Scaling the Non-linear Impact Response of Flat and Curved Composite Panels

Abstract

The application of scaling laws to thin flat and curved composite panels exhibiting nonlinear response when subjected to low-velocity transverse impact is investigated. Previous research has shown that the elastic impact response of structural configurations exhibiting geometrically linear response can be effectively scaled. In the present paper, a preliminary experimental study is presented to assess the applicability of the scaling laws to structural configurations exhibiting geometrically nonlinear deformations. The effect of damage on the scalability of the structural response characteristics, and the effect of scale on damage development are also investigated. Damage is evaluated using conventional methods including C-scan, specimen de-plying and visual inspection of the impacted panels. Coefficient of restitution and normalized contact duration are also used to assess the extent of damage. The results confirm the validity of the scaling parameters for elastic impacts. However, for the panels considered in the study, the extent and manifestation of damage do not scale according to the scaling laws. Furthermore, the results indicate that even though the damage does not scale, the overall panel response characteristics, as indicated by contact force profiles, do scale for some levels of damage.

Some Recommendations for Characterization of Composite Panels by Means of Drop Tower Impact Testing

“…It has been shown frequently in the literature that their value is independent of impact energy, and it stays constant for increasing values of available kinetic energy. [3][4][5][6]11 Furthermore, for modern composite systems, which are relatively strain-rate insensitive in this low-velocity regime, it has been found that the value of the critical force is very close to the mean static failure load (MSFL) of the laminate, obtained by means of quasi-static indentation tests.…”

“…3, but if the signal is very noisy, values of n up to 14 have been used. 4 One of the simplest smoothing techniques, as the one employed in the Dynatup 930 software, 21 takes a running average of the recorded signal over 2n + 1 points. In most cases, a value of n = 1-3 will produce satisfactory results, 11 as shown in Fig.…”

“…The resulting peak force for this pseudoclamped condition is only 3-5% greater, whereas the contact duration is around 8-10% shorter than the purely simply supported condition. 6 If proper torque is applied and is measured by means of a torque wrench, and if the clamping is performed by means of multiple through-thickness holes in the laminate, whereby the faceplate is fastened to the support plate, 4,6 it is possible to achieve peak force values up to 20% greater than simply supported ones. 13 In general, however, the amount of clamping achievable with these methods is not nearly as high as the one predicted by plate theory.…”

“…It has also been found that the signal becomes less noisy and hard to interpret if the size of the impactor is increased, and the linear system approximation becomes less accurate. 4,6 As a generic guideline, the ratio between the in-plane cutout dimension to impactor diameter should not be greater than 10. On the other hand it has been found that a larger impactor diameter has the effect of producing greater contact forces and introduces greater damage within the laminate, and so care needs to be taken in choosing the adequate size for the study.…”

“…As a general guideline, the use of maximum sustained contact force to compare impact events is a very useful but also dangerous practice, because apparently similar force-time curves can hide very different underlying states of damage. 3,4 The amount of energy dissipated in a supercritical event, given by Eqs. (3) and 4, is much larger than that of an elastic test, and it can provide a useful indication of the amount of damage introduced in the specimen.…”