An investigation into the face sheet (skins) debonding of glass balsa sandwich composites Comptest Lausanne 02/2011 Dr M. Colin de Verdiere Professor J.M Dulieu - Barton, Professor R.A Shenoi and Dr J.I.R Blake

Content Introduction. Manufacture of specimens. Material and crack characteristics. Testing for debonding characterisation. Experimental results. Digital image correlation for added material information. Parameters estimation. Conclusion.

Introduction Advantages Requirement Mine blast Mine countermeasure vessels using glass-balsa sandwiched structures Weather (hail) Tool falling on deckRough seas

Variability of balsa wood Defects (gaps in between blocks) Each blocks has different mechanical properties End grain balsa sheets are made of many different blocks Balsa core coated in resin and drying in oven to avoid excessive resin absorption

Specimen manufacture Large panel Reduced panel Crack film Specimen 1 Specimen 2 Specimen 3 Specimen 4 Specimen 5 Specimen 6 Specimen 7 Specimen 8 Specimen 9 Specimen 10 Specimen 11 Specimen 12 Specimen 13 Specimen 1 Specimen 2 Specimen 3 Specimen 4 Specimen 5 Specimen 6 Specimen 7 Specimen 8 Specimen 9 Specimen 10 Specimen 11 Specimen 14 Specimen 15 Specimen 16 Specimen 17 Specimen 18 Specimen 19 Specimen 20 Specimen 21 Specimen 22 Specimen 23 Specimen 24 Specimen 25 Specimen 26

Specimen manufacture Vacuum bag Flow media Peel ply DBL 800 CSM Mat DBL 800 Peel ply Flow media Mould Balsa core Crack film

Specimen characteristics Face sheet (skin) Core Face sheet (skin) Pre crack area Width: 35 mm Crack length: 50 mm Length: 200mm Core thickness: 13-40mm Face sheet (skins) thickness: 4 mm

Specimen characteristics During debonding of the face sheet (skins) the following effects are looked at: Core thickness Crack film thickness CSM mat or no CSM mat Epoxy or vinylester resin 13 mm 40 mm 14 μ m 60 μ m 120 μ m CSM Mat No CSM Mat Epoxy resin Vinylester resin

Debond testing Mode I

Debonding of epoxy specimen Mode I (14μm crack film, core thickness 13 mm, Mat CSM)

Debonding of epoxy specimen Mode I (14μm crack film, core thickness 13 mm, Mat CSM) ( Quispitupa A, Bergreen C, Carlson LA, 2009

Debonding of epoxy specimen Effect of core thickness (14μm crack film, core thickness 40 mm, Mat CSM) Two different modes of failure depending on the specimen position in the panel: interface or wood crack propagation (just below interface) Interface crack propagation Balsa cracks propagation

Debonding of epoxy specimen Effect of crack film thickness ( μ m crack films, core thickness 13 mm, Mat CSM) Crack film thickness 60 μ m Crack film thickness 120 μ m

Mode II Debond testing

Debonding of epoxy specimen Mode II (14μm crack film, core thickness 13 mm) No mat CSM Mat CSM

Mixed Mode Debond testing

Debonding of epoxy specimen Mixed Mode (14μm crack film, core thickness 13 mm, Mat CSM)

Vinylester specimen Mixed Mode (14 μm crack film, core thickness 13 mm, Mat CSM)

Epoxy – vinylester

Digital image correlation Strain versus crack path through time Strain extraction along the crack path. The strain is extracted at different time during crack loading and propagation: Pixels 10% 5% 0% -5% -10% Strain ε yy

Parameter estimation -Face sheet and core stiffness -Stress at which the crack propagate in Mode I and II - G strain release energy rate in Mode I and II F (N) d (mm) σ : Stress to propagate the crack G

Conclusion Variability of balsa wood is important Thicker core specimens provided less reproducible results and lower G IC. Thick crack film led to unsteady crack tip initiation and propagation and should be avoided. The presence of mat layer is beneficial. Vinylester resin was weaker than epoxy resin in mix mode loading. G IC computed by recording of the crack tip location during loading. G MMB calculation scatter due to the difficulty to locate the crack. Mode II crack location was not detectable precisely to the naked eyes. To improve this reading usage of digital image correlation may help to refine the crack location.

Future work Validation of the optical crack location method in Mode I and usage in Mode II for G IIC calculation. Refinement of materials parameters Numerical validation in Mode I, II and mix mode and comparison to experimental results.

Thank you