Comparació amb d’altres modes d’operació en bioreactor.

The first case evaluates the four subframe mounting points of the front axle com- bined with also four mounting locations at the rear axle. As far as the directions are concerned, only the acceleration signals that come from the z-direction at the front and rear subframe mounts are used in this case, as illustrated in figure

6.2.

Figure 6.2: The locations of the accelerometers in the vehicle structure and the directions of the reference signals. : One direction.

Overall, eight reference signals are fed into the adaptive algorithm. The acceleration signals of the z-direction help the ARNC system to reduce by 5 dB(A) the noise levels at the tyre cavity range between 90-220 Hz, a fact that is evident in figures 6.3(a) - 6.3(d). It also help the controller to attenuate the road noise levels for a resonance at 290 Hz for the left hand front microphone. The rumbe is attenuated by at least 5 dB(A) at the driver’s headrest and also of the headrest at the right hand rear seat,yet it is less than 4 dB(A) in the other microphones across the rumble range. This is apparent at the co-driver’s

headrest, where only some part of the rumble resonance is removed. It is likely that more directions are necessary to use as references for noise reduction at the rumble band behind the co-driver’s head.

(a) Left hand front microphone response. (b) Right hand front microphone response.

(c) Left hand rear microphone response. (d) Right hand rear microphone response.

Figure 6.3: Interior noise at the four headrests. −: ARNC off. −: ARNC on.

The average noise reduction of the controller in the range 0-500 Hz is slightly more than 2 dB(A) across the four microphones. The highest reduction is ob- served at the right hand front microphone with 2.8 dB(A) of active reduction in the road noise range.

LHF dB(A) RHF dB(A) LHR dB(A) RHR dB(A) Average dB(A)

2.4 1.5 2.3 2.4 2.2

Table 6.1: Total reduction for 0-500 Hz. LHF: Left hand front, RHF: Right hand front headrest, LHR: Left hand rear, RHR: Right hand rear.

6.5.1.2 Case 2: Subframe mounts and suspension z-axis

The set of the acceleration signals from the z-direction at the subframe mounts of the rear axle is maintained as shown in figure 6.4, but only two symmetrical locations at the front suspension, which observe the vibrations of z-axis, are now included in the ARNC model. Consequently, six acceleration signals are used as reference inputs to the feedforward controller. It must be pointed out that this is relatively low for an ARNC application, since several structure-borne sources are expected to be observed by the controller. The locations at the front axle are in fact rathen close to the wheel located at the lower control arm of the suspension. Therefore sufficient control of tyre cavity resonances should be easier to achieved easier.

Figure 6.4: The locations of the accelerometers in the vehicle structure. : One direction reference signals.

This change in the acceleration signal set may have implications on the con- troller’s performance at other frequencies in the rumble range. For example, in figures 6.5(a) and 6.5(b) the reduction at the rumble of the front headrests is very limited. This may be due to the fact that there is some contribution from the front axle that is not fed into the controller as input, thus it remains uncon- trolled. Interestingly, the reduction levels at the tyre cavity do not improved and the midfrequency resonance at 290 Hz is unattenuated for all the noise spectra. The general poor performance affects on the total reduction for each head- rest location, which becomes less than 2 dB(A). Still, the improvements are greater than to the trade-offs of other NVH control methods, especially the im- provements in the right hand rear microphone of the rumble. Still, they are attractive enough fto turn the implementation of this echnology into a vehicle worthwhile.

The total reduction levels in all the microphones are poorer than the ones of the previous accelerometer configurations. As a matter of fact, only 1.6 dB(A) is the average attenuation across the four headrests.

(a) Left hand front microphone response. (b) Right hand front microphone response.

(c) Left hand rear microphone response. (d) Right hand rear microphone response.

Figure 6.5: Interior noise at the four headrests. −: ARNC off. −: ARNC on.

LHF dB(A) RHF dB(A) LHR dB(A) RHR dB(A) Average dB(A)

1.5 1.2 2.1 2.5 1.8

Table 6.2: Total reduction for 0-500 Hz. LHF: Left hand front, RHF: Right hand front headrest, LHR: Left hand rear, RHR: Right hand rear.

6.5.1.3 Case 3: Subframe mounts and suspension dampers z-axis

An important part of multilink suspension systems in road noise NVH is the top of suspension at the z-direction [Wang (2010)], where high road forces are usually applied. It is worth to investigate how this road noise input location can help the active systems control the rumble. In this case, only two symmetrical subframe mounts are used as displayed in figure6.6. Apart from the top mount of the suspension strut also the bottom mounting point is also included for the front axle, thus in total ten reference signals were used for this case. The attenuation is slightly improved, especially for the front microphones in figures

6.7(a)-6.16. At the rear part of the cabin, the atternuation is improved at the tyre cavity range 180-220 Hz, but worse at the rumble of the microphone at

Figure 6.6: The locations of the accelerometers in the vehicle structure. : One dimensional accelerometers.

the left headrest. This change in the control focus deals with the fact that the acceleration singals at the front axle are close to the wheel, hence they should include spectral components that are related to tyre cavity noise. On the other hand, they do not contain any components that relate to the rumble range and also the midfrequency resonance at 290 Hz that appears at the front headrests. The limited performance of the controller is somehow expected, as they are far from the actual structural sources that cause these two road resonances at the interior noise spectra. In the following cases, we will not use the suspension damper locations anymore, since apparently they do not offer any significant improvements.

(c) Left hand rear microphone response. (d) Right hand rear microphone response.

Figure 6.7: Interior noise at the four headrests. −: ARNC off. −: ARNC on.

In comparison with the previous scenario, the total reduction levels are improved. More specifically, 2 dB(A) in the frequency range of 0-500 Hz are removed.

LHF dB(A) RHF dB(A) LHR dB(A) RHR dB(A) Average dB(A)

2.1 1.3 2.0 2.7 2.0

Table 6.3: Total reduction for 0-500 Hz. LHF: Left hand front, RHF: Right hand front headrest, LHR: Left hand rear, RHR: Right hand rear.