Page 682 - Automotive Engineering Powertrain Chassis System and Vehicle Body
P. 682

Interior noise: Assessment and control    C HAPTER 21.1

           connected to the receiver via sprung or compliant   6. With the engine still removed from the vehicle,
           mounts and there is a reasonable differential movement  measure the noise transfer functions (NTFs)
           across each mount. For example, the complex stiffness
           method is commonly applied to the engine mounting             P
                                                                 NTF ¼      P is the sound pressure ðPaÞ
           problem thus:
                                                                         F i
            1. Identify all probable vibration paths between the                                      (21.1.17)
              source (the engine) and the receiver (a microphone
              positioned in the passenger compartment).          using the force excitation from (5) and a microphone
                                                                 in the passenger compartment.
            2. Organiseaphasereferencesignalforuseinsubsequent
              measurements.Thismightbeanelectricalsignaltaken  7. Measure the dynamic stiffness of each engine
              from the tachometer or the conditioned output from  mount along with its damping. These may be com-
              an accelerometer attached to the engine block.     bined to give a complex stiffness (Cremer and
                                                                 Heckl, 1988)
            3. On the test track, or the rolling-road dynamometer,
              measure acceleration levels in the three axes (x, y, z)  K ¼ Kð1 þ ihÞ                  (21.1.18)
              on the engine side of the engine mounts. Analyse
              these levels either as power spectral densities or as  where K is the dynamic stiffness (N m ) and h is
                                                                                                    1
              ‘order-levels’. Order levels may be determined     the loss factor (dimensionless). A typical example of
              either by synchronous sampling relative to the ref-  the results from a measurement of engine mount
              erence phase signal or they may be extracted from  stiffness is shown in Fig. 21.1-3.
              the time-averaged power spectral density (in which  8. Double integrate the engine side accelerations
              case the magnitude of the order level is taken as the  from (3) and the body side accelerations from (4)
              average of three spectral lines, whereas the phase  and estimate the differential displacements of the
              information is taken from the middle spectral line).
                                                                 mounts. The force applied to the body at each
            4. Repeat (3) for the body side of each engine mount.  mount is therefore
              Ideally steps (3) and (4) should be undertaken
              simultaneously.
                                                                 F i ðf Þ¼ Kðf Þ$ðx s ðf Þ  x r ðf ÞÞ  (21.1.19)
            5. Remove the engine from the vehicle and measure
              the input accelerances in the three axes (x, y, z)on  where x s is the displacement on the engine side of
              the body side of the mounts due to force excitation  the mount (the source side) and x r is the displace-
              applied in each of the three axes in turn. The force  ment on the body side (the receiver side).
              excitation is applied with either a shaker or using  9. Use the forces obtained from (8) along with the
              a hammer fitted with a force transducer at its tip.  NTF from (6) to estimate the partial sound pres-
                                                      2
              The accelerance between the acceleration (m s )at  sure in the passenger compartment due solely to
              point i due to a force applied at point j is given by  vibration power transmitted across each engine
                                                                 mount.
                            a ij
              accelerance ¼                       (21.1.16)   10. Sum all of the partial sound pressure contributions
                            F j
                                                                 from all of the identified noise paths on a polar plot
                                                                 (or Nyquist diagram) and compare the result with
                                                                 the measured total sound pressure for validation of
             40 000                                              the method.
             35 000
                                                              The use of a polar plot for displaying the results is im-
             30 000
            Stiffness (kN/m)  25 000                          formation. Therefore, interference effects between the
                                                              portant as it shows both magnitude and phase in-
             20 000
                                                              contributions from different paths may be investigated.
             15 000
                                                                For noise paths (transfer paths) comprising only rigid
             10 000
                                                              (or fairly rigid) connections, the complex stiffness
              5 000
                                                              method is not suitable as there will be negligibly small
                0
                  0      100     200    300     400     500   relative displacement across the connections.
                                                                In such cases, the force imposed by the source at each
                                 Frequency (Hz)
                                                              input to the system is determined from the inverse of the
                         K [kN/m]                 PolyFit
                                                              full accelerance matrix multiplied by the accelerations
           Fig. 21.1-3 Measured engine mount stiffness (Verstraeten, 2003).  measured on the receiver side of the first connection
                                                                                                     693
   677   678   679   680   681   682   683   684   685   686   687