Papers by Author: Hua Tao

Abstract: A developed method to predict the cutting forces in end milling of generalized corners is proposed in this paper. The cornering milling process is divided into a series of cutting segments with different cutting states. The mathematical model of the geometric relationship between cutter and the corner profile is established for each segment. Cutting forces is predicted by introducing the classical cutting force model. The computational results of cutting forces are in good agreement with experimental data.

Abstract: The GA and multi-objective optimization approach is flexible in application and extremely reliable, providing optimal results for all optimization problems attempted. Aircraft engine maintenance using nonlinear analyses. This considerably slows the entire production and maintenance process. There are three approaches for this purpose. GA with a gradient-based method to improve quality enhancements and moderate convergence efficiency; the multi-objective framework will integrate the analysis codes for multiple disciplines, instead of relying on one code to perform the analysis for all disciplines; using optimize a multi-objective model by implementation of GA instead the computationally expensive functions. Results indicate that is to provide a capability to yield a better production and maintenance in much reduced time and cost.

Abstract: Very high cycle fatigue behavior of high strength steel, were investigated using ultrasonic fatigue testing equipment at 20 kHz up to 109cycles. S-N curves at room temperature with different stress ratio (R=0.01 and R=0.1) was determined. The experimental results show that fatigue strength decrease with increasing number of cycles between 105 and 109. SEM examination of fracture surface reveals that fatigue damage was governed by the formation of cracks, and subsurface crack initiation was in the very long life range. The results shown that the portions of life attributed to subsurface crack initiation between 107 and 109 cycles are 99%.

Abstract: The selection of adaptable machining parameters is a key part in numerical control (NC) milling process. Deterministic optimal parameters are obtained without considering the uncertainties of machine tools. So a reliable optimization method was introduced in this study. We focus on a response surface methodology (RSM) adapted to reliability-based design optimization (RBDO) based on Diffuse Approximation (DA). A variant of the first order reliability method (FORM) is chosen to calculate the safety index. Case study is investigated for the NC milling operations of ultrahigh strength steel. We show how to get the optimal parameters with the safety index by the reliable optimization method.

Abstract: Two kinds of beam specimens made of composite laminated panel were designed and used to do vibration fatigue test for ε-N curves. the shapes of various simulated joint configurations for the specimens were adopted in order to get the accurate results. They were two kinds of double-cantilever beams supplied by T-mode stiffener and L-mode stiffener. All beam specimens were tested under the same shape of narrowband spectrum and the several loadings of random vibration by shaker excitation. The bandwidth of the narrowband random spectrum was 1/3 octave bandwidth whose centre frequency was the first resonance frequency of the cantilever beam, so that it reduced influences of the difference of specimens resonance frequency. Fitting vibration fatigue test results, the ε-N curves would be obtained then. The vibration test shows that this method of getting ε-N vibration curve is feasible. It can be used in practice.

Abstract: Domain knowledge representation is various and domain-concerned. The aircraft assembly process planning (A2P2) is a special domain, a lot of things should be taken into account, and the knowledge representation of A2P2 is complicated. It is focused on the knowledge representation of A2P2 in this paper. Based-on case, the framework of A2P2 knowledge is presented. The main considerations of A2P2 are analyzed, the transformation and reassembly of native A2P2 knowledge is studied, the features of A2P2 is acquired, and the formalizable framework of A2P2 is proposed. With BNF, a formal description of A2P2 knowledge is given.