Be based on the turning quality control of power spectral density

In Comparison With The Presently Successful Turning Quality Control Method Based On The Roughness, it Can Fully Util~ize All Measured Data Of A Real Profile, and Find The Effect Of Each Factor On The Power Spectral Density.

Primary Study Shows That The Turning Quality Control Method Based On The Power Spectral Density Has A Good Future.

Keywords: Surface of spare parts of turning of 　 of Power Spectral Density of 　 of 　 of Quality Control of Turning 　 　 is microcosmic geometrical appearance has very big effect to its function, like wearability, corrosion resistance, contact stiffness, fatigue intensity and cooperate property to wait. Besides, surface of turning spare parts still contained the information of process of a large number of concerned cutting tool, machine tool, turning and hardware data function in microcosmic geometry appearance, if surface roughness of surface of turning spare parts is right turning process the change of each element is very sensitive. According to the change of surface roughness of surface of turning spare parts, the concerned element that adjusts turning process in time controls turning quality effectively, this is current the relatively successful fundamental that is based on quality of surface roughness turning to control a method. Use discovery actually, because exterior surface roughness is parameter of assess of a quality, it desertioned the other in real profile is useful information, come out the influence distinction of all sorts of element very hard, the exact fixed position that gives an issue brings inconvenience. Article key discusses a kind of turning quality that is based on power spectral density to control a method, it can make full use of the data of all actual measurement of real profile and turning process each element is opposite the effect of power spectral density is apparent the area is apart. The abecedarian considers to make clear, this method is a method of turning quality control that has an outlook very much. The power spectral density of 1 turning spare parts takes one real profile in the exterior edge axial of turning spare parts, the power spectral density of this real profile is the power spectral density of turning spare parts with respect to the definition. Normally, what actual measurement gets is real profile only in finite a sampling data alignment, cannot get true power spectral density, can undertake estimating to its only. Be like below a few kinds of chart commonly usedly to estimate a method. 1.

Direct way calls 1 direct way periodic plan law again, the N that it is a real profile X(n) nods sampling data XN(n) to regard one energy as finite signal, the Fourier commutation that takes XN(n) directly gets XN(ejW) , next the square of its amplitude divides reassume with N, the estimation that regards true to X(n) power spectral density as P(ejW) (EjW) : (1) will (1) the EjW in type it is worth while that in unit circle classy interval takes: Brief write down for: The method that afore-mentioned chart estimate included undermentioned hypothesis and measure: ① broad sense X(n)} of ｛ of smooth and random real profile regards each as voice ergodic, with firstly example X(n) will replace ｛ X(n)} , and use the N of X(n) only sampling is worth XN (N) the P(ejW) of power spectral density that will estimate X(n) . ② 　 XN (N) be regarded as cycle in time domain and frequency domain with XN(k) , meantime region cycle is N, corresponding frequency domain cycle is 2 π . Because the XN(k) in direct way can use FFT to be calculated quickly, so (K) also can be begged conveniently piece.

1.

Indirect method of 2 indirect method gives autocorrelation function R(m) by XN(n) estimation first, seek the Fourier shift that gives R(m) next, the power spectral density that gets XN(n) estimates (EjW) : Because this kind of means gets power spectral density secondhand through autocorrelation function, call auto-correlation the law again so. What need points out is, when M bigger, be close to especially when be equal to N-1, (M) greatens to the estimation deviation of R(m) , the quality of power spectral density that because this estimates,gives drops subsequently. Usually, use indirect method takes M<<N-1. 1.

The couplet of 3 direct ways and indirect method uses tool of FFT of have the aid of, direct way and indirect method are OK conveniently couplet is used. 2 quality control the exterior quality of principle turning spare parts to suffer a lot of elements to affect, the influence of error of the vibration that if machine tool bearing wears away,causes, slideway, cutting tool appearance and wear away the influence of degree, the influence of feed, and the influence of hardware data function. With respect to influence degree character, the effect of feed is most apparent, it is appearance of cutting tool geometry next, wear away the action such as degree and machine tool vibration, the influence of material function is relatively a few weaker. Usually, the real profile of turning spare parts is by afore-mentioned each element the result of joint action, decompose its very hard on time domain as corresponding as each element heft. The method that power spectral density of have the aid of analyses, each heft that can overlap time domain comes out in frequency domain depart. Such, according to the analytic result of power spectral density, can judge turning spare parts to whether be put in exterior quality problem not only, and the specific element that can point out influence surface quality further changes. Be based on quality of turning of power spectral density to if the graph is shown 1 times,control a system. When machine tool state, each element and hardware data function produce course of cutting tool, turning when the change, the quality of spare parts surface with good turning will have marked change. Through power spectral density the analysis can discover these change and get the information of corresponding element change, this information seasonable feedback gives concerned operation staff, let them adjust machine tool of working parameter, overhaul, make turning process returns good working position afresh thereby, assure the exterior quality of turning spare parts. Graph 1 be based on quality of turning of power spectral density to control a system the effectiveness of afore-mentioned methods, depend on greatly the proper analysis of power spectral density of pair of turning spare parts. In current system, had not realized the online analysis of power spectral density, the data of real profile needs to be inputted from clavier artificially. Additional, because this method used a kind of brand-new train of thought, do not have off-the-peg data and experience to be able to offer reference, our analytic method and standard are the summary to carrying out experience only, still need farther theory to discuss. Nevertheless, existing experimental data confirmed turning process is occurring between each element and power spectral density of turning spare parts adequately already apparent dependency. If the analyses turning spare parts typical power spectral density that music of 3 turning part pursues pursues 2 are shown. Graph the analysis that the typical power spectral density of 2 turning spare parts passes data of many actual measurement, the power spectral density that discovers turning part has following laws: The typical power spectral density of turning spare parts can be divided commonly for 4 component, among them the peak of taller amplitude 1 as corresponding as feed, the peak with inferior frequency group the state such as the 2 vibration with the machine tool is corresponding, the peak with tallish frequency group the 3 geometrical appearance with cutting tool and wear away rate is corresponding, the peak with very tall frequency group 4 material function with the spare parts are concerned. Besides the qualitative conclusion of above, still can use a peak 1 with the peak group the amplitude of 3 undertakes quantitative analysis than wearing away to cutting tool. Current, the analysis of power spectral density of turning spare parts still is in experience to increase the level that referenced chart pursues, the most difficult question is to lack uniform analytic standard. The standardization that realizes analysis of power spectral density of turning spare parts and intelligence change the way that is further effort, the warehouse of chart of turning part music that builds content to perfect is the main base that this method promotion uses. 4 last words of article introduction be based on new method of pilot of quality of turning of power spectral density, the exterior quality problem that can discover turning part not only and still can point out the reason that poses quality problem further. The abecedarian considers to make clear, the power spectral density of turning spare parts can be divided for 4 component, each element is having they and turning process very good dependency. Be based on quality of turning of power spectral density to control is a new research field, having wide perspective. The standardization of the versatility that how raises this method and analysis, intelligence is changed, the joint efforts that still remains to go at wide Great Harmony. CNC Milling