Vibration prediction capabilities of helicopter airframe connections

Reliable analysis of structural dynamics to avoid resonance has been a long term problem for the helicopter industry. Although efforts such as the NASA DAMVIBS Program have resulted in documentation and improvement of industry modeling practices, the improvements have been limited to lower frequencies. Lack of reliable predictions available in a timely manner have adversely affected the structural design process, and dynamic related problems are still a source of many redesign tasks after the airframe enters service. Better analysis idealization control techniques are needed to improve and qualify the reliability of dynamics analysis results.; This study focuses on a combined experimental/numerical investigation aimed at identifying the physical mechanisms and modeling errors that are the culprits for the lack of a correlation between experimental measurements and analytical predictions for the panels with fastened connections. The first objective is to investigate several factors including dimensional reduction and simplification, idealization errors of the connection details, nonlinearities, or the presence of friction and damping. Two typical connection details using the HI-LITE and the rivet fastener common to the helicopter airframe design are represented as lap joints transferring membrane loads and butt joints carrying transverse loads and stiffening. The second, to identify relative contributions of these various modeling errors and physical mechanisms to the lack of correlation between experimental and numerical results. The third, to propose more realistic idealizations of jointed regions by adjustments to assumptions conventionally made on structural connections modeled as rigid for the ultimate strength analysis.; Measurements obtained by the impact hammer test normally applied to testing linear structures are used to quantify relative contributions of two modeling methods; the beam reinforced element method and the double thickness element method. Next the shaker test is used to investigate connection details and nonlinearities; slip mechanism or softening effect on the fastener assembly. A conceptual methodology for more accurate estimation of joint flexibility is developed to experimentally extract stiffnesses through introduction of elastic constraints using the spring stiffness at the tip of the cantilever beam. A standard test methodology is proposed necessary for investigations of modal parameter sensitivities in terms of natural frequency and mode shape on which experimental predictions of stiffness in a joint are based. With an assumption of nonconstant normal force in a joint, the energy loss coefficient behaves like hysteretic damping, not like Coulomb damping. Actual hysteretic damping levels for isolated beams with a single joint are measured by the triad point method.; Based on experimentally measured stiffnesses, the upper bound of beam element size associated with each effect of moment, shear or axial force is suggested and then its size within the bound is tuned for natural frequency correlations. A circular cross section definition for Timoshenko beam properties is used to minimize the number of fastener stiffness parameters in the proposed N-point contact method where beam elements are equally spaced around the perimeter of the fastener. Natural frequencies predicted from the proposed N-point method are compared with those measured by shaker tests simulating actual loading situations in the joint. Dynamic prediction capabilities are much improved using this proposed model. (Abstract shortened by UMI.)...