IMPACT BEHAVIOR AND RESIDUAL STRENGTH OF SANDWICH STRUCTURAL ELEMENTS UNDER STATIC AND FATIGUE LOADING

Transcription

1 AIAA IMPACT BEHAVIOR AND RESIDUAL STRENGTH OF SANDWICH STRUCTURAL ELEMENTS UNDER STATIC AND FATIGUE LOADING Michael Gaedke * Jens Baaran, Hans Christian Goetting, Raimund Rolfes Institute of Structural Mechanics DLR German Aerospace Center Lilienthalplatz 7, D Braunschweig, Germany ABSTRACT 1 The extend of impact damage in sandwich structures depends on the core material and the stacking sequence of the sandwich skin laminate, the size, mass and velocity of the impactor and on the ability of the component to absorb the shock at the impact point. Because of the complex interaction between these parameters the forecast of the damage and the progress under Tension-Tension (T-T), Tension-Compression (T-C) and Compression- Compression (C-C) fatigue loading is difficult to conduct. Impact damage is most critical, when the skin remains intact. Except of a small dent or blister in the skin surface the impacted zone is often barely visible. Nevertheless, the damage may grow under fatigue loading to a critical size, where the component is endangered. In sandwich structures with foam core a delamination will be detected most likely around the impacted zone between skin and core. The damage progress is triggered by the local buckling of the delaminated skin layer. In structures with Honeycomb core delaminations are hardly detected after impact (see Figure 1). The core is crumpled in the vicinity of the impacted zone and the elastic support of the skin layers is reduced. Under compressive loading the core shrinks in thickness direction and the skin layer may buckle in core direction. INTRODUCTION The impact damage can be analyzed by 3Delements, often combined with extended plate elements for the skin layers. Due to the large deformations the postbuckling behavior of the growing damaged zone must be considered. This kind of analysis is quite expensive and will hardly be applied to actual damage configurations. In order to overcome these disadvantages the Analysis Tool * Dr. of Mechanical Engineering Copyright 2001 by the first author. Published by the American Institute of Aeronautics and Astronautics, Inc. with permission COmposite Damage Tolerance Analysis Code (CODAC) as a Windows and NT compatible computer program was developed. The model is complex enough to cover the essential geometrical and damage configurations and it is numerically inexpensive enough to perform trend analyses of the damage growth in reasonable time. The analysis tool contains free parameters in order to achieve conservative results. With CODAC, an impact damage can be forecasted and/or a detected impact damage can be analyzed in a sandwich plate. Also trends for the damage growth can be evaluated, which allows a first estimation of the damage tolerance behavior of the structure. Additionally it can be used to establish guide lines for the design. As can be drawn from experimental results in the open literature, high stress concentrations appear at damaged zones in structural components which lead to additional damages combined with a damage growth and ultimately to a failure of the component. The so long accomplished researches however show, that fiber-reinforced plastics (FRP) are able to reduce (disintegrate) stress concentrations at notches or damages (such as impact damage) by stress rearrangements ( Notch Root Blunting Effect ). In different researches it was found that the life characteristics (S/N curves) of undamaged and damaged structures are narrowing each other with increasing numbers of load cycles and are intersecting in the range of N = 10 6 to 10 7 load cycles (LC). This leads to the fact that damaged structural components have the same durability as undamaged components at relative very low load levels. These results were mostly stated in relevant investigations with solid laminates manufactured from Prepregs. Only rare experiences are available at this time for structural components out of fabrics and especially out of sandwiches. Therefore investigations were carried out with numerous sandwich samples the skin of which was build up with unidirectional CFRP- Prepregs as well as CFRP- and GFRP-fabrics with different fiber orientations and stacking numbers. 1

2 The core itself consisted of Nomex-Honeycomb with different thickness and weight per volume. In the test program undamaged test pieces as well as tests samples after impact were investigated under quasi-static as well as fatigue loading. The impacts were generated by means of an improved computerized impact testing machine (different values of energy and measurements of time depending load and depth of penetration). The quasistatic tests for the determination of characteristic materials data of stiffness and strength for the different test-configurations were executed with pure material test samples and sandwich specimens under tension and compression loading. The relevant fatigue tests comprised alternating loads in the TCload regime with R = -1 in two different test modes: Fatigue tests in one load stage for every individual test sample cycled at one load level up to ultimate failure or fracture; multistage fatigue tests for every individual test sample loaded at different load levels in a block diagram. The results of both the test procedures were compared via statistical methods and models (such as Miner Rule). During the tests the load and the global distortion - and by this the stiffness change - was measured continuously. In addition an optical microscope was used to detect and measure cracks and delaminations at the surface. For the judgement upon type and extent of impact damages and for the characterization of the damage propagation during the fatigue loading some of the specimens were recorded by X-ray and/or ultrasonic C-scans. ANALYSIS TOOL CODAC Analytical methodology Figure 2 Figure 3 Figure 4 Numerical results Figure 5 Figure 6 Figure 7 Figure 8 Analytical/experimental correlation Figure 9 SANDWICH STRUCTURAL ELEMENT Composite sandwich Figure 10 Impact behavior Figure 11 Damage detection and assessment Figure 12 Figure 13 Figure 14 Figure 15 RESIDUAL STRENGTH Static loading Figure 16 Figure 17 Figure 18 Figure 19 Fatigue loading Figure 20 Figure 21 Figure 22 Damage detection and assessment Figure 23 2

3 CONCLUSION / OUTLOOK 3

4 D 1 D 2 1 REFERENCES M. Gädke and Eggers, H. (in memoriam): Modeling the Damage Behavior in Sandwich Structures after Impact, Sandwich Construction 3, Vol. II, pp , Gädke, M. and Kirschke, L.: Experimental Investigation of Damage in Sandwich Structures after Impact, Sandwich Construction 3, Vol. II, pp , Gädke, M. et al.: Methodology for Residual Strength of Damaged Laminated Composites, paper no.: AIAA , pp , Proceedings of the 38 th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conference and Exhibit, Kissimmee, FL, April 7-10, 1997 (Electronic Library, AIAA Meeting Papers on Disc, Vol. 2, No.:2, 1997). 4 McGowan, D. M. and Ambur, D. R.: Damage Tolerant Characteristics on Composite Fuselage Sandwich Structures with thick Face Sheets, NASA Technical Memorandum , February Damages from Impactanalyses Buckling analysis Soft inclusion A) Delaminationen assumed as elliptical areas/spaces: B) Buckling load determination of individual sub-laminates Detail A-A t2 t4 t3 t1 A A thickness of sub-laminate Figure 3 Buckling analysis with CODAC * * COmposite Damage tolerance Analysis Code N y 1. Basic region 1 st Buckling: N1 N x N x Buckling analysis 2 nd Buckling: N2 2a 2. Basic region 2b N y 3. Basic region 3 rd Buckling: N3 Impact: after Impact: Honeycomb Core Skin webs crumbled or broken D 1 webs drawn out, but damaged! Figure 1 Deformation of Skin and Honeycomb Core of a Sandwich Element Calculation of damages Subdividing of elements in Units Calculation of stresses for each Unit across laminate thickness x y xy yz zx * improved for z Application of the damage criteria for the evaluation of Fiber breakage Matrix cracking Delaminations Figure 2 Analytical Methodology 4

5 Enhancements of CODAC include: Extension of CODAC for application to stringer-stiffened structures Enhanced method for assessment of the complete 3D stress tensor for the 2D FE Method Impact locations from mid-bay impact to impact in the area of the stringer-foot Enhancement of the buckling analysis to account for quasistatic delamination growth Implementation of various damage criteria for fibre breakage, matrix cracking and delamination (Hashin, Puck, Choi/Chang, etc.) DLR Impact Analysis with CODAC * Figure 4 Improvements of Analytical Tool CODAC Perspective view of calculated impact damage delamination matrix crack fibre failure Impact 800 Figure 5 Composite structural element (wing) used as a first benchmark Verlauf der Kontaktkraft bei einer Impactenergie von 30 J Verlauf der Kontaktkraft bei einer Impactenergie von 30 J Simulation CODAC 5.00 Simulation CODAC Experiment Experiment 1 Experiment Experiment Kraft / kn Kraft / kn Zeit / ms Zeit / ms stringer stiffened CFRP panels, impact in the midbay Figure 7 Transient analysis with CODAC in comparison with experiments Figure 6 Numerical results with CODAC for damages in skin laminates stiffened CFRP panel impact stringer foot edge Figure 8 Numerical results with CODAC 5

6 0, , w Interface (IF): (0)2 +45 (-45)2 +45 (0) Impact, 40 J 1. IF: Experiment Simulation 3. IF: Experiment Simulation 5. IF: Experiment Simulation 7. IF: Experiment Simulation No data 2. IF: Experiment Simulation 4. IF: Experiment Simulation 6. IF: Experiment Simulation 8. IF: Experiment Simulation No data Figure 9 Analytical / test correlation stringer stiffened CFRP panel impact in the mid-bay width = 100 mm 160 mm Figure 12 Damage detection in sandwich elements skin delaminations after 10 Joule impact L Wabenausrichtung Z Haut: [+45/0/-45/90/ 90/-45/0/+45] 8 UD-Lagen 6376/HTA 0,25 mm (Soll) Dicke (2 mm Soll) 1,3 mm Ist grd 0 grd Faserrichtungen 90 grd 100 bzw ~18,5 Klebfilm:?M 300 Kern: Coremaster C Nomex Honeycomb 50,16 kg/m, 18,4 mm Aufleimer: E-Glas [ 45] ~3,5 mm Frei ( 18,4 mm Alukern bei Fertigung und Versuch) 2,0 + ~ 3,5 = ~ 5,5 Impact 5 Joule 10 Joule 15 Joule Figure 13 Damage detection in sandwich elements global damage in 160 mm wide samples Updated: gädke Figure 10 Composite sandwich component for experimental studies Impact 5 Joule 15 Joule Figure 14 Damage assessment for analysis tool global damage in 160 mm wide samples C-Scans of B-Scans through the backwall echo of whole thickness thickness Figure 11 US-Inspection of a Sandwich with 4 (top) and 18 (bottom) Joule Impact Width 160 mm 100 mm Figure 15 Detailed damage assessment for CODAC skin damages after 5 Joule impact 6

7 Figure 16 Undamaged reference specimen N = 0 LC N = 90% N to failure N = N to failure Figure 20 Impacted (18 Joules) Sandwich Specimen under Fatigue Loading (R = -1) CAI Sandwich 100 mm Druckbruchlast, kn Impactenergie, Joule Figure 17 Static Compression after impact (CAI) versus Impact energy, 100 mm wide samples Figure 21 Effect of impact energy on damage size for honeycomb sandwiches 40 kn Fundamaged Load Level F F im pa cted R=-1 Configuration: A 2 Configuration: D Configuration: A 1 10 Imp. "0" Imp. " " D Configuration: F Figure 18 Global damage after 20 Joule impact 160 mm wide sandwich component CAI Sandwich 160 mm 4 6,5 10 Joule 18 Impact Energy four different sandwich configurations Figure 22 Effect of impact energy on decrease of fatigue strength of Honeycomb sandwiches Druck-Bruchlast, kn Impactenergie, Joule Figure 19 Static Compression after impact (CAI) versus Impact energy, 160 mm wide samples Global damages skin delamination Figure 23 Global and detail damage assessment for CODAC 5 Joule 100 mm wide sandwich component 7