Size Effects in Impact Damage of Composite Sandwich Panels

Size Effects in Impact Damage of Composite Sandwich PanelsAlan Dobyns - Sikorsky AircraftWade Jackson - Army VTD at NASA LangleyABSTRACTPanel size has a large effect on the impact response and resultant damage level of honeycomb sandwichpanels. It has been observed during impact testing that panels of the same design but different panel sizeswill show large differences in damage when impacted with the same impact energy. To study this effect, atest program was conducted with instrumented impact testing of three different sizes of sandwich panels toobtain data on panel response and residual damage. In concert with the test program, a closed formanalysis method was developed that incorporates the effects of damage on the impact response. Thisanalysis method will predict both the impact response and the residual damage of a simply-supportedsandwich panel impacted at any position on the panel. The damage is incorporated by the use of anexperimental load-indentation curve obtained for the face-sheet/honeycomb and indentor combination understudy. This curve inherently includes the damage response and can be obtained quasi-statically from arigidly-backed specimen or a specimen with any support conditions. Good correlation has been obtainedbetween the test data and the analysis results for the maximum force and residual indentation. Thepredictions can be improved by using a dynamic indentation curve. Analyses have also been done usingthe MSC/DYTRAN finite element code.INTRODUCTIONThe application of composite materials tohelicopter structure often includes use of thinface sheet honeycomb structure (e.g. RAH-66Comanche). These structures must be designedto sustain barely visible impact damage atultimate load, which can result in a reduction of50% or more in the design strength relative to anundamaged structure. The design strength ofsandwich structure is usually related to a face-sheet instability known as wrinkling. Since thefailure mechanics associated with lace sheetwrinkling combined with impact damage are notwell understood, the design allowables forsandwich structure are normally obtainedempirically using coupon tests (ref. 1, 2, 3).However, the damage produced and the residualstrength in actual sandwich structure due to agiven impact energy can be quite different fromthat produced in the small coupon specimens.This is primarily because the larger panel is moreflexible than a small panel of the same basicdesign. The most important parameterdetermining the structural response to an impactis the effective spring constant at the impactpoint. This is influenced by the panel size, panelPresented at the ,.Imerican ltelicopter Socie O,58't_Annual Forum, Montreal. Quebec. June I1 13,2002.Copyright C2002 by the American HelicopterSocieo,, Inc. All rights Reservedboundary conditions, and the impact location onthe panel.This paper discusses impact testing of sandwichpanels and an impact analysis for sandwichstructures which is able to predict the impactdamage resulting from a given impact energy inpanels of different size with user specified impactlocations. The sandwich impact analysis entailsan extension of an existing closed-form platetransient analysis method (ref. 4) to account forthe unique force-indentation behavior ofhoneycomb sandwich. In order to model thesoftening type tlonlinear load-indentationbehavior shown by the sandwich, the impactanalysis was modified to accept an experimentalload-indentation curve instead of the power lawHertz equation normally used for solidlaminates. This method implicitly accounts forthe face sheet failure and core crushing seen insandwich impact tests. It can also be used withdynamic load-indentation Curves obtained frompanel or coupon impact tests.To provide data for correlation of the analysis,impact tests were conducted on various piecesremoved from a 3' x 5' sandwich panel cut froma scrapped component. Impact tests at severalimpact energies have been conducted on theoriginal 3' x 5' panel, and the panel was then cutinto smaller panels to obtain 3' x 2.5', 1.5' x 2.5'panels. These panels were impacted to give arange of panel sizes and impact energies for theanalysis development. Further impact testing isscheduled on 12" x 12", and 6" x 6" panels tocomplete the panel size study. Static indentation