AbstractThe Iosipescu shear test, also known as asymmetric four point bending of a V-notched beam,is frequently used for measuring in-plane shear properties of composites. The ASTM standard(ASTM D-5379-05) regulates how the test is to be performed. It prescribes a notch openingangle of 90° independently of the material tested, although this has proven to produceinhomogenous strain distributions in the test region (between the notches) for orthotropicmaterials. Commonly, strain gauges are attached in the center of the test region where thedeviation from average strain is high. Thus, systematic errors in the measurement in the rangeof 10% or more may be introduced. The modified Iosipescu shear test, presented in this thesis, uses a variable notch opening angledepending on the material orthotropy and orientation to accomplish even stress- and strainfields in the test region. The variable notch opening angle accommodates both anisotropicmaterials and their orientation. Based on an elastic rescaling theory for orthotropic materials,the geometry was rescaled to recreate the same stress distribution in the test region as forisotropic materials. Specifically the notch opening angle was rescaled depending on theorthotropic ratio, the ratio of the two in-plane principal stiffnesses (Ex/Ey), to obtain theoptimal notch geometry. The rescaling procedure has been verified numerically with FEsimulationsand experimentally for several materials of different orthotropic ratio showingthat this was a very feasible method. Using a whole field optical measurement system duringtesting, significantly more homogenous strain fields were observed than for the standardspecimen geometry. Thus, there is no longer any need for correction factors, relying on FEsimulation,to obtain correct shear moduli. Constitutive shear properties and strength can thusbe more accurately measured, more completely and with fewer sources of error. Notablyhigher shear strengths at larger strains were also recorded compared to standard testing.The function of the new fixture was evaluated and compared with the standard Wyomingfixture. Combined in-situ 3D deformation measurements of both the new fixture and thespecimen showed that out of plane specimen deformation was very low and substantiallylower than the Wyoming fixture. Thus considerably lower parasitic stresses are introducedwith the new fixture. Recommendations regarding fastening of the specimen were determined based on simpleanalysis combined with FE-calculations and experiments. For both isotropic and orthotropic itwas found favorable if the clamp load used to hold the specimen and the expected net peakload and were set about equal. This reduces the risk of failure outside the test region bycrushing, brushing, splitting and etc. The same effects as shown in the FE-simulations werealso observed experimentally and of similar relative magnitude.Problems with differences in strains arising on the front and back face of the specimen duringtesting have been frequently reported in the literature. This is believed to stem from deviationsfrom nominal specimen geometry such as non-parallel and/or non-perpendicular boundingsurfaces. Three types of these combinations were evaluated numerically and the two mostsignificant were confirmed experimentally. The most critical geometrical deviation assessedwas a specimen with slightly conical cross section in the gripping region. For both isotropicand orthotropic materials, very small deviations from nominal geometry, caused unacceptablylarge errors in measurements of constitutive behavior / QC 20100827
| Identifer | oai:union.ndltd.org:UPSALLA1/oai:DiVA.org:kth-9234 |
| Date | January 2008 |
| Creators | Melin, Niklas |
| Publisher | KTH, Hållfasthetslära (Avd.), Stockholm : KTH |
| Source Sets | DiVA Archive at Upsalla University |
| Language | English |
| Detected Language | English |
| Type | Doctoral thesis, comprehensive summary, info:eu-repo/semantics/doctoralThesis, text |
| Format | application/pdf |
| Rights | info:eu-repo/semantics/openAccess |
| Relation | Trita-HFL, 1654-1472 ; 0459 |
Page generated in 0.0021 seconds