Search

A machine for testing thin-walled tubes in torsion at shear-strain rates up to 25/sec is described. Results of constant and variable-strain-rate tests are presented for 1100-0 aluminum, AISI 1020 steel, and 50-A titanium. Results indicate that 1100-0 aluminum is very slightly strain-rate sensitive, but steel and titanium are noticeably sensitive to both strain rate and strain-rate history. Variable-rate tests show that subsequent dynamic loading on a statically prestrained specimen causes an increase in the flow stress in steel and a decrease in the flow stress in titanium.

A technique for determining the complex shear modulus of a linear viscoelastic material from damped free vibrations of sandwich beams is presented. The sandwich beams consist of thin steel facings and cores of a filled elastomer. In such construction the facing experiences only extensional vibrations and the core only shear when the difference in stiffness between core and facing is large. The mathematical analysis of such a system involves, in general, two complex moduli of the core. However, the relative rigidity of the facings allows the elimination of the complex Young's modulus from the equations, thereby leading to a direct evaluation of the complex shear modulus. Experimental results for a wide range of temperatures and frequencies are presented for a potassium chloride filled polyurethane solid propellant. With the aid of an Arhenius type temperature-time shift function the shear modulus and damping constant are evaluated over eight decades of frequency. (Author)

The analysis of the split Hopkinson bar experiment for determining dynamic material behavior is examined for several specific examples of specimen materials which exhibit strain-rate-dependent mechanical behavior. The torsional mode of deformation is chosen as more closely representing a one-dimensional state of stress. Details of the propagation and reflection of stress waves within the specimen are studied using a numerical procedure based on the method of characteristics. Reconstituted stress-strain curves calculated from the conventional analysis of the split Hopkinson bar experiment are compared with actual material behavior for several simulated experiments involving variations in input stress, gage length, material behavior, and static stress-strain curves including statically prestressed materials. The validity of the experimental technique is discussed and limitations on its use are delineated.