Vibration often exists alongside airborne noise and this is certainly the case for the vibration that is experienced in residential environments due to railway traffic. Many of the field studies described in the previous sections have shown that per- ceivable levels of vibration in combination with noise from railways can have an effect on the overall annoyance response (Fields, 1979; Gidl¨of-Gunnarsson et al., 2012; Schomer et al., 2012; Waddington et al., 2014). Indeed, Schomer et al. (2012) suggest the need to develop separate predictions for annoyance due to railway noise for railway sources that produce perceivable vibrations and for those that do not. They demonstrate that, even though railway noise is generally believed to be less annoying than road traffic noise (Miedema and Vos, 1998; Moehler, 1988; Moehler et al., 2000), when perceivable vibration is present, railway noise can actually cause more annoyance than road traffic noise.
The first laboratory studies on the subjective response to noise and vibration from railway traffic were performed by Howarth and Griffin (1990). In these studies, subjects were presented with stimuli composed of different combinations of six magnitudes of railway noise and vibration. Twenty four subjects took part in the three part magnitude estimation study. In the first session, the assessment of vibration in the presence of noise was investigated. In the second session, the assessment of noise in the presence of vibration was investigated. In the third session, the combined effects of noise and vibration was investigated. The results of these tests indicated that, within the range of stimuli magnitudes investigated, vibration does not significantly influence the judgement of noise, but the judgement of vibration may be affected by the presence of noise, depending on the magnitudes of the stimuli. They discovered that a reasonable approximation of the total annoyance caused by combined noise and vibration stimuli can be determined from a summation of the effects of the individual stimuli, using the following relations:
ψ = 15.9 + 260ϕ1v.04+ 0.167ϕ0n.039 (2.8)
ϕv = VDVb (2.9)
10 log10ϕn = SELA (2.10)
whereψ is the annoyance response, VDVb is the Wb weighted vibration dose value
and SELAis the A-weighted sound exposure level. In a follow up study, Howarth
and Griffin (1991) investigated the effects of duration, magnitude and frequency of vibration stimuli in combination with noise. This was achieved by exposing subjects to combinations of noise and vibration stimuli with different magnitudes, duration and vibration frequency. Their results suggested that an annoyance rela- tion involving a summation of the individual magnitudes of the noise and vibration stimuli provides a more accurate means of prediction of overall annoyance due to combined noise and vibration than relations based on either the noise or vibration stimuli alone. They derived a new relationship for predicting overall annoyance, which is similar to Equation 2.8 and is shown below:
ψ = 22.7 + 243ϕ1v.18+ 0.265ϕ0n.036 (2.11)
whereϕv andϕn are defined by Equations 2.9 and 2.10 respectively. Although the
relationship shown in Equation 2.11 was derived for stimuli with varying durations, the maximum duration was 29 s, and Howarth and Griffin (1991) state that the relationship may not be appropriate for predicting annoyance for combined noise and vibration stimuli with durations longer than this.
In another laboratory study, Paulsen and Kastka (1995) investigated the effects of combined noise and vibration, from a passing tram and a hammermill, on rated intensity and annoyance. Four levels of noise and vibration levels were presented to subjects in every possible combination. The results indicated that the presence of vibration influences the evaluation of noise annoyance and has a greater influence
on the evaluation of total annoyance. They developed a predictive relationship for total annoyance caused by combined tram noise and vibration exposure, which takes the same form as those developed Howarth and Griffin (1990, 1991), and is shown below:
ψ =−0.15 + 1.58 log10(vrms) + 0.11LAeq (2.12)
where vrms is the rms vibration velocity and LAeq is the A-weighted continuous
sound pressure level.
More recently, Jik Lee and Griffin (2013) performed a laboratory study looking at the combined effects of noise and vibration produced by high speed trains on annoyance. In this test, subjects were exposed to six levels of noise and six levels of vibration exposure, for both windows open and windows closed scenarios. The experiment was divided into four sessions:
1. Evaluation of noise annoyance in the absence of vibration
2. Evaluation of total annoyance from simultaneous noise and vibration 3. Evaluation of noise annoyance in the presence of vibration
4. Evaluation of vibration annoyance in the presence of noise
The results indicated that vibration did not influence ratings of noise annoyance, but that total annoyance for combined noise and vibration was significantly greater than that due to noise alone. The authors developed predictive models of the total annoyance resulting from combined noise and vibration based on two classical models: the dominance model (Rice and Izumi, 1984) and the independent effect model (Taylor, 1982). In the dominance model, the total annoyance is a function of the maximum of the single source (noise or vibration) annoyance, whereas the independence model is a function of of the annoyance of both sources, assuming that the separate sources make independent contributions to the total annoyance.
Jik Lee and Griffin (2013) found that both models provided useful predictions of the total annoyance caused by simultaneous noise and vibration from high speed trains.