Effects of Repair p Procedures on Bonded Repairs of Composite Structures

Transcription

1 Effects of Repair p Procedures on Bonded Repairs of Composite Structures 2011 Technical Review John Tomblin, PhD Executive Director, NIAR Lamia Salah Sr. Research Engineer, NIAR

2 FAA Sponsored Project Information Principal Investigators & Researchers Dr. John Tomblin, Wichita State University Lamia Salah, Wichita State University Mike Borgman, Spirit Aerosystems Dr. Bill Stevenson, Irish Alcalen, Wichita State University FAA Technical Monitor Curtis Davies, Lynn Pham Other FAA Personnel Involved Larry Ilcewicz, Peter Sheprykevich Industry Participation Mike Borgman, Spirit Aerosystems April 21 st,

3 Motivation/ Background In-service experience has shown that a composite part in service will absorb as much as 1.2% by weight of moisture throughout its lifetime and may be exposed to maintenance fluids such as hydraulic fluid, jet fuel and paint stripper. Leaking hydraulic and fuel lines will contaminate the composite part that may later need to be repaired. Damage threats are another concern particularly impact events that result in little or no visual surface indication. Damage to a bonded repair is of particular interest as no visible indication of damage may be present and the resulting damage may be different in size and morphology depending on the impact site in the repair. Ref: Aniversario, R. B., Harvery, S. T., McCarty, J. E., Parsons, J. T., Peterson, D. C., Pritchett, L. D., Wilson, D. R. and Wogulis, E. R., Design, Ancillary Testing, Analysis, and Fabrication Data for the Advanced Composite Stabilizer for Boeing 737 Aircraft, NASA CR , Volume II, Final Report, December April 21 st,

4 Objective To investigate the effects of pre-bond contamination and impact damage on the ultimate strength and durability of bonded scarf repairs to laminate structures To investigate different variables on the performance of repairs applied to moderately thick solid laminates To substantiate the strength and durability (mechanical loading) of OEM repairs applied to moderately thick composite laminates To substantiate the strength and durability (mechanical loading) of field repairs applied to moderately thick composite laminates To evaluate the ultimate strength of bonded repairs subjected to contamination prior to repair To evaluate the damage tolerance of repairs subjected to BVID inflicted at three different locations on the repair scarf joint To provide recommendations pertaining to process improvement to ensure repair bond repeatability and structural integrity April 21 st,

5 Laminate Repair Coupon Configuration Parent Substrate: highly toughened system, 350 F cure, autoclave processed Adhesive System: Cytec metalbond 1515, 350 F cure (OEM repair) Adhesive System: FM300-2, 250 F cure (Field repair) Repair Material, same as parent but processed under vacuum, cured at 350 F (OEM repair) Repair Material, ACG MTM45-1 processed under vacuum, cured at 250 F (Field repair) Single Scarf Joint, 4 wide to isolate the variables investigated April 21 st,

6 Pre-Bond Contamination Investigation To evaluate the strength of contaminated repairs applied to laminate configurations. Five different contaminants are considered: Hydraulic oil (skydrol), jet fuel (JP8), paint stripper, water and perspiration. The effects of each one of the contaminants is being evaluated according to the proposed test matrix. Variables: substrate Moduli, scarf rates (1:10 and 1:20), 32ply unidirectional laminates Parent Material 350 F cure system, 250 F out of autoclave repair material, FM300-2 adhesive Contamination application: Jet Fuel and Skydrol-30 day exposure perspiration p (salt water) applied to the surface prior to bonding paint stripper panels exposed to the contaminant for 6 days Plies 32 Modulus E1 E2 Scarf Rate Test Contaminant Condition SH JF PS PR WA75 WA50 WA25 WA0 CT RTA RTA W RTA CT RTA W April 21 st,

7 Water Contamination WA75/ WA50/ WA25/ WA0 Pre-Bond Moisture: Water contaminated panels were conditioned prior to repair until moisture equilibrium was achieved (1% weight gain referred to as 100% saturation level) ply-WA-1 32ply-WA ply-WA ply-WA % Saturation/ Moisture Equilibrium i re Uptake % Moistu % Saturation/ Moisture Equilibrium 50% Saturation/ Moisture Equilibrium % Saturation/ Moisture Equilibrium Time (days) April 21 st,

8 Pre-Bond Contamination Investigation Scarf Machining Scarfed Panels Adhesive Layer FM300-2 Repair Implementation April 21 st,

9 Pre-Bond Contamination Investigation Repair Implementation Repair Bagging Tabbed Panel Mechanical Testing April 21 st,

10 Pre-Bond Contamination Investigation ACG RTA ACG RTF RTA Water Contaminated Panel Jet Fuel Contaminated Panel Process yielded repairs with various levels of porosity as illustrated by the C-Scan images. Possible source of variability in the mechanical data TTU does not detect weak/ Contaminated bonds April 21 st,

11 Surface Analysis of Contaminated Scarfed Panels Dr Bill Stevenson/ Irish Alcalen High surface free energy = efficient wetting before cure High surface free energy DOES NOT NECESSARILY EQUAL a good bond A surface contaminated with polar compounds analyzed with water will yield a high SFE Contaminated surfaces with SH, PR and WA75 had a high SFE but a poor bond Effect of the contaminant Contaminant t occupies secondary bonding sites on the scarfed surface (cured resin system and fibers in parent) Contaminant prevents the uncured adhesive from occupying those sites This lessens full adhesion of the cured adhesive to the fiber/resin surface During cure cycle, contaminant migrates to and from the surface into the adhesive and can chemically react with the adhesive at high temperatures Contaminant, in sufficient quantities, brought to the surface during cure may phase separate to form porosity April 21 st,

12 Surface Analysis of Contaminated Scarfed Panels Dr Bill Stevenson/ Irish Alcalen All contact angle/sfe experiments used water as the testing liquid Water is a very polar liquid (with a powerful permanent dipole, caused by the hydrogen adopting a partial positive charge and the oxygen a partial negative charge) This allows the water molecule to strongly bind to polar sites on the scarfed surface. WATER as a contaminant Increased the apparent scarfed surface surface free energy (free energy does not translate into a better bond) For the purposes of the contact angle measuring experiment, the surface contaminated with water is very polar and the analysis is conducted with water itself which is very polar causing the water drop to wet the water contaminated surface SH (skydrol) as a contaminant Exhibited a high SFE because SH is a very polar compound, thus attracting the very polar water drop, which in turn wets the surface April 21 st,

13 Pre-Bond Contamination Investigation Sources of Variation in Mechanical Data: Thickness variation/ Variation in scarf length from panel to panel/ Variation in the lay-up process (RH, operators, surface preparation, repair ply orientation), Variation in compaction level/ bondline porosity from panel to panel (debulking the repair system has not yielded repeatable results) :10 Scarf Rate Tested at -65 F Perspiration WA75 WA50 WA25 WA0 1:10 Scarf Rate Tested at RTA 1:20 Scarf Rate Tested at RTA 1:20 Scarf Rate Tested at 180W 100 % Baseline Strength E E1-10-RTA E1-20-RTA E W April 21 st,

14 Pre-Bond Contamination Study- Results :10 Scarf Rate Tested at -65 F Perspiration WA75 WA50 WA25 WA0 1:10 Scarf Rate Tested at RTA 1:20 Scarf Rate Tested at RTA 1:20 Scarf Rate Tested at 180W % Baseline Stre ength E E2-10-RTA E2-20-RTA E W Ref: Abdelkader, A., F. and White, J.R., Water Absorption in Epoxy Resins: The effects of the crosslinking agent and curing temperature, Journal of Applied Polymer science, vol 98, water absorption may cause irreversible ibl changes in the epoxy network, absorption and diffusion i of water in polymeric material is related to the free volume which depends on molecular packing (degree of cure), that water molecules that attach to the polymer through H bonds disrupt the interchain H bonds, induce swelling and plasticize the polymer. April 21 st,

15 Pre-Bond Contamination Study- Summary The mechanical test data confirmed the detrimental effects of pre-bond contamination on the static strength of bonded scarf repairs. Individual data point recorded for perspiration and water contaminated panels yielded strength reductions exceeding 33%. Individual data points recorded for jet fuel, paint stripper and skydrol contaminated panels yielded strength reduction exceeding 7%. Environmental durability and effects of cyclic loading have to be considered Proposed method (contact angle measurement) provide promising results Good knowledge of the chemical composition of the parent and repair systems/adhesives is required for proper interpretation of the results Substantiation has to be performed at the most aggressive environments the repaired structure will be subjected to. Rigorous material, structural and process substantiation is crucial to ensure the structural integrity of repairs. April 21 st,

16 Damage Tolerance Study - Objectives To evaluate the strength, durability and damage tolerance of scarf repairs applied to laminate structures (To substantiate the effects of different impact sites on bonded repairs and their effect on residual strength) Tip of the scarf far side TF Tip of the scarf TN Parent Panel Repair Panel Parent Panel Repair Panel Parent Panel Repair Panel Center Impact P Parent Panel Repair Panel Parent Panel Repair Panel Parent Pane el Repair Panel Reference CMH17-3G Chapter 12 Damage Resistance, Durability and Damage Tolerance: Damage tolerance provides a measure of the structure s ability to sustain design loads with a level of damage or defect and be able to perform its operating functions. Consequently, the concern with damage tolerance is ultimately with the damaged structure having adequate residual strength and stiffness to continue in service safely until the damage can be detected by scheduled maintenance inspection (or malfunction) and be repaired or until the life limit is reached. The extent of damage and detectability determines the required load level to be sustained. Thus, safety is the primary goal of damage tolerance. April 21 st,

17 Damage Tolerance Study Summary Plies Modulus E1 E2 E1 E2 Scarf Rate Test Condition Impact Site TN TF CN RTA RTF RTA RTF RTA RTF RTA RTF RTA RTF RTA RTF RTA RTF RTA RTF Variables Investigated: Three impact sites on scarf joint (TN, TF or CN) Substrate thickness, Moduli, Scarf rate Same normalized energy level for all configurations Tension Loading mode (ultimate strength, residual strength after fatigue Parent material used as repair (processed under vacuum) 18 ply configurations (1.2 impactor) Impact Energy Level 200 in-lbs, Depth: ply configurations (1.2 impactor) Impact Energy Level 400 in-lbs, Depth: 0.01 April 21 st,

18 Damage Tolerance Study NDI Results Same impact energy level yielding different damage sizes depending on impact site (TN, TF or CN) 18ply-E2-20-TN-180W W Damage area (0.642, 0.61,0.6444) in 2, Depth (0.0085, 0.008, ) 18ply-E2-20-TF-180W Damage area (2.28, 0.96, 0.95) in 2, Depth (0.0095, , ) 18ply-E2-20-CN-180W Damage area (1.8852, , ) in 2, Depth (0.009, 0.007, ) April 21 st,

19 Damage Tolerance Study NDI Results Same impact energy level yielding different damage sizes depending on impact site (TN, TF or CN) 48ply-E2-10-TN -RTA Damage area (1.905, 1.878,1.869) in 2, Depth (0.0085, 0.011, ) 48ply-E2-10-TF -RTA Damage area (2.5648, ,1.7528) in 2, Depth (0.011, 0.009, ) 48ply-E2-10-CN -RTA Damage area (7.218, 6.30,6.9596) in 2, Depth (0.008, 0.008, ) April 21 st,

20 Damage Tolerance Study - Results ply substrate 1:10 scarf rate 14 in 2 bond area 48 ply substrate 1:20 scarf rate 28 in 2 bond area 6 Damage Are ea [in 2 ] % 15% 17% 18% 32% 40% 9% 7% 11% 10% 14% 15% 10-TN 10-TF 10-CN 20-TN 20-TF 20-CN 48 PLY-E1 48 PLY-E2 April 21 st,

21 Damage Tolerance Study - Results ply-E1-10-TN-RT Epsilon Y [% %] Section Length [in] April 21 st,

22 Impact Damage Study - Results 48ply-E2-10-TF-RT Ep psilon Y [%] Section Length [in] April 21 st,

23 Damage Tolerance Study - Results 48ply-E2-10-CN-RT-1 Epsilon Y [% %] Section Length [in] April 21 st,

24 Damage Tolerance Study - Results Undamaged Strength 1:20, E1 48PLY-E1 48PLY-E Undamaged Strength 1:20, E2 Ulti imate Load [lbs] Undamaged Strength 1:10, E1 Undamaged Strength 1:10, E TN-RTA 10-TF-RTA 10-CN-RTA 20-TN-RTA 20-TF-RTA 20-CN-RTA April 21 st,

25 Damage Tolerance Study - Summary Strength degradation is a function of damage area The same impact energy level applied to various locations in the scarf joint yielded different damage areas Coupons impacted at the center of the scarf repair, yielded the largest damage area and the lowest static strength The residual strength is also dependent on the residual bond area. The largest repairs were observed to be more damage tolerant than the smaller repairs The 48-ply-E1-10 specimens survived cycles at a load level equivalent to 2500με, 2000με and 1000με for TN, TF and CN. The 48-ply-E2-10 specimens survived cycles at a load level equivalent to 1500με for all configurations tested All 48-ply-E1-20 specimens were cycled at a load equivalent to 3000με far field and all specimens tested survived cycles of fatigue except 48-E1-20-CN- RTF-01 which h failed at cycles. All 48-ply-E specimens were cycled at a load equivalent to 3000με far field and the resulting average cycles to failure were cycles, and cycles for the TN, TF and CN respectively Conclusions are based on specimens tested at room temperature. Need to establish fatigue capability of repairs April 21 st,

26 A look Forward/Benefits to Aviation Repairability has to be addressed during the design phase (OEM/ factory repair system substantiation-limitations associated with out of autoclave systems) To understand the effects of contamination and the resulting implications on the strength of a repaired structure To provide recommendations pertaining to repair technician training and repair process control To understand the effects of damage on residual strength of a repaired structure To identify the crucial steps in bonded repair that can be used to develop rigorous repeatable repair processes April 21 st,

27 End of Presentation. Thank you. April 21st,