Shafts measuring and analysing sound produced by a splined shaft hob with changeable teeth during milling process of grooves.

Ciofu, Ciprian Dumitru ; Nedelcu, Dumitru ; Pruteanu, Octavian 等

Abstract: The accuracy of a machined part depends on the precision motion delivered by a machine tool under static, dynamic, and thermal loads. Improper working parameters is often a serious limitation to achieving higher rates of removal, as it adversely affects the surface finish, reduces dimensional accuracy, and may damage the tool and machine. The adaptation of process variables for the purpose of enhancing process efficiency is addressed within the area of control for process optimization. Machine tool monitoring and control provide the bridge between machining research and the production line (The Mechanical systems design handbook, 2002).

Key words: splined shaft hob, milling, monitoring, process optimization

1. INTRODUCTION

The accuracy of a machined part depends on the precision motion delivered by a machine tool under static, dynamic, and thermal loads. The accuracy is evaluated by measuring the discrepancy between the desired part dimensions identified on a part drawing and the actual part achieved after machining operations (Altintas & Yellowley 1987).

Usually, sound produce by a machine tool is used to identify any anomaly that could appear during production process (Moriwaki, 1980). But the sound intensity depends also by the cutting force (Koenigsberger, 1967). This correlation permits us to use measurement of the sound to appreciate the cutting force intensity and variation.

2. EXPERIMENTAL

2.1 Equipments

For testing we used a sound level meter from "Chauvin Arnoux" and measure with it sounds emitted by a milling machine for gear tooth FD-320 during production process. The device was bringed near working zone tool and it take over the noise produced during milling process. (fig.1.).

Measurements were made using a new gear milling hob for grooves shafts and compare results obtained for each ten work piece obtained with different milling parameters and with different tool wear stage and different materials. After we obtained an significant lot of pieces, using sound spectral density of the process signal delivered by the sound level meter, using Sigma Plot 2000 software, Fig. 2., we trace a graph where we superpose data obtained on one type of material with another one using same working parameters, or on the same material using different parameters and last same material, same working parameters but different stage of wearing of the tool.

2.2 Methodology

The data from the decibel meter was stored in a data file (.dat extension) which later was processed with SE322 program supplied by manufacturer. The data from file was exported in Sigma Plot 2000 program that allowed us to compare working behavior of the splined shaft hob and determinate the regress curves of noise models.

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

Using Sigma Plot software capabilities, the data form retrieved from decibel meter was statistically analyzed, obtaining mathematical models of variation in time of the noise. To not affect the result, was rejected all noise data corresponding to the approaching stage to the material of the tool. Mathematical models tested, was from Waveform family, chose from Equation Categories menu (fig. 3).

[FIGURE 3 OMITTED]

Mathematical equation most appropriate with sound variation is:

y = [y.sub.0] + a x sin(2[pi]t/b + c), [dB] (1)

where: [y.sub.0], a, b and c are constants and which are calculated by the software; t - time [s].

Mathematical function it's a sinusoid, displaced with [y.sub.0] value from origin, with 2a amplitude, 2[pi]/b frequencies and phase difference c.

There are statistical parameters and constants calculus (correlation coefficient, standard error) - fig. 4.

[FIGURE 4 OMITTED]

Regression equation obtained is:

y = 80,009 + 0,93 x sin (2[pi] x t / 12,38 + 3,775), [dB] (2)

Sigma Plot software permits us a graphical representation of this equation (fig. 5):

[FIGURE 5 OMITTED]

After processing of the experimental data, was obtained the followed regression equation: -OL60 material sample

y = 78,929 + 0,47 x sin(2[pi]t / 12,92 + 3,313 (3)

-OLC45 material sample

y = 78,957 + 0,37 x sin(2[pi]t / 12,17 + 3,18 (4)

- 34MoCr11material sample

y = 78,949 + 0,64 x sin(2[pi]t / 12,32 + 3,41 (5)

Resultant graphs from equations, for that 3 materials, was superpose for a better interpretation - fig. 6.

[FIGURE 6 OMITTED]

3. CONCLUSION

As result, using this method of measurement and control, we was able to control milling process results modifying the milling regime and parameters therefore noise produced to be as lowered as possible, to avoid or control chatter appearance before it could reach dangerous values for wok piece and hob by modifying milling parameters and re-sharpening the hob before tooth failure appear.

Also this method of control allow us to identify any kind of misaligned or whipping of the work piece or hub indicate by the general shape of the graph which could had wave form or excessive values for some work tooth's of the hob.

Expanding research, we think that is possible to partially predict behavior of the tool with different materials or different working parameters.

4. REFERENCES

Koenigsberger, F. & Tlusty, J. (1967). Machine Tool Structures, Vol. I: Stability against Chatter, Pergamon Press, Oxford,

Moriwaki, T. (1980), Annals of the CIRP, Detection for tool fracture by acoustic emission measurement, 29, 1, 35-40

Altintas, Y. & Yellowley, I. (1987), Sensors for Manufacturing Process detection of tool failure in milling using cutting force models, ASME, New York, 1-16

The Mechanical systems design handbook--Modeling, Measurement and Control, (2002) CRC Press LLC

C. Ciofu--PhD Thesis--unpublished results