Research On Sampling Points Distribution And Data Processing For Blisk Measurement

There is a trend in aviation industry that blisk which has advantages of lightweight, fewer parts, higher reliability quickly spread to aviation engines instead of theblade structure. Vane-integrated Disk which is a high-precision parts, high precisionmanufacturing, twisting the blade angle, blade passage space narrowing, facesproblem of manufacture. How to process precise and the high qualified rateVane-integrated Disk have become the global focuses of research. The detection is akey to ensure high quality blisk, which purpose is to obtain allowance guiding thesubsequent finishing of the workpiece. Two-column series coordinate axis machinetools is adopted, which is high control accuracy and realize integration of processingand measuring. The touch trigger probe is adopted for measurement, which accuracyand repeatability is high but the efficiency is low. Improving measurement efficiencyon the premise of ensuring the accuracy of blisk has a great significance for shorterdevelopment cycle. Around this difficulty, the following tasks are completed.After the adequate research for the free-form surface characterization, the methodof three non-uniform B-spline is adopted to interpolate the raw data points. Theacquired interpolated curve is regarded as a design curve which provides the basis forsubsequent measurement. The global B-spline approximation algorithm is studieddeeply and the closed data points’ reconstruction algorithm is deduced. Theapproximation of the measured data is finished on the matlab software, whichapproximate extent is adjusted by the reconstruction error E.The basic principles and common methods of complex surface sampling pointsdistribution are analyzed. To overcome the difficulty of improving measurementefficiency on the premise of ensuring the accuracy of blisk, an algorithm suited for theblade section of the sampling point distribution is proposed, named as adaptiveplanning method. The simulation system used to compare the merits of the samplingpoint distribution algorithm is established. Based on the system the chord toleranceand adaptive planning method are compared. The data in the863project by the chordtolerance and adaptive planning method are tested. It is proved that the proposedmethod can meet the principle of sampling point density is positively correlated with the curvature, which result is better than the result of chord tolerance, improving themeasurement accuracy and measurement efficiency.The sort algorithms based on distance, based on polar coordinates, anddichotomy are interpreted, which shortcomings are discussed. Pointing out these sortmethods cannot effectively sort profile section data of the blade. To solve the problemof profile section data, an algorithm suited for the blade section is proposed, namelymean line. The work about sorting of profile section data is done well.Machining allowance and processing orientation were confirmed throughcalculating the distance between the measurement point and its nearest point on thedesign curve in this article so as to avoid the complicated intersection operationsbetween the measurement curve and the normal vectors of the points on the designcurve. Two field tests were taken to verify the correctness of the algorithm raised inthis article.