Influence of gear grip quality on noise levels of gearbox

1. Introduction

The tested gear must be degreased and it is first mounted into special housing which has cut holes. Through these holes, a special coating is applied on the tested gear and after the running-in, the test results are recorded with a laparoscopic camera. The complete gearbox is fixed to the relevant engine block and the whole configuration is placed in a test stand by using silentblocks. Everything is done in a way so that the situation is as close to the real vehicle installation as possible. The electric motor driving is provided only from the engine (in both senses of speed – for simulation of engine braking). The half-axes are braked by real disk brakes which simulate the road resistances and load. The braking force is expressed as a load in Newton-meters. The maximum value is 180 Nm. However, the measurement is generally made up to 120 Nm – in dependence on the load capacity of the gearbox. When the values are higher, the trace does not change significantly (by reason of elastic deformation of gearing – full supporting colour appears. The maximum revolution is 30 min^-1.

2. Basic types of supporting colours

a) Full supporting colour (Fig. 1).

Full supporting colour is characteristic for a heavy load – this effect occurs due to elastic deformations and correctly set corrections and it is not acoustically evident in traffic.

b) Mirror supporting colour (Fig. 2).

Mirror supporting colour has a smaller bearing area than full supporting colour. It usually develops when the load is low (smaller elastic deformations) in normal traffic by reason of barrel gearing ($C\alpha$ a $C\beta$). When the load is heavy, it changes to full supporting colour.

c) Addendum/dedendum supporting colour (Fig. 3)

This type of supporting colour is given by an incorrect profile of the tool. It is being solved by a profile modification. Addendum/dedendum supporting colour is dangerous because of the appearance of pitting (big pressure in a small area – Hertz pressure); and in terms of noise it is one of the worst cases.

d) Corner supporting colour (Fig. 4).

In this case, the bearing area is situated in the gearing edges. It leads to unsmooth running owing to a small supporting part; again, there is a threat of pitting appearance. The main reason for this is an unsuitable angle of inclination $\beta$ or – from the technological point of view – deviation of the helix slope $fH\beta$.

e) Conical supporting colour (Fig. 5).

Conical supporting colour is formed especially because of a badly fixed workpiece in the machine or by a large deformation of a cogwheel.

f) Crossing trace.

Crossing traces can appear along the helix (Fig. 6) but also along an involute. This happens owing to a big circumferential runout $F_r$. Possible causes: eccentric hole, heat deformations, badly fixed workpiece in the machine. It is demonstrated by a fluctuating character of vibrations.

Fig. 1Full supporting colour [1]

Fig. 2Mirror supporting colour [1]

Fig. 3a) Addendum supporting colour and b) dedendum supporting colour [1]

Fig. 4Edge supporting colour [1]

Fig. 5Conical supporting colour [1]

Fig. 6Crossing trace [1]

3. Results of tracing on a serial type of construction

Fig. 7Tracing 1° backward – series

a)$M_{MOT}=$ 0 Nm a 20 Nm, 1° driving (Festrad), right side

b)$M_{MOT}=$ 0 Nm a 20 Nm, 1° driven (Schaltrad), left side

4. Tracing test results

Fig. 8Tracing 1° backward – proposal

a)$M_{MOT}=$ 0 Nm a 20 Nm, 1° driving (Festrad), right side

b)$M_{MOT}=$ 0 Nm a 20 Nm, 1° driven (Schaltrad), left side

4.1. Tracing results evaluation – proposal

When comparing the results of the previously made calculation and the results of the practical tracing test, we can state that the results correspond. Compared with the serial type of the construction, the trace and the edge supporting colour – which negatively influenced the traces on the serial type of the construction – were reduced. Due to the extension of the lateral barrel gearing, the “full supporting colour” was changed into “mirror supporting colour” which is characteristic for big values of barrel gearing.

Especially when simulating the zero load ($M_{MOT}=$ 0 Nm), the difference is clearly visible. The trace of the supporting colour was considerably reduced and it also changed its shape from an angular shape of the full supporting colour to an elliptical shape of the mirror supporting colour. In this regime, the edge supporting colour was also completely removed.

Fig. 9Results comparison from EOL – the whole spectrum

5. Results from the noise level state (EOL)

As well as all the serial gearboxes, also the test ones are assembled on the same assembly line so that the principle of the serial assembling is kept. Within the assembly, each gearbox goes through a so-called test bench (EOL). The following graphs are the results of the order analysis from this noise level in the regime of the first gear on the backward side. The figure allows us to see the differences between the gearbox and new, proposed modifications (blue curve) and serial measuring of safe gearboxes from the same day (black curves).

In the following figures, we can see the comparison of the amplitude size of the first harmonic frequency (Fig. 10) and the second harmonic frequency (Fig. 11) depending on the speed of the gearbox drive shaft.

Fig. 10Results comparison from EOL – amplitude of first harmonic frequency

Fig. 11Results comparison from EOL – amplitude of second harmonic frequency

Fig. 12Amplitudes of acceleration comparison: green – first harmonic frequency, blue – second harmonis frequency, yellow – first interharmonic frequency

Fig. 13Amplitudes of sound pressure comparison: green – first harmonic frequency, blue – second harmonis frequency

6. Conclusions

The results clearly show that – with increasing speed – the amplitude of the first harmonic frequency increases on a linear basis. Modificated gearing shows the increase of this amplitude by an average of 5 to 10 dB in comparison with the serial type of the construction of the gear.

The second harmonic frequency has a slightly different course. We can observe a local minimum of the values of the amplitude which is – in serial type of construction – ranged between 2400 and 2700 min^-1. When considering the modificated type of construction, it is slightly moved to higher speed. As for the amplitude size, the results are comparable in low speed, higher in middle speed and rather lower in high speed – as shown in the graph.

Graphs (Fig. 12 and 13) show the comparison of amplitudes:

1) Graph in the Fig. 12 compares the amplitudes of acceleration of the first and second harmonic frequencies, including the interharmonic one.

2) Graph in the Fig. 13 compares the amplitudes of the sound pressure of the first and second harmonic frequency from the driving test, from the microphone. The first harmonic frequency of the sound pressure in the vehicle cab was reduced by 5 dB. Therefore, the noise impact on the vehicle crew was slightly reduced as well.

From the diagnostic point of view, the amplitude modulation around the first harmonic frequency was also completely removed.