Objective The objective of this research program is to assess the effects of different variables on the static and fatigue performance of scarf repairs applied to moderately thick composite laminates representative of the 7E7 fuselage configuration and the long time durability of these repairs especially when a faulty process has been implemented and was not detected by NDI. The main research program will be divided into three tasks. • The first task consists of investigating the effects of different variables on the strength performance of repairs applied to moderately thick solid laminates. Variables considered include different substrate stiffness, lap length, laminate thicknesses. The only loading mode considered will be tension. Coupons will be tested for static and fatigue properties under room temperature c bond ETW. Constant amplitude fatigue will be conducted on the coupons and residual strength will be measured after fatigue cycles. • The third task consists of OEM as well as field repairs will be considered. The goal is to evaluate the fatigue life knock‐down when one of the repair steps was not properly implemented. The project will also be investigating the effects of the process parameters on the strength and durability of repairs. • The second task will consist of a validation of safety standards required for composite repair and inspection technicians as related to composite repairs. The goal of this task will be to evaluate existing CACRC standards as related to technician skill level using different repair geometries and establish the value of existing CACRC standards for composite repair technician qualification. • FEM validation of experimental results will be conducted to predict scarf joint failure.
Research Methodology The proposed methodology consists of generating static and fatigue data to understand the effects of different variables on the performance of the repair and the basic degradation mechanisms that bonded repairs undergo under sustained long‐term mechanical loads. The ultimate goal will be to assess the level of criticality of each step in the repair, to determine the fatigue life, and/or the difference in the fatigue performance of weak joints and finally to be able to make recommendations that will increase the level of confidence in bonded repairs.
16
Coupon Fabrication Coupons will be manufactured at the OEM (Boeing, Wichita) to replicate actual aircraft parts. All repairs will be implemented at NIAR according to the OEM recommended repair procedure. Weak repairs will be implemented as follows: light hand sanding instead of grit sanding will be used to simulate poor surface preparation, soaking the coupons in an environmental chamber prior to bonding the repair will be used to simulate pre‐bond moisture, modifying the heat blanket cure cycle will be used to simulate an under‐cured patch and finally failing to thoroughly clean the surface prior to bonding will be used to simulate a contaminated repair/ parent structure interface. All coupons will be subsequently machined and inspected using through transmission ultrasonics for possible defects induced during manufacture. Coupon Configuration/ Experimental Set‐up/ Methodology All coupons used for this program will be unidirectional tension solid laminates tabbed at the ends for loading purposes. Laminates of different thicknesses and substrates of different stiffnesses will be considered for the purpose of this investigation
Test Set‐up Validation of Safety Standards Required for Composite Repair and Inspection Technicians The commercial aircraft composite repair committee (CACRC) developed an industry standard for the certification of composite repair technicians. A previous study has indicated the quality of training and experience of repair technicians may have a much larger role in the technician’s successful development of a repair (see Figure 4 – ref. [2]). This study has indicated the quality and reliability of a composite repair is much more directly linked to the skills/knowledge of the repair technician than was previously believed and specified in the CACRC standard. This task will address this issue to verify if the proposed CACRC standards for composite repair technicians are appropriate. This research validation task will re‐evaluate the existing CACRC technician standard in order to provide information leading to a new FAA composite repairman qualification for 14 CFR Part 65. In addition, a proposed regulatory activity (Part 66) by FAA, seeks to create many new standards for mechanics and certain specialists.
17
The main goal of this task will be to generate a report evaluating the technical value of the CACRC standards for technicians performing aircraft composite material repair and inspection. The report’s task qualification evaluation will be formatted to be suitable for incorporation into FAA Advisory Circulars and related regulations. Expected Outcomes A working group comprised of representatives from the FAA, WSU, and industry, will develop a detailed plan and schedule for the research program. NDI inspection of repaired coupons will be conducted prior to destructive mechanical evaluation as well as specific bonded repair coupon configuration for the evaluation of weak bonded repairs. Results of experimental tests and supporting data analysis will be detailed in a final report. The program will also evaluate the technical value of the CACRC standards for technicians performing aircraft composite material repair and inspection. The report’s task qualification evaluation will be formatted to be suitable for incorporation into FAA Advisory Circulars and related regulations.
18
VARTM Variability and Substantiation Principle Investigator(s): Dirk Heider, Assistant Director UD‐CCM,Associate Professor Electrical & Computer Engineering Department, University of Delaware and Crystal Newton, Scientist, Center for Composite Materials, University of Delaware Background Vacuum‐assisted resin transfer molding (VARTM) has the potential advantages of relatively low cost with sufficiently high volume fractions of reinforcement and can be readily applied to large‐scale structures. However, for many aircraft applications, VARTM does not currently provide sufficient repeatability or control of variability. Such variability is commonly observed when processing with the VARTM process. In order to routinely produce VARTM parts of aircraft quality, the variability must be understood.
Objectives The long‐term objectives are repeatability equivalent to autoclave processing with specific properties (property/weight) that are close to autoclave processed part levels at a lower cost. • There are many factors that influence the variability of the final part. The factors that play a major role in the cause of this variation need to be identified and the causes and effects of changes in these factors understood. Three main VARTM process variations will be considered:
1. The SCRIMP process, patented by TPI Composites is a vacuum infusion process using a high‐permeability layer to rapidly distribute the resin on the part surface and then allow through‐thickness penetration;
2. The CAPRI process, patented by Boeing Co. is a SCRIMP variation where a reduced pressure difference is used to minimize thickness gradients and resin bleeding;
3. The VAP process is another SCRIMP variation, patented by EADS where a air‐permeable membrane is used on top of the distribution media to allow continuous and areal venting reducing void content and creating a robust process variant.
Repeatability
Cost
Actual
Goal
A lower variation (higher repeatability) in properties improves the allowable design
Cost
Property
Weight
Actual
VARTM
Autoclave
19
• Use Liquid Injection Molding Simulation (LIMS), a comprehensive simulation software for mold filling, which can explore a variety of filling scenarios and provide for the investigation of the influence of various processing conditions in VARTM and the SMARTMolding Intelligent Process Control (IPC) system, for which Boeing is a beta‐site, will be used, enabling material, process, and part traceability along with semi‐automated material lay‐up, automated resin mixing, and resin infusion and control of dwell times and cure cycles. The automation capabilities enable monitoring of cycle times for all processing steps, sensing of the important process parameters through embedded sensors and QA/QC of the complete process.
• The proposed work will statistically evaluate material, process, and quality variation. A link
will be developed between incoming precursor material quality and process variation to understand quality drivers.
• Develop a model to predict part variation and/or optimize material/process requirements. Once the sources of variability are understood, substantiation issues for a risk‐mitigated approach to incorporate VARTM parts of increasing size and complexity into large commercial transport aircraft will be addressed.
• A step‐by‐step development from flat panels to structural elements that include stiffeners and core materials will be used at UD‐CCM.
• Following completion of the initial phase, efforts will coordinated with an industrial partner to approach subcomponents and more complex levels of the building block approach.
Objective • Aerospace materials such as resin, fabric and core material for the VARTM process will be
reviewed and one candidate system will be down‐selected. • Flat panel with uniform resin input from one end will be used to evaluate the processing
conditions and part performance. The three processes will be evaluated on process repeatability, part quality and mechanical performance.
• Process repeatability will evaluate the resin fill times and flow rates, part quality will asses the dimensional tolerances and properties such as void content and fiber volume fraction, and mechanical performance will evaluate tensile, compression and shear properties.
• The developed database will be used to compare the processes and to understand current state‐of‐the‐art VARTM capabilities to produce aerospace structures.
Figure 1. UD-CCM’s Tool Kit of VARTM Capabilities
20
Expected Outcomes Database of material properties across the three main VARTM variants for aerospace structures This program will create a model to predict part variation and optimize material/process requirements. It is anticipated that this model will interface with UD‐CCM’s tool kit of design and control software.
21
Wednesday, May 25, 2005
Laminate Statistical Allowable Generation for Fiber Reinforced Composite Materials Principle Investigator(s): Waruna Seneviratne, Manager, NIAR Structures Lab Background The proposed research will focus on using a reduced testing methodology related to the building block approach for these laminate tests. The proposed research will also interface with other past and ongoing efforts under investigation by the FAA. Previously, FAA document DOT/FAA/AR‐00/47 entitled “Material Qualification and Equivalency for Polymer matrix Composite material Systems” was funded under the NASA Advanced General Aviation transport Experiments (AGATE) and a subsequent policy was issued from the Small Airplane Directorate in Kansas City, MO which allowed this document to use as a means of compliance with Federal Aviation Regulations. Presently, a companion effort is ongoing which is funded by the FAA involving the standardization of Material and Process Specifications. The proposed research represented by this proposal will interface with both efforts and is primarily meant to provide guidelines and develop statistical methodology for the use of laminate (notched and un‐notched) data in future designs while maintaining an acceptable level of safety. In general, analysis alone is not considered adequate for the substantiation of composite structures used in aircraft designs and commonly a “building block approach” is used to in conjunction with analysis. The building block approach basically consists of testing at lower levels of the building block to characterize material performance, usually at the coupon level, and gradually increasing complexity and structural detail, usually at the element or subcomponents level, as one progress towards the full‐scale components test. Figure 1 shows the building block approach schematically as a pyramid of tests.
22
Objective Referring to the building block approach, the major objective of the work proposed in this statement of work will attempt to link the steps at the lowest level of the building block consisting of lamina tests with the next level in the building block consisting of notched and un‐notched laminate tests. In order to achieve the background for the statistical assumptions required to reduce the number of tests at the next higher level, a three batch laminate test program is proposed which is outlined in the next section. Using the larger number of tests at this higher, laminate level, comparisons will be made which evaluate the reliability of this reduced testing methodology as one proceeds higher in the building block. A number of other sub‐objectives will also be addressed in the proposed program. The overall objectives and goals for the proposed research are outlined as follows:
1. Generate a multi‐batch, nested database of notched and un‐notched laminate properties for two typical, commonly used composite material systems (unidirectional tape and plain weave fabric) in which lamina level statistics already exist.
2. Evaluate common ASTM test methods that are used to generate the notched properties for database generation and develop guidelines for use and make recommendations to the Military Handbook 17 committee for possible inclusion into the handbook.
3. Generate typical effects of width, thickness and notch diameter with respect to ultimate strength for inclusion into the laminate database.
4. Using the lamina and laminate databases, compare multiple statistical methodologies that are currently proposed which show the effects of using smaller sample sizes of the statistical reliability at the laminate level of the building block.
5. Generate notched and un‐notched laminate data for inclusion into Military Handbook 17 that may be used in future design efforts.
23
(6) Recommend a reduced laminate qualification test matrix and statistical methodology for inclusion into Military Handbook 17 and publish a guideline and methodology document. Technical Approach Using the objectives described in the previous section, the proposed research program may be broken down into five tasks, which are summarized below. The program data generation will be focused on carbon reinforced material systems with a unidirectional tape and two plain weave fabrics in two different resin systems from two different material suppliers. The materials proposed for this investigation are listed in Table 1.
Task 1: Data Generation The primary scope of this task will be data generation for statistical analysis using the two material systems described previously. The tests proposed are based upon tests suggested in section 7.2 of MIL‐HDBK‐17. For ease of use, the proposed test matrices are separated by the two materials systems, i.e., unitape and plain weave, and are divided into tension, compression, bearing, and rail shear. Task 2: Specimen and Joint Geometry Effects The main objective of this task is to generate these strength reduction curves for each specific test method (open hole, tension, and 50 percent bypass tension) for a range of W/D and t/D. Task 3: Data Reduction and Statistical Methodology Development Once the testing is completed for Task 1 and 2, the multi‐batch data will be reduced to generate statistical allowable using traditional analysis methods. These methods are (1) Allowable generation using MIL‐HDBK‐17 methods (2) Allowable generation using the method recommended in DOT/FAA/AR‐00/47 Task 4: Data report in MIL‐HDBK‐17 format This task will primarily consist of using the generated data for this program and making it available for use in future programs. All data generated from this program will be submitted to the MIL‐HDBK‐17 Data Review committee for inclusion into the handbook for future use.
24
Task 5: Methodology and Guideline Report Using the results from the preceding tasks, a generic document will be written to accompany DOT/FAA/AR‐00/47, which covers laminate testing. This document will include test method guidelines, test matrices and statistical reduction procedures to be used for future material qualifications. It is also intended for this document to filter into various sections of MIL‐HDBK‐17 as well. Expected Outcomes Multi‐batch, laminate data resulting from the material systems mentioned in Task 1 and 2. All tests will be submitted to MIL‐HDBK‐17 for inclusion into the handbook and used for developing statistical methodologies, which may be applied for future testing and material qualifications. This project will also establish possible statistical engineering link between lamia and laminate data.
25
Combined Global/Local Variability and Uncertainty in Integrated Aeroservoelasticity of Composite Aircraft
Principle Investigator(s): Eli Livne, Professor, Department of Aeronautics and Astronautics, University of Washington
Background
The rigorous design, analysis, testing, and certification effort in the development of a new airplane comes to an end once the airplane enters service or some set time thereafter. During service, changes in airplane characteristics from the certified original configuration are usually addressed by maintenance procedures aimed at detecting such changes and by guidelines to determine those that are acceptable and those that require corrective action. In addition to maintenance, possible variations of airplane structural characteristics over time are addressed during the design phase to obtain robust design. Two technological developments have made the study of airplane variability problem more worthwhile: the increasing usage of composite materials in load‐bearing major airplane components and the increasing power and authority of digital active control systems. With composite structures, the potential sources of structural variation and deviation from original characteristics of an airframe over its lifetime in service are numerous: moisture absorption, crack and delamination progress, softening of bonded joints, damage due to impact, and material degradation resulting from radiation and other environmental effects. These variations and deviations from the nominal design may lead to stiffness and mass variation with time. They can start as localized effects, but develop to potentially affect the overall stiffness and mass distributions of major structural components. This may lead to increased loads caused by changes in aeroelastic deformation under load and to aeroelastic instabilities such as divergence and flutter. The problem seems to be particularly severe for composite control surfaces. Over time, moisture absorption can lead to increased mass and inertia and lack of balance, and wear of hinges and linkages can lead to reduced stiffness or nonlinear stiffness of hinges. The combined effect might lead to flutter or limit cycle oscillations. With digital flight control systems, the ease with which control laws can be changed throughout the lifetime of an airplane has greatly increased. Pilot feedback, avionics, and actuation system changes and changes in operational requirements or mission needs all lead to changes in control laws as the airplane is modified over time. The problem is that with high‐authority active control systems the control system is an integral part—with the loads, structures, and flight mechanics models—of the simulations and tests that demonstrate fatigue life for the airframe. Modification of control laws means modification of airplane response, changes in dynamic loads and loads spectra, and resulting changes in fatigue life. The problem has been encountered in modern fighter aircraft, where late changes in control laws were found to lead to major effects in fatigue life. Objective
• Develop better understanding of effects of local structural and material variations on overall aeroservoelastic integrity.
• Develop computational tools (validated by experiments) for local/global linear/nonlinear analysis of integrated structures/aerodynamics/control systems subject to multiple local variations/ damage.
26
• Establish a collaborative expertise base for future response to FAA and industry needs, R&D, training, and education.
Payoffs
• Better understanding of the underlying physics. • Tools for rapid evaluation of structural uncertainty and digital flight control system
modifications on load redistribution, local stresses, and resulting aeroservoelastic integrity.
• Identification of damage sensitive areas. • Development of cost‐effective fleet maintenance for a consistent level of safety. • Foundation for future extension to advanced structures technology.
Approach Computational capability development will focus on quantification of effects on stiffness of key local effects in composite structures, global aeroelastic/aeroservoelastic (ASE) analysis capable of evaluating variations and uncertainty to such local effects, and integrated local/global modeling capability for uncertain composite structures. Capabilities for simulation of the effects of control surface nonlinearities on aeroelastic and aeroservoelastic behavior of full scale airplanes will be developed and used to study effects of nonlinearity and uncertainty mechanisms and guide maintenance practices. Simultaneously, an experimental structural dynamic/aeroelastic testing capability for composite airplane structure models will be developed at the UW, and tests will be planned and conducted to study effects of damage on stiffness of components and models. The analytical, numerical, and experimental technology developments will all be done in close collaboration with The Boeing Company.
27
Approach (continued): 1. Evaluate state‐of‐the‐art and report. 2. Create Aeroelastic (AE) / Aeroservoelastic (ASE) sensitivity analysis tools for sensitivity‐
to‐damage evaluation. 3. Create Aeroservoelastic uncertainty analysis tools for composite aircraft (statistical and
deterministic). 4. Develop tools for identifying most critical scenarios due to local structural variation:
aeroelasticity (AE). 5. Develop tools for Identifying most critical scenarios and for control system modification
with minimal effects on fatigue life: aeroservoelasticity (ASE). 6. Develop tools for simulating effects of uncertainty and nonlinearity of control surface
attachments and for devising maintenance guidelines. 7. Build a structural analysis / aeroelastic experimental capability for composite structures
at the University of Washington. 8. Create an Industry‐quality expertise base for AE and ASE of composite aircraft in
anticipation FAA and, possibly, NTSB future needs.
Statement of Work Work is planned to progress simultaneously in task groups A, B, and C.
Group A: Tasks that focus on the integrated linear local ‐ global aeroelastic problem: linking of local damage mechanisms to local component stiffness changes, sensitivity of global linear aeroelastic behavior to damage and uncertainty at the local level, uncertainty and variability of global aeroelastic behavior, and identification of worst‐case damage scenarios. Group‐A work includes investigation of local damage effects on aeroelasticity in load‐carrying structures.
Group B of tasks focuses on the nonlinearities of control surfaces on their hinges and actuators and gradually builds the capability for simulating full‐scale real aircraft for subsequent investigations.
C‐tasks: Tasks associated with the construction of the experimental capabilities.
Expected Outcomes 1. Selected study cases (with FAA Grant Monitor concurrence). 2. Two Interim reports summarizing one‐year efforts—12 and 24 months after go‐ahead. 3. A final report in the FAA approved format will be submitted covering the three‐year
effort 34 months after go‐ahead. This report will be published as an FAA report.
28
Evaluation of Friction Stir Weld Process and Properties for Aerospace Application Principle Investigator(s): Dale Cope, Director of NIAR’s Aging Aircraft Research Laboratory and Advanced Joining Technology Lab; Bryan Tweedy, Research Associate; Christian Widener, Research Associate; Adam Jahn, Graduate Student Background Invented in 1991 by The Welding Institute (TWI), friction stir welding (FSW) is a solid state joining process, which is emerging as a viable manufacturing process with many applications. Due to interest of the local aviation industry, the National Institute of Aviation Research (NIAR) in Wichita Kansas conducted a feasibility study of the friction stir welding process. The study focused specifically on the development of FSW with respect to aerospace applications and to the feasibility of using FSW as an aircraft manufacturing technique. 2XXX and 7XXX series aluminums were identified as aircraft specific materials and served as a focal point for the study. The research proposal for the “Evaluation of Friction Stir Weld Process and Properties for Aerospace Application” consists of 5 major focus areas: (1) material properties testing, (2) destructive and non‐destructive inspection techniques, (3) development of pin tools and process 3 parameters, (4) modeling of the FSW process, and (5) manufacturing and testing prototypes of complex aircraft components. The greatest hurdle to the introduction of friction stir welding into aerospace applications is the lack of standard properties data. The principal goal of this proposed research will be the development of that data for use in the design of FSW aerospace vehicle structures. This data will require properties data for strength, fatigue life, fracture toughness, fatigue crack propagation, corrosion fatigue, and environmentally assisted cracking. The aircraft industry uses 2XXX and 7XXX series aluminum extensively. Both of these alloy series traditionally have been considered difficult to weld by conventional fusion methods. This portion of the research program seeks to determine the material, fatigue, and corrosion properties of 2XXX and 7XXX butt and lap welded aluminum sheets with thicknesses ranging from 0.040‐in to 0.125‐in, which are of particular interest to the aircraft industry. In conjunction with the materials properties testing of basic joint configurations at the coupon level, built‐up structures such as flat and curved stiffened panels will also be investigated. Evaluations of their strength in tension, compression, and shear, along with damage tolerance will be made and compared to similar riveted panels. Objectives – The objectives for the evaluation of FSW for aerospace applications include:
• Determine the tensile strength properties of FSW butt and lap joints in common aerospace alloys, beginning with Al 2024 and 7075; • Characterize the fracture properties of those joints; • Develop S‐N curves for thicknesses and joint configurations of interest; • Assess the fatigue crack growth rates (da/dN vs. ∆K) in the joint and in the heat affected zone adjacent to the joint; • Determine the effect of exfoliation and stress corrosion cracking on FSW joints, and investigate methods to enhance their resistance; • Fabricate and test flat and curved stiffened panels for comparison with riveted structure; • Develop finite element models for predicting the strength of FSW built‐up structures in tension, compression and shear.
29
Testing Methods – The material properties will be determined by following methods. • Tensile properties will be determined using ASTM E8‐03 Standard for the different thickness of welded sheets. • Determination of the fracture resistance of welded sheets for both the thin and thick sheets will be determined using C(t) and M(t) coupons. • Fatigue crack propagation rates for the weldments of different thicknesses will be established using K decreasing or K increasing tests, depending on the range of da/dN vs. ∆K curve considered important to the industry. • Effect of corrosion on the weldments will be determined by determination of the fatigue strength and crack propagation characteristics in salt fog and high humidity environments. Other environments if required could also be used. Materials – Initial material testing will be carried out on 2024 and 7075 material with sheet thicknesses of 0.040‐in, 0.063‐in, 0.125‐in and 0.25‐in. All tests will be carried out on sheets welded using procedures developed by and to flaw acceptance criteria agreed to by the industrial group. Fatigue Tests Environments – Tests will be carried out at ambient temperature to establish the fatigue performance of the selected materials in comparison to base material. The simulated corrosive environment will be salt fog and 95% relative humidity. Other corrosive environments can be considered if so recommended. Destructive Testing Methods for Standardization Four destructive testing methods will be conducted in order to evaluate weld properties and weld quality. These methods are: 1) Metallographic Examination 2) Tensile Testing 3) Three Point Bend Testing 4) Four Point Bend Testing Non‐Destructive Inspection of FSW Structure Non‐destructive testing will be performed on welded panels and structure in order to establish their effectiveness in finding weld defects. The methods of inspection that yield the most reliable results will be evaluated for overall effectiveness, and standard testing methods may be developed if they differ from already established standards. The non‐destructive tests would consist of the following: 1) Dye Penetrant Inspection 2) Phased Array Inspection 3) Eddy Current Inspection Surface Treatments for Properties Enhancement In this section of FSW testing, various types of surface treatments will be evaluated in order to determine if they can improve material property characteristics such as corrosion resistance. Initially, the four types of surface treatments that will be evaluated are: 1) Surface Milling 2) Shot Peening 3) Low‐Plasticity Burnishing 4) Laser Shock Processing
30
Investigation II: Tool and Process Parameter Development, Weld Process Modeling, and Welded Structures Prototype and Testing This investigation will enhance and be performed in conjunction with the first investigation, and it will mainly support specific applications for the aircraft industry. Described in the following subsections, the proposed research would consist of a) Tool and Process Parameter Development b) Weld Process Modeling c) Welded Structures Prototype and Testing Tool and Process Parameter Development One of the areas of FSW requiring significant development is the standardization of tooling. Pin tool design is as important as the weld parameters that are used to create the weld since both will change the character of the weld metal. It is also the area of FSW with the least amount of published data, since specific tool designs tend to be considered proprietary information. However, since the quality of the weld may depend also on tool design, then the standardization of those designs is needed in order to develop standards for creating acceptable quality welds. Once tool designs have been standardized, then process parameters can also be standardized for given tool designs. Weld Process Modeling The stirring action of the FSW tool creates a very complex three dimensional flow profile that depends on a large number of variables, and this flow profile is not yet fully understood. It is very important, however, to gain a better understanding of the complex flow phenomenon in order to better predict the effects of changes in weld parameters without the need for testing every permutation. Many other complex flow phenomena are now being successfully modeled with the aid of recent advancements in computing power and software modeling capabilities; FSW research could also greatly benefit from the advent of a successful flow model. Welded Structures Prototype and Testing Another area that is undeveloped is the testing of FSW structures. The complex states of stress that exist under real loading conditions can produce results that are difficult to predict with plane stress or plane strain models. Therefore, prototype aircraft structural components will be constructed in order to develop standardized testing methods of welded versus riveted structures. Static and fatigue tests will be conducted on typical joint configurations along with actual built‐up components, such as flat and curved stiffened panels that simulate fuselage skin panels and wing structure. Through the research in this area, a better understanding of crack growth and damage tolerance of welded structures for aircraft application will be developed. Expected Outcomes– Evaluation of FSW for aerospace applications will include: • Investigate sensitivity to various process parameters:
• Appropriate post weld heat treatments for the enhancement of corrosion resistance of FSW joints in 2024‐T3 and 7075‐T73/T6, and dissimilar joints of those same two alloys; • Joint properties of 0.125” FSW butt welds of those same alloys, including tensile, fatigue, fracture toughness, fatigue crack propagation, stress corrosion cracking, and microhardness properties;
31
• Joint properties of 0.040” FSW lap welds tested in tension and fatigue, including an investigation into methods for enhancing corrosion resistance; • Fabrication and testing of butt and lap welded flat stiffened panels for comparison with riveted panels tested in tension, compression, shear, and damage tolerance.
• Standardizing a process for qualifying a friction stir weld structure • Standardization of property requirements • Access possibilities of creating standard data (MMPDS‐MIL‐HANDBOOK 5)
32
Fatigue Behavior of VARTM Manufactured Affordable 2x2 Biaxial Carbon Braided Composites Principle Investigator(s): Ajit Kelkar, North Carolina A & T State University; John Whitcomb, Texas A & M University Overview Braided composites have good properties in mutually orthogonal directions, more balanced properties than traditional tape laminates, and have potentially better impact resistance due to the interlacing. Another benefit is reduced manufacturing cost by reducing part count. Because of these potential benefits these braided composites are being considered for various applications ranging from primary/secondary structures for aerospace structures. These material systems are gaining popularity, in particular for the small business jets, where FAA requites take off weights of 12,500 lb. or less. The advantages of braids are a result of the interlacing, but this interlacing also complicates the stress analysis and so makes it more difficult to design with confidence. A complex three‐dimensional stress distribution exists in a ʺsimpleʺ plain weave even for uniaxial loading. However, it is important to recognize that the complexity depends strongly on the waviness of a particular weave. More important, not every complication is necessarily important in determining performance. A critical part of the work proposed herein will be identification of the critical complexities that must be considered. Obviously, this requires an integrated experimental and analytical investigation. The other obstacle to widespread use of weaves is shared by all composites: the cost of manufacturing. The conventional autoclave processing which consists of a vessel which is pressurized internally up to 5 bar (~ 100 psi) and then the contents are heated. The main aim of the autoclave process is to manufacture the laminate with uniform thickness and to ensure minimum porosity. The drawback of the autoclave processing technique is that the pressure vessel must be sufficiently large to accommodate large components. This in turn results in a high capitalization cost and stringent pressure code regulations. The new process, Vacuum Assisted Resin Transfer Molding (VARTM), is low cost, affordable and suitable for high volume manufacturing environment. Recently the aircraft industry has been successful in manufacturing wing flaps, using carbon fiber braids and epoxy resin and the VARTM process. We propose using this system and processing for the research herein. The fatigue performance of fiber reinforced composites is bound by two limiting factors: the fiber strength and the fatigue limit of matrix material. The relative magnitude of their values determines the shape of the fatigue life curve. Even though the macroscopic fatigue behavior is somewhat similar in tape and some braided composites, it is dangerous to assume that the insight from research on tape laminates is directly transferable to braided composites. Very little work has been done in fatigue damage, the relationship between internal damage and macroscopic properties such as stiffness and residual strength, interlaminar stresses and delamination, and inter‐tow stresses and tow debonding in braided composites.
33
Objective The objectives of the research are to
• Develop VARTM processing procedures for vinyl ester and epoxy resin composites • Characterize the static and fatigue performance of 2x2 braided composites. Some of the key
data to be obtained are • Fiber distribution and tow architecture • Engineering moduli • Static and fatigue strengths as a function of braid angle and
material system • S/Su‐N curve • Stiffness degradation as a function of braid angle and fatigue
loading • Description of microscopic damage initiation and growth • Macroscopic fatigue model
• Develop three‐dimensional micromechanics models for braid architectures tested. This includes:
• Prediction of effective engineering properties • Prediction of detailed local stress states • Prediction of damage initiation and growth • Expediting interpretation of experimental data via modeling • Validation of models via comparison with experimental
observations • Identification of critical characteristics that determine performance • Simplified models based on insights from detailed analysis
Achievements: Significant progress was made on experimental, analytical, and combined experimental and analytical tasks. The primary achievements are listed below.
• VARTM Process Development Although VARTM is a simple process, as far as equipment, is concerned, it is quite complicated and has many variables. Designing the proper VARTM process for a particular material combination is a challenge. The series of experiments were performed to optimize the VARTM process for vinyl ester and epoxy resin systems. The following process variables were optimized:
• Processing temperature and viscosity of resin, • Temperature of mold, • Degassing, • Proper placement of resin distribution media, • Placement of resin and vacuum distribution lines, • Use of air release agents, • Use of flow control device, • Double bagging
34
• Static and Fatigue Characterization
The performance carbon/vinyl ester and carbon/epoxy biaxial braided composites were studied under static tension and tension‐tension fatigue loading (R = 0.1, 10 Hz). The major conclusions are
• The axial modulus, Poisson’s ratio and ultimate tensile strength decrease as the braid angle increases from 25º to 45º. • The endurance limit of the braided composites is 40 to 50% of the UTS for the braid angles between 25º and 45º for carbon/epoxy and carbon/vinyl ester material systems. • The S‐N diagram of braided composites for various braid angles can be approximated using the Sigmoidal (Boltzmann) function. • The failure of braided composites always occurred suddenly in the last 10% of the fatigue life without any visible matrix cracking or delamination of plies. • Statistical model is developed to predict ultimate strength as function of fiber volume fraction and braid angle. New approach is suggested for fatigue study to minimize scatter in the fatigue data.
• Stiffness Degradation Model ‐ The Stiffness degradation curves were studied for both carbon/vinyl ester and carbon/epoxy resin systems. Some of the highlights of study are:
• The Stiffness degradation curves for braided composites exhibit a typical three‐stage pattern as that of woven composites. The fatigue life consumed is 5%, 85%, and 10% for stage I, II, and III, respectively.
• The braided composites loose 50% stiffness till failure compared to stiffness in 1st cycle. This indicates that the damage accumulation rate is much higher for braided composites compared to woven composites.
• Endurance limit can be predicted by testing specimens only in the stage I
• A unique analytical model is developed based on stiffness degradation curves to predict the residual fatigue secant modulus. This model takes into account the material heterogeneity and nonlinearity.
• Analysis Development and Validation There were three aspects of the analytical effort: development of the analysis, parametric study using the analysis, and comparison with experiments. The progress is summarized below.
o Three‐dimensional finite element analysis was developed for 2x2 braids. This included
very efficient automated mesh generation that accounts for unusual tow shapes in braided composite
utilities for automated generation of input data required for micromechanics analysis
utilities to expedite interpretation of output from analysis options for
• material nonlinearity
35
• geometric nonlinearity • effective property calculation • progressive failure analysis
• Numerous parametric studies were performed. Some of the highlights of the parametric studies are
• In‐plane properties of carbon/epoxy material system with flattened tow cross‐section can be predicted very well by using a simple laminate model. However, for the glass/epoxy and carbon/epoxy with lenticular cross‐section, the laminate analysis can produce as large as 16% error.
• The most sensitive effective properties were found to be the transverse properties (G13, G23, υ13 and υ23). This suggests that simple laminate theory cannot be used to determine reasonable approximations for the transverse properties of the braid.
• The G13 and G23 can be as much as 72% greater than the laminate value, which means a considerable increase in transverse shear modulus can be achieved using the 2x2 biaxial braid as compared to the equivalent angle‐ply tape laminate. This can be significant for structural applications in which higher G13 and G23 are desirable.
• A complex 3 D stress state which is fully three‐dimensional exists in the tow even for simple uniaxial loading.
• A considerable volume of the tow has larger in‐plane stresses than an equivalent tape laminate.
• Much (but not all) of the variation in stress volume distribution with braid angle is due to simple orientation effects, such as exist even for a tape lamina.
• The severity of the peaks increases linearly with an increase in waviness ratio for all stress components (except for σ12, for which there is little variation).
• The effect of variation in braid parameters on the progressive failure behavior of a 2x2 braided composite laminate was studied. It was seen that once the variation in braid angle or/and tow volume fraction was considered at the laminate scale, the “saw‐teeth” that are typically seen in response curves of uniform braids are smoothed out.
• Progressive failure analysis of different weave architectures was performed. Several property degradation models were considered. It was seen that property degradation models require dependence on material system.
• Comparisons of experimental observations and analytical predictions were performed. Predicted moduli fell within the range of experimental measurements.
36
Damage Tolerance and Durability of Fiber‐Metal Laminates for Aircraft Structures Principle Investigator: J.M. Yang, University of California at Los Angeles Background Fiber‐reinforced metal laminates (FML) are hybrid composites consisting of alternating thin layers of metal sheets and fiber‐reinforced resin prepreg. The most commonly used metal for FML is aluminum, and the fibers can be Kevar or glass. The FML with glass fibers (trade name GLARE), and Kevlar fibers (trade name ARALL) have been evaluated for potential applications in aircraft structures. These laminates possess excellent properties of both metal and fibrous composite materials. This combination results in a new family of hybrid laminates with an ability to impede and arrest crack growth caused by cyclic loading, with excellent impact and damage tolerance characteristics and a low density. Also, the corrosion resistance is excellent because the prepreg layers are able to act as moisture barriers between the various inner aluminum layers, whereas the metal layers protect the fiber/epoxy layers from picking up moisture. The laminate also has an inherent high burn‐through resistance as well as good damping and insulation properties. Furthermore, this material can be produced as sheet material, but also be cured in an autoclave as a complete structure, e.g., a large curved panel with co‐cured doublers and stiffening elements. As a result, GLARE laminates offer the aircraft structural designer a damage‐tolerant, light‐weight, cost‐effective solution for many applications. GLARE laminates seem poised for a much larger future in the primary structure of pressurized transport fuselages. Objectives • Damage Tolerance Modeling and Validation‐ The damage tolerance of GLARE laminate will
be investigated experimentally and analytically. • Durability Modeling and Validation‐ In this task, the durability of GLARE laminates will be
investigated. The constant amplitude fatigue is necessary to determine the damage initiation sites as well as the final fatigue failure mechanisms of a GLARE structure.
• Constant amplitude fatigue (tension‐tension) testing of GLARE laminate with impact induced damage will be conducted. The influence of loading parameters on damage initiation and accumulation during fatigue loading, interaction between different damage modes, and their effect on life and residual properties will be identified.
• The damage evolution and property degradation during fatigue testing will be characterized. Data from constant amplitude fatigue will be analyzed to develop a predictive cumulative damage models.
• Information System for Certification ‐ Development of an information system for damage tolerance design, and certification of GLARE laminate will be conducted. This system will be based on the knowledge database that contains results of current experimental program as well as summary of the experimental data available in the literature. This information system will facilitate retrieval of critical data during design process and in making certification decisions regarding damage tolerance and durability of GLARE structures.
Expected Outcomes There are still little and insufficient information available about mechanical behavior of GLARE in published literature. More research and testing in basic mechanical behavior such as in‐plane
37
shear strength, bearing strength and tensile/compressive behavior at different environments, estimation of fatigue lives and crack growth rates, notched sensitivity, impact behavior, delaminations and damage characterization are necessary to generate adequate data to facilitate greater utilization of GLARE in future aircraft structures. The certification of aircraft structures made of GLARE requires the material qualification, the establishment of new strength analysis methods and their validation by test. Also, the damage tolerance and durability certification methodology of a GLARE laminate in comparison with a certification of aluminum structures needs to be established.
38
Crashworthiness of Composites ‐ Material Dynamic Properties Principle Investigator(s): Suresh Raju, WSU Assistant Professor, Aerospace Engineering and Hamid Lankarani, WSU Professor, Mechanical Engineering, NIAR Fellow Driven by the need for weight reduction and increased fuel efficiency, composites are increasingly being considered in the development of energy‐absorption (EA) devices. The tailorability of the composites, in addition to their attributes of high strength‐to‐weight and stiffness‐to‐weight ratios, corrosion resistance and fatigue resistance have made them quite attractive in the field of crashworthiness, and in particular in the development of new composite fuselage structures. The limited number of dynamic and drop tests performed on fully composite fuselage structures, however, have indicated differences in the crush patterns, stiffness and other structural properties of composite fuselage structures compared with the traditional metallic fuselage structures. Objective The aim of this research is to gain a better understanding of the crashworthiness of the fully composite fuselage structures by means of a series of quasi‐static, dynamic, and impact tests as well as development of analytical models such as hybrid and finite element models of representative sample of composite structures. The following tasks summarize the work to be done at NIAR/WSU on the crashworthiness of composites material dynamic properties project. All the tasks shall be coordinated with the FAA‐TC COTR.
• Survey of strain rates observed in drop test experiments and finite element simulations of fuselage drop tests.
• Dynamic material properties at different strain rates will be generated for NB321/3k70P plain weave carbon fabric/epoxy prepreg and NB321/7781 fiberglass/epoxy prepreg material systems, and Plascore Nomex PN2‐3/16‐3.0 honeycomb cores of 0.75 and 1.125”” thickness. These material systems and their equivalents are used in airframe structures employing sandwich construction.
• Quasi‐static crush testing of some representative sandwich panels and tubes (EA device(s)) mostly used in the construction of composite fuselage structures in order to obtain the pertinent data characterizing the energy absorption of these panels. The failure mechanism(s) dominating the energy absorption as well as crush strength shall be identified.
• The dynamic properties of the same panels under appropriate constant rate(s) will be studied. The stiffness and crush strength under the dynamic load tends to be higher compared to the ones from the quasi‐static tests. The panels will be loaded at constant stroke rates up to 30in/s using a high‐speed MTS servo hydraulic testing machine.
39
• Impact drop testing of the same panels shall be conducted to understand the response of these panels under impact loads which are representative of the crash loads where the strain rate typically varies as dictated by the dynamic equilibrium.
Statement of Work The following tasks summarize the work to be done at NIAR/WSU on the crashworthiness of composites material dynamic properties project. All the tasks shall be coordinated with the FAA‐TC COTR. 1. Survey of strain rates observed in drop test experiments and finite element simulations of fuselage drop tests. 2. Dynamic material properties at different strain rates will be generated for NB321/3k70P plain weave carbon fabric/epoxy prepreg andNB321/7781 fiberglass/epoxy prepreg material systems, and Plascore Nomex PN2‐3/16‐3.0 honeycomb cores of 0.75 and 1.125”” thickness. These material systems and their equivalents are used in airframe structures employing sandwich construction. The variation of tensile, compressive and shear strength and modulii at different strain rates will be studied and are summarized in table (1). A high‐speed servo‐hydraulic testing machine capable of stroke rates up to 30 in/s shall be used for testing strain rates up to 15s‐1. A split‐Hopkinson Pressure Bar apparatus will be used for strain rates above 100s‐1 and up to 2000s‐1.
Material System Properties Loading type
NB321/3K70 P Plain weave carbon fabric/epoxy prepreg
Strength & modulus
In‐plane tension, compression and shear (off‐axis test)
NB321/7781 Plain fiberglass/epoxy prepreg
Strength & modulus
In‐plane tension, compression and shear (off‐axis test)
Plascore Nomex PN2‐3/16‐3.0 honeycomb core
Strength & modulus
Transverse shear, out‐of‐plane compression
Table (1): Summary of material systems and properties to be studied at different strain rates.
3. Quasi‐static crush testing of some representative sandwich panels and tubes (EA device(s)) mostly used in the construction of composite fuselage structures in order to obtain the pertinent data characterizing the energy absorption of these panels. The failure mechanism(s) dominating the energy absorption as well as crush strength shall be identified. 4. The dynamic properties of the same panels under appropriate constant rate(s) will be studied. The stiffness and crush strength under the dynamic load tends to be higher compared to the ones from the quasi‐static tests. The panels will be loaded at constant stroke rates up to 30in/s using a high‐speed MTS servo hydraulic testing machine. 5. Impact drop testing of the same panels shall be conducted to understand the response of these panels under impact loads which are representative of the crash loads where the strain rate
40
typically varies as dictated by the dynamic equilibrium. The schematic of the test set up for the drop tests is shown in figure (1). The energy absorption device is mounted underneath a constrained mass (simulating the occupant). A piezoelectric load cell is placed in between the mass and the EA device to measure the dynamic loads. Additional strain gages will also be mounted on the EA device to monitor the dynamic strains. The entire assembly will be dropped from different heights onto a base to simulate a vertical drop. Compliant material (eg. Foam) will be placed on the plate to vary the pulse shapes. High‐speed video of the drop test will be filmed for post‐test analysis of the failure modes and their sequences. 6. Hybrid analytical models of the panels shall be developed for evaluating the dynamic crush strength. Finite element models shall be developed and validated in order to optimize the honeycomb dynamic crush properties. Expected Outcomes To develop a database of composites material dynamic properties in order to better understand the crashworthiness of composite fuselage structures.
41
Nanophased Skin‐Stringer Assembly for Aircraft Structures Principle Investigator(s): Hassan Mahfuz, Professor of Ocean Engineering, Florida Atlantic University The subject of nanocomposites is especially interesting in the sense that in nanocomposites at least one of its phases has one or more dimensions ‐ length, width, or thickness ‐ in the nanometer size range, usually defined as 1 to 100 nm. This is the range where phenomena associated with atomic and molecular interactions strongly influence the macroscopic properties of the materials. But this is also the length scale where our knowledge of how to synthesize and process materials is weakest. Nevertheless, it is very much known that the catalytic, mechanical, electronic, optical, and other properties of a material can significantly and favorably be altered when that material is fashioned from nanoscale building blocks. It is our intent in this proposal to incorporate the benefits of nanomaterials into structural composites. One of the important areas for future investments in research identified by the NATIONAL NANOTECHNOLOGY INITIATIVE (NNI) is the “beyond nano,” which notes that the advances at the nanoscale will be meaningless if they cannot be interfaced well with the technology at larger material components, systems and architectures to produce usable devices. One of the primary objectives of the NNI is therefore, to integrate nano‐objects and nanoscale phenomena into larger hierarchical systems. Development of large structural level components from nano‐infused polymers as being pursued in this proposal precisely addresses such a problem. It has been established in recent years that polymer based composites reinforced with a small percentage of strong fillers can significantly improve the mechanical, thermal and barrier properties of pure polymer matrix. Moreover, these improvements are achieved through conventional processing techniques without any detrimental effects on processability, appearance, density and aging performance of the matrix. The benefits of nanoparticle infusion comes from the fact that the large amount of interphase zones in nanocomposites may serve as catalysts for prolific crack growth creating a much greater amount of new surfaces. The creation of new surfaces as we know can serve as efficient mechanisms to dissipate kinetic energy, for example, in the event of an impact. These interphase zones can also be visualized as defects, the density of which will be very high in nanocomposites such that the spacing between them will approach interatomic distances and a large fraction of atoms will sit very adjacent to a defect. Any brittle crack developed in the material will therefore get deflected and branched out into these defects attributing a crack‐blunting feature to the composites. If these unique features can be imparted to structural composites through the modification of matrix, it may bring about significant improvement in the performance and structural integrity of the resulting composites. While nanoparticles have attractive attributes, their usage in structural composites which are relatively large in dimension is almost non‐existent. The main objective of this proposal is to disperse nanoparticles into the matrix of a composite material, and fabricate structural components out of those nanophased matrices. The goal is to impart the strength and superior quality of nanomaterials on to structural composites. Objective The research effort is aimed at the establishment of a science basis for a new class of nanophased materials for aircraft applications. Specifically the objectives are to:
• Develop new generation of polymer‐matrix nanocomposites systems, which will be based on the nano‐scaled dispersion of an inorganic phase in a thermally stable polymer.
42
• Utilize the modified polymers with compatible fibers through an affordable and cost effective process to construct structural components such as skin‐stringer assembly. • Determine the structural integrity of the nanophased skin‐stringer assembly under suitable and appropriate loading conditions.
Expected Outcome
• A methodology for infusion of nanoparticles into matrices through optimal sonic cavitation • An understanding of fundamental science behind particle and polymer interaction • A procedure to construct composite Skin‐Stringer Assembly using VARTM • Nanophased skin‐stringer assembly with superior structural integrity.
Collaboration with Boeing As stated earlier the basic premise of the proposed work began with a funding from Boeing couple of years ago to develop composite skin‐stringer assembly with continuous reinforcement of fibers between the skin and the stringer. Although the funding for the previous work has ended, Boeing is still very supportive of the proposed work to take it to the next phase by modifying the resin with nanoparticle infusion.. We will persistently try to secure new funding from Boeing and use it as matching for the FAA grant.
43
Course Development: Maintenance of Composite Aircraft Structures Principle Investigator(s): Larry Ilcewicz, Chief Scientific and Technical Advisor, Advanced Composite Materials, FAA
Presenter: Larry Ilcewicz, Chief Scientific and Technical Advisor, Advanced Composite Materials, FAA Objective The goal of this proposal is to develop, in conjunction with AMTAS academic and industry partners, a syllabus and course material for a short course addressing the maintenance and repair of composite aircraft structures. Expected Outcomes
1. A three‐ to five‐day (equivalent) course, including both distance learning and the regional workshop.
2. Terminal Course Objectives (TCOs). 3. Curriculum development and delivery of first course, not later than September, 2005.
Progress A curriculum development timeline is shown in Figure 2. A curriculum workshop to help define Terminal Course Objectives (TCOs) was held near Seattle November30–December 2, 2004. Approximately 60 international subject matter experts attended. A final report was prepared and posted on the FAA Center of Excellence for Advanced Materials in Transport Aircraft Structures website for comment by the participants and other interested parties.
• TCOs form the framework for curriculum development in composites maintenance and repair, in support of high‐composite materials usage aircraft such as the Boeing 787. TCOs have been refined based on feedback from workshop participants; some have been assigned to a second ‘prerequisite’ course in order to prepare students for the compressed 5‐day course.
• An expert consultant, contracted by Edmonds Community College, is refining the course objectives and storyboard, and assisting in safety message development.
• Additional subject matter experts (SMEs) have been contacted for various portions of the course development, including industry, government, academic and independent consultants.
Next Steps:
• Work scope for the FAA grant is expanding, and discussions are on‐going with the FAA concerning increased funding
• SMEs are being contacted to provide ‘testimonials’ and safety messages which will be integrated into course modules
• Industry contacts are being made for composite repair video modules to incorporate into the course via ‘video streaming’
• Efforts continue to recruit SMEs to provide input into the TCOs and develop the regional laboratory component of the course.
• Laboratory equipment lists are being developed and will be compared to the current inventory list in order to order equipment and materials
44
• A teleconference was conducted for those workshop attendees interested in reviewing and providing input into the course development. Edmonds Community College is developing synchronous communications which will be a combination audio and visual discussion from remote computer stations.
Figure 1: Composites Curriculum Overview
Prerequisite Development –
Web BasedWeb Based
Aug to Sept ‘05
Lab - Regional
Content – Web Based
2006
Lab - Regional
Content - Classroom
Repair Course Development
Apr to Jul ‘05Lecture/PowerPointStreaming VideoTestimonialsSafety Messages‘BlackBoard’ or equiv.
TCO Development
Organize Course Modules
WorkshopFeedback
Nov/Dec ’04 Workshop Jan to Apr ‘05
45
Figure 2: Curriculum Development Timeline
46
Development of Reliability‐Based Damage Tolerant Structural Design Methodology
Principle Investigator(s): Kuen Y. Lin, Professor, Professor, Department of Aeronautics and Astronautics, University of Washington
Background
Traditional design procedures for aircraft structures are based on combinations of factors of safety for the loads and knockdown factors for the strength. Both the factors of safety and knockdown factors have been obtained from the past five decades of design for metal aircraft. There are at least two shortcomings to these traditional design procedures. First, because the procedures were developed for conventional configurations, metallic materials, and familiar structural concepts, they may be difficult to apply to aircraft that have unconventional configurations, use new material systems, and contain novel structural concepts. Consider, for example, the case of composite materials. Adaptations of traditional design procedures to account for larger scatter in composite properties and the sensitivity of composite structures to environmental effects and to impact damage have led to a very conservative approach for designing composite structures. This approach, in essence, assumes that a ʺworst case scenarioʺ occurs simultaneously for each design condition—temperature, moisture, damage, loading, etc. This results in substantial and unnecessary weight penalties. Another shortcoming of traditional design procedures is that quantitative measures of reliability are not available. As a result, it is not possible to determine (with any precision) the relative importance of various design options on the reliability of the aircraft. In addition, with no measure of reliability it is unlikely that there is a consistent level of reliability and efficiency throughout the aircraft. That situation can lead to excessive weight with no corresponding improvement in overall reliability. To overcome these problems, a new design approach to quantify the reliability of aerospace structures has been proposed by Lin, et al [1, 2]. In this approach, the “Level of Safety (LOS)” of an existing structural component is determined based on a probabilistic assessment of in‐service accumulated damage and the ability of non‐destructive inspection methods to detect such damage. Specifically, the discrete LOS for a single inspection event is defined as the compliment of the probability that a single flaw size larger than the critical flaw size for residual strength of the structure exists, and that the flaw will not be detected. The cumulative LOS for the entire structure is the product of the discrete LOS values for each damage type present at each location in the structure. This approach can be utilized to develop a design process that evaluates the equivalent LOS of an existing structure, and use this value in the design of a new structure that matches or exceeds the existing LOS value. The LOS method enables the characterization of uncertainty associated with damage accumulation, inspection reliability and residual strength behavior of the structure. Using the general concepts of this design methodology, the reliability of a structure can be assessed on a quantitative basis, allowing aircraft manufacturers, operators and flight certification authorities to evaluate the risk associated with structural failures in an aircraft fleet. Objective The overall objective of the proposed research program is to develop a probabilistic method to estimate structural component reliabilities suitable for design and inspection, and regulatory
47
compliance. The proposed research spans over a three‐year period, consisting of two phases of study. The first year will focus on methodology development and validation while the second and third year will concentrate on the application of the developed technology, such as inspection scheduling and maintenance service guidelines. The cooperative strategy is to partner this research with the FAA and Boeing. Phase I Research Tasks • Develop a probabilistic method to determine inspection intervals for composite aircraft
structures. • Develop computing tools and algorithms for the probabilistic analysis. • Establish in‐service damage database from FAA SDR and other sources. • Demonstrate the developed method on an existing structural component. Work Accomplished
• A probabilistic method for determining inspection intervals of composite structures has been formulated. Two basic probabilistic models have been developed: The Integration and Full Monte‐Carlo models. The first model uses the Importance Sampling and Latin Hypercube techniques to compute the Probability of Failure (POF) integral using the Monte‐Carlo integration. The second uses a traditional Monte‐Carlo simulation of major uncertain parameters contributing to the structural reliability of damage‐tolerant composite structures. Both models are based on random simulation of individual residual strength histories shown in Figure 1. These models take into account the following uncertain parameters: types of damage, number of damages per life, damage size, damage initiation time, time of damage detection, external loads, structural temperature, initial failure load, and residual strength degradation due to the damage and possible strength degradation after repair.
Figure 1
Residual Strength
Time
1 2
3 4
5
6 7
8
ith interval of constant strength: Time interval ti[T1,T2…,D], Damage type TDi, Residual strength Si(D,TDi),
48
• The development, testing and documentation of both models have been completed. The Integration model provides reasonably good speed and accuracy, while the Full Monte‐Carlo model allows for better temperature simulation and failure data output, thus providing a detailed understanding of the underlying probabilities.
• Basic computer software has been developed. It uses MS Excel for database management, Excel Visual Basic Scripts for running common ActiveX component written in Visual C++, which invokes in its turn the different integration or simulation modules written in Visual Fortran. The software debugging has been completed.
• The validity of developed method and software has been demonstrated on four existing structural components. In addition, a generic example problem has been analyzed. This problem includes most features of the four problems mentioned above, but the initial data are simplified to provide better understanding of the main ʺrisk driversʺ affecting the Inspection Interval. The output of the study shows how the Inspection Interval varies with parameters used in the model. The parametric study shows that:
1. The damage occurrence data are extremely important. As expected, the POF is
proportional to the damage occurrence rate and the Inspection Interval is reduced rapidly to maintain the same high level of reliability. Statistically meaningful damage occurrence data are critical in this analysis.
2. The POF is not highly sensitive to the detection probability. Figure 2 shows the
variation of the Inspection Interval with the average detected damage size as measured by the Weibull scale parameter �. As can be seen, the inspection interval decreases as detected damage size increases. However, the rate of change is not rapid. Therefore, it is possible to use an inexpensive NDI method for a smaller inspection interval instead of an expensive accurate inspection procedure.
49
Effect of Average Detected Damage
50
70
90
110
130
150
170
0.0 2.0 4.0 6.0 8.0 10.0
Beta parameter, in
Insp
ectio
n in
terv
al, F
light
s
Figure 2
3. The inspection interval required versus strength restoration after repair is shown in Figure 3. Under certain conditions, the inspection interval is not sensitive to percentage of the strength restoration. For example, 80% of the residual strength restoration is sufficient to maintain a reasonably high reliability level with the same inspection interval.
Inspection Interval required to maintain the same POF vs. % of Strength Restoration after Repair
19
110135 138 138
0
50
100
150
60% 70% 80% 90% 100% 110%
Strength Recovery after Repair
Insp
ectio
n In
terv
al.
Flig
hts
Figure 3
50
Methods for the Evaluation of the Fitness of Fiber Reinforced Composite Surfaces for Subsequent Adhesive Bonding Principle Investigator(s): William T.K. Stevenson ,Wichita State University Chemistry Professor, Charles Yang, Associate Professor, Mechanical Engineering, NIAR Fellow The use of carbon based composite materials in primary load bearing applications in the commercial aerospace industry has increased dramatically over the past decade. With user safety in mind, the FAA has determined to coordinate and lead an effort to improve our understanding of current composite bonding methodology in the commercial aerospace and scientific community, with special emphasis on surface preparation and characterization prior to the bonding event. To this end, FAA has issued a RFP to instigate a project that will (1) evaluate and archive current industrial methodology, (2) survey the scientific and industrial literature with a view to selecting the technique(s) best able to detect surface structures that promote good and poor bonding, and (3) begin to develop protocols for the use of said technique(s) of surface analysis in the industrial setting. Objective To develop a test (or tests) that will address a surface property that is associated with the formation of a “good” or “bad” bond, said test to be non‐destructive and applicable to the pre‐bonded composite surface, so as to be able to identify that surface as conducive or non‐conducive to the formation of a good or deficient bond line, and to have potential for use on the shop floor In addition, the FAA has identified three milestones required for successful completion of this project.
• Milestone #1 Identify Contaminates • Milestone #2 Identify Potential Chemical Analysis Technologies • Milestone #3 Establish Appropriate Chemical Analysis Level
Our initial literature review has allowed us to address the first 2 milestones. We will provide proof of concept then establish appropriate levels of analysis (Milestone #3) during the integrated program of work that is detailed in this proposal. Technical Tasks Surface Contaminants To simplify discussion, we have distributed surface contaminants into two categories – water and “everything else”, the latter consisting of (among other things)….. • Silicone mould release agents • Peel ply residues • Residues from gloves • Other oils and greases • Food products, hair products • Material deposited from the atmosphere during long term storage • Solvent residue
51
Specific Probes for Surface Contamination For water • Attenuated Total Reflectance mid Infrared spectroscopy • Attenuated Total Reflectance near Infrared spectroscopy • Diffuse reflectance near Infrared spectroscopy For everything else • Laser desorption mass spectroscopy • ESCA, XPS, and SIMS (for reference purposes) For water and everything else • In situ liquid droplet spreading – contact angle measurements Plan of Work Carbon fiber reinforced composite plaques will be made under contract in the composites lab at NIAR using standard technology and shaped to conform to either of the static or dynamic wedge test geometry. Step 2: Grit blast composite test piece surface then apply surface treatment Step 3: Apply test for surface and produce measurement Step 4: Glue test pieces together into wedge test or lap shear configuration Step 5: Run static or dynamic wedge test and obtain measure of crack growth or bond line strength
Step 6: Analysis of the experiment. We will correlate the results of steps 3 and 5 and establish relationships between the measurement used to characterize the surface and bond integrity. Specific Surface Tests
• Surface Analysis for Water by Fourier Transform Infrared (FTIR) Spectroscopy. Fourier Transform Infrared spectroscopy will be used to determine the levels of water at and near the surface of the composite prior to bonding • Surface analysis by ESCA/XPS and SIMS. XPS/ESCA and SIMS will be used as reference techniques against which to evaluate the efficacy of techniques of surface characterization to be developed in this project. • Surface Analysis by Laser Desorption/Ionization – Mass Spectrometry (LD‐MS).
52
• Surfaces will be characterized by LD‐MS spectroscopy prior to bonding as a means of detecting trace levels of surface contaminant • Surface analysis by Measurement of Contact Angle /Droplet spreading on the Surface • We will measure the contact angle that a droplet of clean solvent makes with the prebonded composite surface and from the contact angle measure the surface free energy of the surface
Expected Outcomes We envision this work to result in organizing a planning meeting at NIAR, to which we will invite interested industry players., for the purpose disseminating our goals and invite input into how to best tailor the work to meet current industry needs. It may be advantageous to invite participants from FIU to discuss how to plan and coordinate our efforts. In response to our objectives and the FAA derivatives, we will develop non‐destructive chemical tests to determine the presence of, and levels of, water and other contaminants at the surface of cured composites prior to bonding – said tests to be potentially able to be applied using apparatus that can be positioned and repositioned close to the part, then correlate the results of those tests with bondline integrity so developed.
53
Identification and Validation of Analytical Chemistry Methods for Detecting Composite Surface Contamination and Moisture Principle Investigator(s): Xiangyang Zhou, Florida International University Background Adhesive bonding has been used in the manufacture and repair of primary aircraft structures for over 50 years and is still in use on current aircraft projects as a direct competitor to riveting. Adherend surface preparation is a critical issue to structural integrity of bonded structures. Inadequate surface roughening, possible chemical contamination on peel ply, release fabric and release film, and surface water moisture result in poor adhesion, i.e. a weak bond between the adhesive and adherend, and reduced long‐term durability. The problems with chemical contaminations from peel ply, release fabric and release film that prevent adhesion of the adhesive to the substrate are now fairly well known. What is far less understood is the adverse influence of pre‐bond water moisture that is unable to avoid during manufacture, repair, and service. Water inclusion in pre‐bond adherends could affect short‐term or long‐term strengths of adhesive bonding depending on how fast are the diffusion and accumulation processes. As being presented in the recent FAA meeting on bonding structures, water moisture is claimed as one of the most adverse factors in adhesive bonding processes. Current adhesive bonding quality assurance practice relies on tightened surface preparation process control and mechanical testing on bonded specimens and non‐destructive inspection (NDI) after bonding. Thus, in the absence of a definitive surface quality control method, laborious and sometimes inadequate measures are used to ensure the quality of adhesive bonding, thereby creating an undue expense on an otherwise economic manufacturing process. Objectives The objectives of the proposed research are: 1) identify surface quality assurance methods that are currently being used by aircraft manufacturers and repair service providers and determine whether the current quality assurance tests including the wedge test are sufficient to ensure the contaminated peel plies are detected and not used, and 2) to identify and validate definitive analytical chemistry methods to provide sufficient in‐field quality assurance. Approaches In the present phase of this research, FIU will benchmark or improve understanding of surface preparation processes, surface assurance and certification procedures using mainly information collection and analysis approach. FIU will also use atomic force microscopy (AFM), scanning electron microscopy (SEM), and energy dispersive X‐ray spectroscopy (EDS), and electrochemical measurements to study the surface morphology, surface chemistry, and activity of surface for peel ply samples. These analytical studies in coupling with the information analysis should allow establishment of criteria for in‐field, online analytical chemistry methods for the surface preparation assurance. In addition, FIU will identify and validate technologies that are promising for the in‐field surface preparation assurance.
54
Preliminary Result Outcomes FIU has reviewed a large number of articles and data with emphasis on quality control procedures for fabricating environmentally durable adhesive bonds to benchmark the understanding of current adhesive bonding technology. Analyses on conflicting results of surface pretreatments lead to conclusions as follows:
1. Post‐bond mechanical strength tests including the Boeing wedge test are not sufficient for certifying environmentally durable adhesive bonds.
2. Variations in bond strength and durability with the same pretreatment‐bonding method implied that an effective quality control procedure is needed to control and reduce the variations. Adhesive bonded joints fabricated under quality control procedure will have predictable in‐service performance.
3. It was found that bond quality was affected by the nature and timing of surface hydrocarbon contamination during pretreatment peel ply or tear ply procedures, while pre‐bond moisture on the adherends was the most detrimental to bond integrity.
4. The analysis indicates that a contaminate‐free adherend surface is a pre‐requisition but not a sufficient condition for forming a strong and durable adhesive bond. A chemically activated adherend surface can enable covalent bonds between the adherend and adhesive. The covalent bonds can effectively inhibit the bond displacement due to contaminants and ingress water during service. The surface preparation certification criteria should evaluate both cleanliness and activity of the surface.
5. Certification of pre‐bond surface preparation quality requires implementation of effective surface chemistry inspection technologies for each and every steps of the surface preparation procedure to ensure the strength and durability of the bonded aviation structures.
6. A prototype carbon nanotube (CNT) based humidity sensor has been developed (Figure 1). Experiment studies indicate that this sensor is sensitive to moisture (Figure 2).
Figure 1. The schematic depicts the possible humidity sensing mechanism by PVA
functionalised Y‐junction single wall carbon nanotubes.
55
Figure 2. Effect of relative humidity change on conductivity of PVA functionalized and pristine
Y‐junction single nanotubes. Expected Outcomes The benefits of this research to the aviation industry are as follows:
1. Better understanding of the pre‐bond surface preparation methods 2. Better understanding of bond strength and durability versus surface preparation 3. Novel in‐field, online certification and assurance technology for surface preparation and
adhesive bonding processes 4. Reduced costs for surface preparation and adhesive bonding processes
56
Adhesive Characterization and Element Testing of Fatigued and Damaged Bonded Joints Principle Investigator(s): Waruna Seneviratne, Manager, NIAR Structures Lab Background Development of advanced material and process technologies has significantly increased the use of adhesively bonded joints in aircraft primary and secondary structures due to favorable characteristics of these joints in comparison with riveted, spot‐welded, and mechanically fastened structures. Applications on primary structures require rigorous characterization of the material. This includes quality assurance and durability investigation. These short‐term and long‐term issues must be investigated so that the certification of these joints can be instituted to ensure the structural integrity of these joints throughout the service‐life of the aircraft. This practice can avoid a catastrophic failure of bonded structures similar to Aloha Airline’s Boeing 737 incident that led to the emphasis on damage tolerance investigation of mechanically fastened metallic structures. Objective Purpose of this research was to investigate the structural integrity of bonded joints during operation through damage tolerance and effects of defects investigation and propose a qualification methodology for certification and quality assurance of these joints. This study investigated the concerns related to manufacturing defects of bonded joints and their impact to the structural integrity. These integrity issues are categorized under concerns on the adhesive, adherend and/or bondline and are highlighted below: 1. Adhesive – During the manufacturing process the adhesive properties can be affected by the process, contamination and cure profile. Porosity can also affect the material properties and degrade the joints as they are exposed to humidity and fatigue. In addition to the aforementioned two conditions, during service, the integrity of joints can be affected by exposure to heat and ultraviolet radiation, and stress relaxation. 2. Bondline – This issue concerns the integrity of the bondline or the adhesion between the adherend and adhesive. This has commonly been referred to in the literature as adhesive failure, but for the purpose of this white paper will be referred to as adhesion. Bondline integrity can be attributed to imperfections of adherend surface, pre‐bond moisture, poor surface preparation of the adhesive surface, and aging of the adhesive layer in composite joints. Additionally, in metal bonded joints, the oxide layer and the primer can contribute to the integrity of the structure. Poor surface preparation of these joints is a major concern and one of the leading causes of joint failure. The process and handling can also create significantly large voids (debonds) that will reduce the load carrying capabilities of the joint and may be the location for cracks to start in the adhesive. The changes in the bondline thickness within the joint can also lead to unfavorable stress concentration that can affect the joint durability in service. The effects of curved and joggle joints may also pose a threat to the durability of the joint as well. The loss of surface energy that may have been the result of poor surface preparation or oxidation, in time directly affects the structural integrity of the joint. Exposure to ultraviolet radiation, humidity, and fatigue conditions over a long‐period of time have also not received a large amount of investigation.
57
These conditions along with impact, overload, and heat may result in debonds that alter the load path. 3. Adherend – The defects of composites adherend such as delamination, under‐cured resin, resin contamination, as well as the wrinkles/depression resulting during fabrication, have received some investigation in the past. However, the effects of such defects on the integrity of the bondline have not received much attention. In addition, delamination of composites adherend, impact damages, exposure to heat and humidity, and overload may also deteriorate the integrity of bonded joints. The proposed research is divided into three tasks listed below which outline the technical approach and objectives of the program. The program will focus on adhesive characterization issues, bondline variation affects and element damage tolerance of bonded joints. Technical Tasks Task 1 – Coupon and Sub‐Element Characterization The purpose of this proposed research is to investigate the structural integrity of bonded joints during operation through damage tolerance and effects of defects investigation and propose a qualification methodology for certification and quality assurance of these joints. This phase of the program will be conducted jointly in parallel with the adhesive survey and bonding workshop tasks. Phase II of the experimental program will be to look at larger structures as indicated in the initial proposal. Loctite EA 9392 two‐part paste adhesive system will be selected for this phase of the investigation. Test Matrices The structural integrity of bonded joints can be weakened by a number of factors, which include manufacturing defects that occur during airframe production and operational defects and/or damages that occur during operation. These integrity issues are categorized under concerns on the adhesive, adherend and/or bondline. The apparent shear strength of adhesively bonded joints with three different defect configurations will be investigated: 1. Variable bondline thickness 2. Disbond 3. Impact In addition to the effects of defects investigation, a newly developed V‐notched rail shear test method along with several standard test methods will be evaluated for the qualification and quality control of adhesives. Variable Bondline Thickness Effects on Bonded Joints (Sub‐Task 1.1) The goal of this task will be to generate bondline variability data using element tests which represent typical aircraft construction and loading. A small working group will be formed to identify representative bonded joints to be studied for thickness variation. This will be used to establish guidelines of the bondline variability limits within a joint and to provide guidance as to the effect this variability has with the load carrying capability of the joint. This sub‐task will focus only on either flat or joggle joints tested in a torsion only loading configuration. A total of twenty four (24) specimens will be tested to investigate the effects of variable bondline thickness.
58
Effects of Disbonds in Joints (Sub‐Task 1.2) For safety assurance of adhesively bonded joints, it is vital for adhesive joints to demonstrate “no‐growth” behavior with the presence of discrete damage. There is also a need to develop and demonstrate the ability of adhesive joints to contain larger amounts of damage/disbond that might be incurred during flight or during production. Successful demonstration of this will aid in the deployment of adhesive joints to other portions of aircraft primary structure. The effects of different disbond geometries, lightening strike damages, and fastener installation on the residual strength of the composite bonded joint were investigated using picture frame test setup. Effects of Impact Damages on Bonded Joints (Sub‐Task 1.3) In this sub‐task, the impact damages due to different impact diameters will be investigated. Impact testing will be conducted using a gravity assisted drop tower with a high speed data acquisition system. Impacted specimens will be subjected to TTU C‐scanning to quantify the planar damaged area using image analysis software. In addition, the residual indentation will be measured. Adhesive Qualification Methodology (Sub‐Task 1.4) The purpose of this task is to begin the development of a recommended standardized characterization and procurement methodology for the development of an adhesive to be used in structural bonding applications. Task 2 ‐ Adhesive Survey and Summary Document The primary objective of this task is to conduct a survey on adhesive bonding which will be used to as a basis for a FAA sponsored workshop in June 2004. Task 3 ‐ Damage Tolerance on Full‐Scale Bonded Assembly The main goal of this proposed task will be to test and explore the limits of damage tolerance of the bonded airframe. The objective of this task will be to provide damage tolerance data to demonstrate scaleup issues of full‐scale bonded assemblies. Once several joints are identified, the objective of the program will be to investigate the damage tolerance aspect of the assemblies under static and fatigue loading.
59
Expected Outcomes
1. A working group will be established at the beginning of the program which includes representatives from the FAA, WSU, participating aircraft industries and adhesive suppliers. Additional members will be added to the working group as needed and directed by the FAA program monitor. This working group will develop a detailed plan and schedule for the research program which focuses on the element bonded joint testing and characterization under damage and fatigue. (2) Develop characterization and procurement specifications for adhesives based on the work currently funded by the FAA for Liquid Resin Molding (LRM) materials.
2. Specific variable thickness bonded joint tests as determined by the working group.
Results of experimental tests and supporting analysis to be detailed in FAA final report.
3. Specific bonded joint element tests with damage as determined by the working group. Results of experimental tests and supporting analysis to be detailed in FAA final report.
4. Specific bonded joint element tests under fatigue loading as determined by the working
group. Results of experimental tests and supporting analysis to be detailed in FAA final report.
5. Survey the industry and collect information from past military and commercial
applications of bonding to fabrication and repair of aircraft structures. Organize a FAA workshop on bonding to be held in Seattle, WA in June 2004 and produce a final FAA report which summaries the results from the survey.
6. Help conduct the Bonded Structures Workshop. This will include providing organization
and leadership in the breakout sessions based on results collected to date in the survey. Additional data collected during the bonded structures workshop will be added with results from the bonded structures survey in documenting a contractor report. Recommendations for future efforts in benchmarking the use of bonding in aircraft products will also be derived and documented.
60
Damage Tolerance and Durability of Adhesively Bonded Composite Structures Principle Investigator(s): Hyonny Kim, Assistant Professor, School of Aeronautics & Astronautics, Purdue University Significant amounts of composite materials will be used in new commercial jetliners such as Boeing 7E7 and Airbus 380, as well as in small and mid‐sized aircraft. Many composite‐to‐composite and composite‐to‐metal components are bonded by adhesives, in both secondary and primary load bearing structures. It is well known that failure prediction remains a difficult problem in composites and in adhesively bonded joints. For adhesively bonded, built‐up structures, failure is even more difficult to predict due to the complexities of geometric details and loading. Being able to definitively predict failure is a necessary technology that needs further research for obvious reasons related to safety, performance, and cost. Flaws and damage such as matrix cracks and delamination within the composite, and debonding within joints, are difficult to detect and yet can potentially lead to catastrophic failure. Such damage forms, when detected, also present significant challenges within the context of repair. Efficient repairs to cracked metal and composite structures are achieved by adhesively bonding a composite patch over the damaged zone. An additional component of complexity arises when considering the durability of bonded composite structures: degradation of polymers (composite matrix and adhesive) can occur due to long‐term environmental exposure, thereby negatively impacting the fracture and fatigue characteristics of these materials. This proposal is composed of three activities, each focusing on aspects related to the effect of variable adhesive bondline thickness on the failure of bonded composite structures. Objectives This project is focused on understanding the effects of bondline thickness on the damage growth mechanisms in adhesively bonded composites structures. The overarching objective is the investigation of the physical phenomena and processes that lead to joint failure, and the development of models describing these phenomena so as to make possible the prediction of the tolerance of bonded structures to damage, whilst accounting for bondline thickness variations and environmental effects. Research activity will be concentrated on three areas, all revolving around the dependency of joint failure on bondline thickness:
1. Establishing a methodology for consistently relating the intrinsic material properties of adhesives in bulk and confined joint form, so as to define bondline‐independent constitutive behavior,
2. Accounting for bondline thickness effects in the development of crack growth criteria which will be based on fracture mechanics and on cohesive zone modeling approaches, and
3. Investigating how varying bondline thickness interacts with moisture effects in the fracture behavior of adhesively bonded joints.
61
Technical Approach The objectives of this project will be accomplished by three closely tied activities:
• Adhesive Constitutive Behavior in Bonded Joints • Bondline Thickness Dependent Fracture Criteria • Influence of Moisture and Bondline Thickness on Joint Fracture
Adhesive Constitutive Behavior in Bonded Joints This activity will investigate the effect of adhesive bondline thickness on measured apparent adhesive properties with the intent of determining an intrinsic set of properties that can describe the adhesive material behavior independent of bondline thickness. The objective of this component of the project is to develop a methodology for defining the constitutive law representing the intrinsic material properties of an adhesive, and using detailed FE models to account for geometric effects so as to reconstruct the bondline thickness dependent apparent material behavior. Bondline Thickness Dependent Fracture Criteria Fracture Mechanics Approach The ultimate objective of this activity is to account for the effects of bondline thickness on the mixed mode fracture of bonded lap joints and to establish relationships for predicting the total critical strain energy release rate as a function of mode mix ratio and bondline thickness. Influence of Moisture and Bondline Thickness on Joint Fracture This activity will focus on how varying bondline thickness affects the fracture behavior of adhesively bonded joints under exposure to a wet environment. Increased moisture content in the adhesive layer (polymers in general) causes significant changes to the constitutive properties and failure behavior. An important aspect of this problem that will also be accounted for is localized material heterogeneity which can exist in the form of a gradient in moisture content, i.e., a non‐uniformly saturated adhesive layer along the bondline direction. Expected Outcomes The outcome of this research will provide the end user community with useful tools that can be employed to interpret bondline thickness dependent test results, and to relate these results to the prediction of fracture in bonded joints. This has implications related to the amount of testing that is needed in order to characterize an adhesive system, as well as to certify adhesively bonded composite structures. Ultimately, bondline thickness can largely be eliminated as a test parameter over which data must be gathered, by the use of mechanics‐based modeling, and thereby significantly reducing the amount of testing that is generally needed for measuring thickness and moisture content dependent adhesive properties. The ultimate motivation for this activity is in response to the difficulty expressed by the end user community in the characterization of adhesive properties, and in particular the issue of thickness dependency and environmental effects. The expected outcome will be the development of the
62
capability to predict fracture in general joints using fracture data derived largely from a combination of bulk adhesive tests and a select number of joint tests, as opposed to an exhaustive adhesive characterization test matrix spanning a range of adhesive bondline thickness.
At the conclusion of the first year of this research activity, it is expected that the following items will be provided to the FAA.
• Models and modeling guidance to account for the adhesive constitutive behavior and its relationship to varying adhesive bondline thicknesses
• Bondline thickness dependent mixed mode fracture data for use in correlations with model development
• Moisture dependent Mode I fracture test data from DCB tests • Cohesive zone traction‐separation laws accounting for bondline thickness and moisture
content in the adhesive • Fracture models accounting for material heterogeneity, i.e., local spatial variations in adhesive
properties
Production Control Effect on Composite Material Quality and Stability Principle Investigator(s): Dr. John Tomblin, Executive Director, NIAR; Yeow Ng, Associate Director, National Center for Advanced Materials Performance Advanced composites have emerged as the structural materials of choice for many aerospace applications because of their superior specific strength and stiffness properties. First developed for military applications, composites now play a significant role in a wide range of current generation military aerospace systems. There has been a significant increase in the use of composite materials by the large commercial transport aviation industry during the past 25 years, and many advances have been made in general aviation and rotorcraft vehicles where composites are utilized for primary structural applications. Unlike metallic materials used in structural part manufacturing processes, the material properties of composite structures are manufactured into the structure as part of the fabrication process. Therefore, it is essential that material and process specifications used to produce composite structures contain sufficient information to ensure that critical parameters in the fabrication process are identified to control production and adherence to the engineered part requirements. Due to the wide variety of composite structures now emerging for certification (particularly for general aviation aircraft), control of the materials is rapidly becoming a vital issue with respect to the overall assurance of safety. This project will interrogate industry sources to determine which issues are understood and which may need further investigation. These issues will be explored by interfacing with appropriate members of aerospace supplier organizations and material user organizations to understand the successes and failures (lessons learned) under the current operating strategy of both aerospace and commercial products. Objective
1. Identify what fiber, resin, and interface issues are possible which could lead to loss of material control in the product produced for aerospace applications.
2. Identify control strategies for process lines, levels control and their importance in the final product form reliability (correlation between constituent materials, fiber resin‐mixing and final material form [Prepreg, VARTM]).
3. Review acceptance tests, and specific selected acceptance values (minimum, average and maximum) affect on the detection of safety concerns (vs. economic).
4. Determine how associated safety risks could be mitigated. Expected Outcomes This program will develop essential information on the nature of the controls required at the producer level to assure the continuation of stable and reliable composite raw material for aerospace usage. The intent of this investigation is to determine the level and types of material and process control that would confirm that the property values established during initial evaluation and characterization are not changed over time. In recent experience, a required level of control has been prescribed by aerospace suppliers, which is not necessarily maintained for less critical needs (e.g., sports equipment, infrastructure).
64
This program will review the delineation in control between aerospace and commercial products. The investigation will identify the differences and similarities in the control process and determine their effect on the reliability of the product produced. It is recognized that variation control related to high volume production is sometimes more restrictive than the controls provided on aerospace products due to economic factors. This may mean that the controls on “commercial” products meet the needs of the aerospace community but are not specifically aimed at end product assurance testing. The suitability of these alternate control strategies and their effect on the ultimate reliability for aerospace applications will be explored.
65
Thursday, May 26, 2005 Full‐Scale Damage Tolerance of Adhesively Bonded Composite Joints Principle Investigator(s): Suresh Raju, Wichita State University Background Sandwich constructions are widely used in airframe structural applications due to the distinct advantages they offer over other metallic and/or composite (monolithic laminate) structural configurations in terms of stiffness, stability, specific strength, corrosion resistance, ease of manufacture and repair, and above all the weight savings. However, the sandwich structures are very susceptible to localized transverse loads, due to their inherent construction. These loads are transient in nature and could be inflicted on the airframe structure during various stages of the aircrafts life. The response of sandwich structures to the transient loads, the resulting damage states, their detectability and the effects of the damage states on the residual properties have been widely investigated using experimental and analytical methods. However, most damage resistance and tolerance investigations have been limited to laboratory coupons and the studies on full‐scale airframe components are rare. In this report, a brief summary of the lessons learned from the coupon level testing of sandwich panels is summarized, based on which damage resistance and tolerance investigation of full‐scale components are proposed. The damage resistance and tolerance characteristics of flat sandwich panels with thin facesheets were observed to be highly dependent on the impactor size (DOT/FAA/AR‐00/44). The blunt impactors (3” diameter, hemispherical) produced large damage areas, which were subsurface in nature and predominantly core crushing with almost negligible residual indentation, while the sharper impactors (1” diameter) produced smaller damage areas accompanied by skin fractures and considerable residual indentation depths, as illustrated in figure (1a). The residual properties of the impact damages sandwich panels were studied using in‐plane compression tests. The damage states due to blunt (3”) impactors promoted a stability‐induced failure; with the core crush region and slight indentation constituting a geometric imperfection. The damage states due to 1” impactor, produced stress concentration induced compression failures of the skins. The residual strengths associated with stability‐induced failures were consistently lower than that of compression failures (figure 1b). The implementation of these observations to address the damage tolerance issues during the design process would require additional knowledge of the effects of scaling, combined loading, presence of substructures, stress raisers (cut outs) etc.
66
Objective The primary objective of this proposed program is to conduct a series of full‐scale tests on curved panels at the FAA’s curved panel test facility in Atlantic City, NJ. The activities outlined in this proposal describe a joint effort between WSU/NIAR and the FAA Technical Center to conduct these full‐scale tests. A composite sandwich test article was designed to study the damage tolerance characteristics under combined longitudinal, hoop and pressurization loading using the FASTER test fixture. Unlike aluminum test articles, the composite test article is desired to be representative of a monocoque fuselage, i.e., without any frames or stringers. The objective of the test is to simulate the strain fields that exists under the combined longitudinal and pressurization loading in a portion of the fuselage structure. The test article should be designed in such a way that the strain fields are close to that of the actual fuselage structure, over a significant portion of the test article. The test article has to be suitably reinforced around the periphery for external/reactive load introduction to eliminate undesirable failures near the edges. The edge stiffening resulting from reinforcements along the edges must not however alter the strain distributions at sufficient distances from the edges. The following constraints and features were imposed on the design of the sandwich test article.
1. The test article should have an internal radius of 74” 2. The circumferential length must not exceed 68” and the longitudinal length must not
exceed 120”. These lengths are inclusive of the edge reinforcements for load introduction.
67
3. Due to the absence of frames, alternative method of reacting unbalanced radial loads must be identified and appropriate modifications/additions to the FASTER fixture must be made.
4. Design appropriate attachment members to connect the test article to the fixture, which was originally designed for semi‐monocoque metallic test articles.
Expected Outcomes (a) At least 18 full‐scale articles manufactured and delivered to the FAA technical center FASTER facility for full‐scale testing. These coupons will be fabricated by Adam Aircraft of Englewood, CO. (b) Coupon tests to represent the material configurations selected for the production of the full‐scale articles. (c) Perform full‐scale tests of 10 panel configurations at the FAA FASTER facility.
68
Aging of Composite Aircraft Structures: Beechcraft Starship Teardown and Decommissioned Boeing 737 Tail Principle Investigator(s): Dr. John Tomblin, Executive Director, NIAR; Lamia Salah, Manager, Fatigue & Fracture Lab; Melinda Laubach, Manager, Aging Aircraft Laboratory Background With the opportunity to use the composite decommissioned 737 tail structure that had a commercial service history of 20 years or greater, this research will focus on providing insight into the aging aspects of composite aircraft structures. Figure (1) shows a schematic of the 737 composite stabilizer to be used for this investigation. These composite stabilizers were originally put into service in 1980 and most of them have been recently decommissioned except two that are still in service by a local airline.
Figure (1). Schematic of a 737 Boeing Composite Stabilizer With the opportunity to use the Beechcraft Starship structure, this proposed research will focus on providing insight into the aging aspects of composite aircraft structures. Figure (1) below shows a picture of the Beechcraft Starship to be used for this investigation.
69
The Beechcraft Starship was developed in the late 80’s using an all‐composite construction, a variable sweep forward wing, and rear‐mounted Pratt & Whitney turboprops. The development of the Starship meant mastering a new technology, building a new manufacturing facility and training a workforce. Beechcraft/Raytheon only built 53 Starships when production was halted due to poor commercial demand. Of the 53 built, only a small handful was ever actually sold. Objectives The research proposed will be sub‐divided into small sub‐tasks to understand the aging mechanism of the composite structure which includes (but is not limited to) the following:
• Investigate the structure for cracks, delaminations, damages, repair and bond integrity if applicable • Change in mechanical properties and resin chemistry • Material degradation due to heat, humidity, ultraviolet (UV) radiation, oxidation, etc • Evaluate bearing conditions around holes and fasteners • Investigate possible bearing failures or delaminations around the holes • Evaluate effectiveness of repairs • Establish micro‐cracking fracture toughness as a function of aging (moisture exposure, heat, UV, etc) • Calculate effective diffusion constants for water absorption • Evaluate effectiveness of repairs
70
Technical Tasks Non‐destructive inspection This task will be divided into two phases. The first phase will involve inspections in accordance with current procedures. The tail sections may be sent to American Airlines facility in Tulsa, OK and/or Sandia National Laboratories for ultrasonic inspection. The flaws that are detected or undetected in this phase will be compared with those found in subsequent non‐destructive and destructive tests. The second phase will involve more advanced inspection techniques that are not yet utilized by certified inspectors (these can be done in collaboration with Sandia National Laboratories as well). Techniques such as optical inspection technology (3‐D photogrametry), laser holography, and more rigorous ultrasonic which are available at NIAR will also be used. The purpose is to determine the potential and practicality of these newer techniques with respect to composite structures. Close coordination with aircraft/airline service stations, other institutions and equipment manufacturers will ensure the success of this task. The advanced techniques which may be used of the structure during the non‐destructive and destructive phases of the program will be carried out in several stages; visual inspection (in‐fleet type of inspection) and detailed NDI. Destructive Evaluation and Inspection Verification of component structural equivalency can only be accomplished by the destructive testing of a full‐scale article. This is not the same as testing an element, specimen or subcomponent. As a rule, the design of elements and subcomponents includes provisions for load introduction that would not be a part of the production design. In addition, the tooling concepts are not identical between test parts and production parts. Therefore it is advantageous to destructively inspect a part fabricated with production tooling and processes. The objectives for performing a destructive inspection on a part are to:
• Verify that the performance properties established during coupon and element level testing (qualification and allowables) are the same in the component. • Quantify internal (hidden) defects or indications detected by non‐destructive inspection, i.e. validate non‐destructive inspection methods. • Validate laminate physical properties (resin content and thickness). • Verify fiber path continuity within joints and complicated geometries (typically features that can not be verified through a Discriminator Panel or by non‐destructive inspection).
It is important to understand the current condition of the composite structure compared to its undamaged state. Composite structure most certainly undergoes synergistic effects of the following parameters during its service life:
• Microstructural and compositional changes • Time‐dependant deformation (fatigue, creep, creep‐fatigue, stress relaxation) and resultant damage accumulation • Environmental cracks and accelerated effects of elevated temperatures (thermomechanical and environmental conditions)
71
Since both the starship and the stabilizers were in service for over 20 years, a number of repairs had also been completed on the stabilizers. Another goal of the investigation is to evaluate the effectiveness of the repairs and relate the state of the repairs to long‐standing certification, inspection and repair philosophies. The exposure of polymeric composite to the elements of nature over the years might have degraded the properties of the polymer matrix due to moisture diffusion, exposure to ultraviolet radiation and the thermal cycling of the component depending on the geographical location of the airline.
• The moisture gained by the composite component may be measured by suitably drying out the specimen. • The degradation of the polymer matrix may be in the form of a reduced glass transition temperature and reduced mechanical properties due to disintegration of long polymer chains due to ultraviolet exposure. These changes may be appraised using dynamic mechanical analysis (DMA) and thermo gravimetric analysis (TGA). The degradation of matrix and matrix/fiber interface may also be appraised by conducting short beam shear tests.
In addition, a typical geometric feature that can only be evaluated by a destructive inspection is the cocured joint. Cocured joint quality is strongly dependent on the tooling approach and design features. It is only through the fabrication of full‐scale hardware with the actual production tooling that cocured joint strength can be evaluated. Another design feature that is difficult to evaluate a destructive test is the honeycomb sandwich panel where foaming adhesive is used to splice honeycomb core to solid structure (such as ribs or spars). In many cases the foaming adhesive will migrate from the core to spar bond line into the skin to spar bond line. This migration is detrimental to bond line strength and almost impossible to detect by standard NDI techniques. It is only through destructive inspection that it can be verified that the processing techniques imposed to prevent foaming adhesive migration actually work. Possible tests for this phase of the program include moisture content, microscopy and visual inspection and thermal analysis. Expected Outcomes Data generated for this program will provide a better understanding of the aging phenomenon on the composite aircraft structure. This data will be used by the FAA to assess the efficacy of the current/ emerging certification methods; it will also be used to issue policy pertaining to usage of composites with respect to aging factors.
72
Improving Adhesive Bonding of Composites through Surface Characterization
Principle Investigator(s): Brian Flinn, Research Associate Professor, Department of Materials Science and Engineering, University of Washington
Background:
Joints represent one of the greatest challenges in the design of structures in general and in composite structures in particular. The reason for this is that joints entail interruptions of the geometry of the structure and often, material discontinuities, which almost always produce local highly stressed areas. In principle, adhesive joints are structurally more efficient than mechanically fastened joints and constitute a resource for structural weight saving because they provide better opportunities for eliminating stress concentrations; for example, advantage can be taken of ductile response of the adhesive to reduce stress peaks. Mechanically fastened joints tend to use the available material inefficiently.
Unfortunately, because of a lack of reliable inspection methods and a requirement for close dimensional tolerances in fabrication, in the past aircraft designers have generally avoided bonded construction in primary structure. Adhesive joints tend to lack structural redundancy, and are highly sensitive to manufacturing deficiencies, including poor bonding technique, poor fit of mating parts and sensitivity of the adhesive to temperature and environmental effects such as moisture. Assurance of bond quality has been a continuing problem in adhesive joints; while ultrasonic and X‐ray inspection may reveal gaps in the bond, there is no present technique that can guarantee a bond that appears to be intact does, in fact, have adequate load transfer capability. Surface preparation and bonding techniques have been well developed for metal to metal bonding, however this is not the case for composite to composite or composite to metal bonding. Techniques to achieve good bond strength in composites have been developed, but there is not a fundamental understanding of the role of surface preparation techniques at the atomistic level.
Defects in adhesive joints that are of concern include surface preparation deficiencies, voids and porosity, and thickness variations in the bond layer. Of the various defects, which are of interest, surface preparation deficiencies are probably the greatest concern. These are particularly troublesome because there are no current nondestructive evaluation techniques, which can detect low interfacial strength between the bond and the adherends. Most joint design principles are academic if good adhesion between the adherends and bond layer is poor. The principles for achieving this are well established for adherend and adhesive combinations of interest. Hart‐Smith, Brown and Wong give an account of the most crucial features of the surface preparation process. Condensate on adhesive that had not been properly stored in a sealed bag in the refrigerator has also resulted in kissing bonds that separate because of close to zero peel strength. The mechanical interlock achieved by filling the cavities in peel ply surfaces creates a ‘Velcro’‐type bond with sufficient strength to pass initial inspections, but without the durability to last in service. Use of peel plies on surfaces to be bonded has been effective in reducing contamination. Co‐cured joints have demonstrated significantly less susceptibility to shop contaminants; therefore, it is anticipated that co‐bonding will be somewhat less susceptible to improper surface preparation than secondary bonding.
73
Objective
Further understand the effect of peel ply surface preparation on the durability of primary structural composite bonds through surface analysis coupled with mechanical testing and fractography.
Investigate the effect of peel ply material, texture, and moisture content on the surface structure and bond performance of BMS8‐276 form 3 (Toray) laminates using two different adhesives.
• Peel Ply/Release Ply
o Materials: polyester, nylon and SRB
o Texture: Fine, medium and coarse weaves
o Moisture Content: dry to saturated
• Adhesive type
o MB1515‐3 vs. AF555
Expected Outcomes
To further understand the effect of surface preparation on the durability of co‐bonded and secondary bonded composite joints, samples were prepared using peel ply removal and examined by surface analysis (ESCA, SEM and Profilometry) coupled with mechanical testing and fractography. In addition, we hope to develop a fundamental understanding of the role of surface preparation techniques at the atomistic level.
1. Produce laminates with 10 different peel plies and 1 release film.
2. Send samples to other JAMS investigators (FIU and WiSU) for other characterization research.
3. Characterize chemical structure of laminates after peel ply removal using ESCA, SEM, SIMS and other techniques such as profilometry and contact angle measurement as applicable.
4. Bond laminates with Metal bond 1515‐3 and AF555 film adhesives after peel ply removal.
5. Measure Mode I fracture toughness and perform fractography.
6. Correlate surface characterization with mechanical properties.
7. Coordinate results with parallel study on long term durability of similar samples conducted by Lloyd Smith at WaSU.
74
Preliminary Results
Preliminary results to date include surface chemistry using ESCA, the mode I fracture toughness, GIC, and fractography of laminates prepared with polyester, nylon and SRB peel plies and bonded with MB 1515‐3. The fracture toughness of laminates bonded with MB 1515‐3 are given Table 1. The fracture toughness and fracture mode was strongly influenced by the peel ply material used during layup.
Polyester peel ply SRB release film Nylon peel ply
Average GIC 3.25 0.38 0.70
Standard Deviation 0.34 0.07 0.08
Failure Mode Cohes/Interlam 100% Adh. 100% Adh.
Table 1: Mode I fracture toughness of laminates bonded with Metal Bond 1515‐3 Adhesive
Clearly, the sample prepared with polyester peel ply has the best bond quality based on mode I fracture testing. Additionally, the mode of failure (cohesive or interlaminar) is more desirable than the adhesive (interfacial) failure seen in the other two samples. The interfacial failure of samples prepared with nylon and SRB is reflected in their very low GIC values. In order to better understand the reasons for these results, comparison should be made to the ESCA analysis of the laminate surfaces before bonding. Figure 1 shows the spectra collected during the survey scan for composition, while Table 2 gives the amounts of each element shown by the peaks.
Figure 1 and Table 2: ESCA Composition Scan Spectra after peel ply removal
Peel Ply % C % N % O % Si
Nylon 77.5 9.8 12.6 Poss trace
Polyester 75.5 1.9 21.6 1
SRB 68 0.9 24.2 6.9
0
500
1000
1500
2000
2500
3000
3500
02004006008001000Binding Energy (eV)
Coun
ts
Polyester
Nylon
SRB
Si
C
N
O
75
The high content of carbon in all samples is expected due to the epoxy matrix and carbon fibers. The low fracture toughness of the SRB peel ply specimens are most likely explained by the presence of the significant silicon contamination shown in the survey scan. The amount of nitrogen, especially in the sample prepared with nylon peel ply is also surprising. To gain some molecular information and to further investigate the presence of nitrogen in the samples, high resolution scanning of the C (1s) region is underway. Initial indications are that amide groups are present on the laminate surfaces after nylon peel ply removal. At this point of the research, it is still difficult to explain the poor performance of the samples prepared with nylon peel ply. The amount of C, N, and O found on the surface do not seem to correlate directly to bond quality. The transfer of an amide groups to the surface are possible cause of the poor bonding of the samples.
Summary
The fracture toughness of laminates bonded with MB1515‐3 is strongly affected by the type of peel ply material used for surface preparation. High toughness bonds and cohesive failure was obtained when polyester fabric was used. Surfaces prepared with nylon and siloxane coated polyester (SRB) fabrics resulted in very low fracture toughness and adhesive failure at the interface between the adhesive and the laminate. Characterization of the surface chemistry of laminates using ESCA after peel ply removal revealed Si on the laminate surface prepared with SRB fabric. Si compounds are known to interfere with adhesive bonding and this is the most likely explanation for the low fracture toughness. Amine groups and higher nitrogen content were detected on the laminate surfaces prepared with nylon peel ply. This may have contributed to the low fracture toughness, and further investigation is on going.
76
The Effect of Surface Treatment on the Degradation of Composite Adhesives
Principle Investigator(s): Lloyd Smith, Associate Professor, School of Mechanical and Materials Engineering, Washington State University
Background
To ensure the longevity of the commercial aircraft fleet, the long term durability of primary aircraft structure must be understood. The degradation of metals and their attachments (mechanical and adhesive) has been rigorously studied over the years. The introduction of composite materials in aerospace applications has presented challenges as methodologies that have successfully been used for metals do not always produce reliable results with new materials. Project Motivation
• Higher efficiencies in commercial aircraft are being realized through composite materials.
• Bonded joints contribute to the weight savings afforded by advanced materials. • The resistance of adhesives to long‐term degradation is not understood as well as their
adherends. • Stress can accelerate degradation and is often not considered in degradation studies.
Objective This project will consider the effect of surface treatments on composite adherends and accelerated test methods that may be used to reliably compare their long term degradation. Follow‐on projects will consider improving durability using nano‐reinforced adhesives. Goals
• Surface treatment effects o Strength o Fracture toughness o Durability
• Accelerated test methods o Wedge crack
• Model degradation o Geometry o Temperature
Expected Outcomes (first year) • Long term durability
o Consider current bond preparation practices o 140F water immersion o Residual fracture toughness after sustained environmental exposure (DCB) o Residual strength and modulus after sustained creep and cyclic loading (WLS)
• Accelerated testing o Wedge crack coupons o Adherend compliance proportional to fracture toughness o Compare crack growth as a function of bond quality
77
• Modeling degradation o Expose polymer to aggressive solvent with measurable weight change (gain or loss) o Characterize fundamental degradation parameters (D, cm, k) o Consider temperature and geometry effects o Describe the effect of surface preparation on long term degradation
• Accelerated testing o Determine if a modified composite wedge crack specimen can reliably accelerate degradation
• Modeling degradation o Experimentally verify a method to that describes polymer degradation
Fig. 1. Compact pneumatic creep that provides a constant tensile stress to coupons immersed in aggressive environments.
Fig. 2. Low profile grips that transfer load from the compact creep frame into the test coupons.