Characteristics Of Waveform Reproduction Using The Combination Of Thin-walled Circular Energy-absorbing Tubes

The sled crash test has low cost, high fidelity and good repeatability, so it iscommonly used in the performance study of passenger restraint system as a substitutefor the real vehicle collision. The crash energy-absorbing device is the key componentin the sled crash test, it should not only absorb the energy generated in the collisionprocess, but also reproduce the real vehicle’s front crash acceleration waveform.When simulating the acceleration curve, it is necessary to confirm that the curvegetting from sled tests has the same trend as from the real vehicle, and also to makesure that there is a correspondecce of the magnitude and time of the peaks, with therelative error into10%.After analyzing the various collision energy-absorbing devices at home andabroad, this paper proposes a method of reproducing the real vehicle front crashacceleration curve through the composite structures of multiple thin-walled circularenergy-absorbing tubes. Through reading references and the sled crash simulationverification, the failure modes of thin-walled circular energy-absorbing tubes underthe axial loading are summarized, and the various factors that affect the deformationmodes are analyzed one by one.Using the front crash acceleration waveform of a mass production compact car asthe simulated object, the initial combination of energy-absorbing tubes is got by sledcrash simulatin analysis. By adjusting the material strain rate parameters C, P andfailure parameter of plastic strain EPPF to consider the combined effect of the failuremode and the strain rate, a well fitted sled simulation acceleration curve is obtained.The number, length, diameter, wall thickness, parameters of induction holes, anddistance between the induction hole and the front of energy-absorbing tubes isredefined by modified material model. It’s proved to be feasible through sled test.On the basis of the fitting results of equivalent dual-trapezia wave, trapezoidalwave approximation, and bi-slope approximation by using five thin-walledenergy-absorbing tubes, it’s conclued that all kinds of real vehicle accelerationwaveforms can be well fitted by changing the length and wall thickness, with thediameter of energy-absorbing tubes being Φ114mm.In order to improve efficiency of getting parameters of energy-absorbing tubes,It’s used that the combination of predicting with artificial neural networks and verifying with sled crash simulation. The energy-absorbing tube parameter predictionmodel based on BP neural network is established. According to the fertures of realvehicle acceleration waveforms, the neural network can predict the parameters ofthin-walled circular energy-absorbing tubes, and the relative error is less than10%.