Abstract
Experiments on the forming limit strains of several metal sheets are performed exclusively under the (quasi-) plane strain tension. As well known, sheet breaks generally due to localized necking (localized-type instability or bifurcation). All theories predicts that any sheet would break at the tensile strain comparable with its n-value (strain-hardening exponent) under such conditions. It is clarified here why there exist so many materials whose forming limit strains for plane strain tension are not coincident with their n-value. It is verified that the theoretical critical condition for localized necking is basically correct, in which the first-order derivative of the stress-strain relation (the strain-hardening characteristic) of the material plays the decisive role. The conventional n-th power law approximation of the strain-hardening characteristic is not reliable when its first-order derivative must be used. It is emphasized that the strain-hardening characteristic has to be formulated precisely up to its first order derivative when determining the forming limit strain.
