Vibration and comfort I. Translational seat vibration

A series of studies of discomfort caused by multi-axis vibration at the seat, feet and back of seated persons is described. This first paper reports on studies with translational seat vibration. Two experiments concerned with the effects of level, frequency and direction of the translational vibration of a firm flat seat are reported.

At octave centre frequencies from 1 to 63 Hz the first experiment determined the levels of fore-and-aft, lateral and vertical seat vibration which caused discomfort equivalent to 0.5 and l.25m/s²r.m.s. 10 Hz vertical seat vibration. In the second experiment, comfort contours equivalent to 0.8m/s²r.m.s. 10 Hz vertical seat vibration and subject transmissibilities were determined from 18 males and 18 females at preferred third-octave centre frequencies from 1 to 100 Hz. In both studies the feet of subjects were not vibrated and there was no backrest.

It was concluded that the shapes of equivalent comfort contours need not normally depend on vibration level. The forms of both individual and group equivalent comfort contours and seat-to-head transmissibilities are presented. Significant correlations were found between subject characteristics (size and transmissibility) and subject relative discomfort. The males and females produced similar equivalent comfort contours.

Information on the computerized application of the method of constant stimuli which was developed for the series of experiments is presented together with a consideration of alternative methods of determining the central tendency of the data. A method of assessing the effect of vibrator distortion on judgements of equivalent discomfort is also defined.     

Bedside safety rails: assessment of strength requirements and the appropriateness of current designs

This second paper in a series of studies of the discomfort produced by multi-axis vibration is concerned with rotational seat vibration. The effects of level, frequency and direction of the roll, pitch and yaw vibration of a firm flat seat have been studied in two experiments. At octave centre frequencies in the range 1-31.5 Hz the first experiment determined the levels of roll, pitch and yaw seat vibration which caused discomfort equivalent to 0-5 and l.25m/s²r.m.s. 10 Hz vertical seat vibration. In the second experiment, comfort contours equivalent to 0.8 m/s² r.m.s. 10 Hz vertical seat vibration were determined from 18 males and 18 females at preferred third-octave centre frequencies from 1 to 31.5 Hz. In all cases the axis of rotation passed through the centre of the seat surface. There was no vibration of the feet and no backrest.

It was concluded that the shape of equivalent comfort contours need not normally depend on vibration, level. Both individual and group equivalent comfort contours are presented. Although there were significant correlations between subject size and subject relative discomfort it is not thought that these correlations have much practical application. In all three axes the median contours of vibration acceleration increase in proportion to vibration frequency. Sensitivity is greatest for roll vibration and least for yaw vibration of the seat.     

The vibration of inclined backrests: perception and discomfort of vibration applied parallel to the back in the z-axis of the body

This study determined how backrest inclination and the frequency of vibration influence the perception and discomfort of vibration applied parallel to the back (vertical vibration when sitting upright, horizontal vibration when recumbent). Subjects experienced backrest vibration at frequencies in the range 2.5 to 25 Hz at vibration magnitudes up to 24 dB above threshold. Absolute thresholds, equivalent comfort contours, and the principal locations for feeling vibration were determined with four backrest inclinations: 0° (upright), 30°, 60° and 90° (recumbent). With all backrest inclinations, acceleration thresholds and equivalent comfort contours were similar and increased with increasing frequency at 6 dB per octave (i.e. velocity constant). It is concluded that backrest inclination has little effect on the frequency dependence of thresholds and equivalent comfort contours for vibration applied along the back, and that the W _d frequency weighting in current standards is appropriate for evaluating z-axis vibration of the back at all backrest inclinations.

Statement of Relevance: To minimise the vibration discomfort of seated people, it is necessary to understand how discomfort varies with backrest inclination. It is concluded that the vibration on backrests can be measured using a pad between the backrest and the back, so that it reclines with the backrest, and the measured vibration evaluated without correcting for the backrest inclination.     

Vibration and comfort IV. Application of experimental results

In this conclusion to a four-part series of papers, an interpretation of the experimental data presented in the earlier three papers is given. The results obtained from the experiments are compared with those obtained by other researchers. A set of simplified approximations to the various equivalent comfort contours are defined and a general approach to the prediction of discomfort due to multi-axis, multi-input complex vibration is given.

It is shown that the results for translational vibration of a seat surface supporting the body are reasonably consistent with some previous studies. Although the contours for fore-and-aft and lateral vibration are in agreement with those in ISO 2631 (1978) there is evidence that sensitivity to vertical seat vibration above 8 Hz should be greater than defined in that standard. It is shown that the results for rotational seat vibration differ by an order of magnitude from those published by other authors but that this difference may be explained by the positions of the centres of rotation used in other studies.

It is shown that simple approximations to the median equivalent comfort contours for translational and rotational seat vibration and translational foot and back vibration may be defined by contours on acceleration frequency graph having a slope of O dB per octave (constant acceleration) and 6 dB per octave (constant velocity). For most conditions these contours fall within the 25th-75th percentile of the experimental data. The application of these contours as frequency weightings for general vibration evaluations is discussed.     

The vibration of inclined backrests: perception and discomfort of vibration applied parallel to the back in the z-axis of the body

This study determined how backrest inclination and the frequency of vibration influence the perception and discomfort of vibration applied parallel to the back (vertical vibration when sitting upright, horizontal vibration when recumbent). Subjects experienced backrest vibration at frequencies in the range 2.5 to 25 Hz at vibration magnitudes up to 24 dB above threshold. Absolute thresholds, equivalent comfort contours, and the principal locations for feeling vibration were determined with four backrest inclinations: 0° (upright), 30°, 60° and 90° (recumbent). With all backrest inclinations, acceleration thresholds and equivalent comfort contours were similar and increased with increasing frequency at 6 dB per octave (i.e. velocity constant). It is concluded that backrest inclination has little effect on the frequency dependence of thresholds and equivalent comfort contours for vibration applied along the back, and that the W (d) frequency weighting in current standards is appropriate for evaluating z-axis vibration of the back at all backrest inclinations. STATEMENT OF RELEVANCE: To minimise the vibration discomfort of seated people, it is necessary to understand how discomfort varies with backrest inclination. It is concluded that the vibration on backrests can be measured using a pad between the backrest and the back, so that it reclines with the backrest, and the measured vibration evaluated without correcting for the backrest inclination.     

Response of the seated human body to whole-body vertical vibration: discomfort caused by sinusoidal vibration

Frequency weightings for predicting vibration discomfort assume the same frequency-dependence at all magnitudes of vibration, whereas biodynamic studies show that the frequency-dependence of the human body depends on the magnitude of vibration. This study investigated how the frequency-dependence of vibration discomfort depends on the acceleration and the force at the subject–seat interface. Using magnitude estimation, 20 males and 20 females judged their discomfort caused by sinusoidal vertical acceleration at 13 frequencies (1–16 Hz) at magnitudes from 0.1 to 4.0 ms^? 2 r.m.s. The frequency-dependence of their equivalent comfort contours depended on the magnitude of vibration, but was less dependent on the magnitude of dynamic force than the magnitude of acceleration, consistent with the biodynamic non-linearity of the body causing some of the magnitude-dependence of equivalent comfort contours. There were significant associations between the biodynamic responses and subjective responses at all frequencies in the range 1–16 Hz.

Practitioner Summary: Vertical seat vibration causes discomfort in many forms of transport. This study provides the frequency-dependence of vibration discomfort over a range of vibration magnitudes and shows how the frequency weightings in the current standards can be improved.     

Equivalent comfort contours for vertical seat vibration: effect of vibration magnitude and backrest inclination

This study determined how backrest inclination and the frequency and magnitude of vertical seat vibration influence vibration discomfort. Subjects experienced vertical seat vibration at frequencies in the range 2.5-25 Hz at vibration magnitudes in the range 0.016-2.0 ms(-2) r.m.s. Equivalent comfort contours were determined with five backrest conditions: no backrest, and with a stationary backrest inclined at 0° (upright), 30°, 60° and 90°. Within all conditions, the frequency of greatest sensitivity to acceleration decreased with increasing vibration magnitude. Compared to an upright backrest, around the main resonance of the body, the vibration magnitudes required to cause similar discomfort were 100% greater with 60° and 90° backrest inclinations and 50% greater with a 30° backrest inclination. It is concluded that no single frequency weighting provides an accurate prediction of the discomfort caused by vertical seat vibration at all magnitudes and with all backrest conditions. PRACTITIONER SUMMARY: Vertical seat vibration is a main cause of vibration discomfort for drivers and passengers of road vehicles. A frequency weighting has been standardised for the evaluation of vertical seat vibration when sitting upright but it was not known whether this weighting is suitable for the reclined sitting postures often adopted during travel.     

Transmission of vertical vibration through a seat: Effect of thickness of foam cushions at the seat pan and the backrest

The perception of vehicle ride comfort is influenced by the dynamic performance of full-depth foam used in many vehicle seats. The effects of the thickness of foam on the dynamic stiffness (i.e., stiffness and damping as a function of frequency) of foam cushions with three thicknesses (60, 80, and 100 mm), and the vibration transmitted through these cushions at the seat pan and the backrest were measured with 12 subjects (6 males and 6 females). With increasing thickness, the stiffness and the damping of the foam decreased. With increasing thickness of foam at the seat pan, the resonance frequencies around 4 Hz in the vertical in-line and fore-and-aft cross-axis transmissibilities of the seat pan cushion and the backrest cushion decreased. For the conditions investigated, it is concluded that the thickness of foam at a vertical backrest has little effect on the vertical in-line or fore-and-aft cross-axis transmissibilities of the foam at either the seat pan or the backrest. The frequencies of the primary resonances around 4 Hz in the vertical in-line transmissibility and the fore-and-aft cross-axis transmissibility of foam at the seat pan were highly correlated. Compared to sitting on a rigid seat pan with a foam backrest, sitting with foam at both the seat pan and the backrest reduced the resonance frequency in the vertical in-line transmissibility of the backrest foam and increased the associated transmissibility at resonance, while the fore-and-aft cross-axis transmissibility of the backrest was little affected. Compared to sitting without a backrest, sitting with a rigid vertical backrest increased the resonance frequency of the fore-and-aft cross-axis transmissibility of the seat pan cushion and increased the transmissibility at resonance.Relevance to industryThe transmissibility of a seat is determined by the dynamic properties of the occupant of the seat and the dynamic properties of the seat. This study shows how the thicknesses of foam at a seat pan and foam at a backrest affect the in-line and cross-axis transmissibilities of the foams at the seat pan and the backrest. The findings have application to the design of vehicle seats to minimise the transmission of vibration to the body.     

Transmission of roll and pitch seat vibration to the head

A series of experiments has investigated the transmission of roll and pitch seat vibration to the heads of seated subjects. Head motion was measured in all six axes using a light-weight bite-bar while seated subjects were exposed to random motion at frequencies of up to 5 Hz at 1.0 rad.s ^?2 r.m.s. Subjects sat on a rigid flat seat in two body postures: ‘back-on’ (back in contact with backrest) and ‘back-off’ (no backrest contact). The influence of the position of the centre of rotation was also investigated.

Motion at the head occurred mostly in the lateral, roll and yaw axes during exposure to roll seat vibration and in the fore-and-aft, vertical and pitch axes during exposure to pitch seat vibration. A reduction in the magnitude of head motion occurred when the subjects sat in a 'back-off' posture compared with a 'back-on' posture. Varying the position of the centre of rotation along the lateral axis during roll seat vibration affected vertical and pitch head motion: least head motion occurred when the centre of rotation was in line with the subject's mid-sagittal plane. Varying the position of the centre of rotation along the vertical axis during roll seat vibration affected head motion in the mid-coronal plane: roll head motion decreased as the position of the centre of rotation was raised from below the seat surface to above the seat surface. Varying the centre of rotation (along the fore-and-aft and vertical axes) during pitch seat vibration altered head motion in the mid-sagittal plane. Head motion increased with increasing distance of the centre of rotation in front or behind the subject's ischial tuberosities and increased as the seat was raised from below the centre of rotation to above the centre of rotation.     

Equivalent comfort contours for fore-and-aft,lateral, and vertical whole-body vibration in the frequency range 1.0 to 10 Hz

Abstract

Standards assume vibration discomfort depends on the frequency and direction of whole-body vibration, with the same weightings for frequency and direction at all magnitudes. This study determined equivalent comfort contours from 1.0 to 10?Hz in each of three directions (fore-and-aft, lateral, vertical) at magnitudes in the range 0.1 to 3.5?ms^?2?r.m.s. Twenty-four subjects sat on a rigid flat seat with and without a beanbag, altering the pressure distribution on the seat but not the transmission of vibration. The rate of growth of vibration discomfort with increasing magnitude of vibration differed between the directions of vibration and varied with the frequency of vibration. The frequency-dependence and direction-dependence of discomfort, therefore, depended on the magnitude of vibration. The beanbag did not affect the frequency-dependence or direction-dependence of vibration discomfort. It is concluded that different weightings for the frequency and direction of vibration are required for low and high magnitude vibration.

Practitioner summary: When evaluating whole-body vibration to predict vibration discomfort, the weightings appropriate to different frequencies and different directions of vibration should depend on the magnitude of vibration. This is overlooked in all current methods of evaluating the severity of whole-body vibration.     

Effects of elastic seats on seated body apparent mass responses to vertical whole body vibration

Apparent mass (AM) responses of the body seated with and without a back support on three different elastic seats (flat and contoured polyurethane foam (PUF) and air cushion) and a rigid seat were measured under three levels of vertical vibration (overall rms acceleration: 0.25, 0.50 and 0.75 m/s²) in the 0.5 to 20 Hz range. A pressure-sensing system was used to capture biodynamic force at the occupant-seat interface. The results revealed strong effects of visco-elastic and vibration transmissibility characteristics of seats on AM. The response magnitudes with the relatively stiff air seat were generally higher than those with the PUF seats except at low frequencies. The peak magnitude decreased when sitting condition was changed from no back support to a vertical support; the reduction however was more pronounced with the air seat. Further, a relatively higher frequency shift was evident with soft seat compared with stiff elastic seat with increasing excitation.     

The application of SEAT values for predicting how compliant seats with backrests influence vibration discomfort

The extent to which a seat can provide useful attenuation of vehicle vibration depends on three factors: the characteristics of the vehicle motion, the vibration transmissibility of the seat, and the sensitivity of the body to vibration. The ‘seat effective amplitude transmissibility’ (i.e., SEAT value) reflects how these three factors vary with the frequency and the direction of vibration so as to predict the vibration isolation efficiency of a seat. The SEAT value is mostly used to select seat cushions or seat suspensions based on the transmission of vertical vibration to the principal supporting surface of a seat. This study investigated the accuracy of SEAT values in predicting how seats with backrests influence the discomfort caused by multiple-input vibration. Twelve male subjects participated in a four-part experiment to determine equivalent comfort contours, the relative discomfort, the location of discomfort, and seat transmissibility with three foam seats and a rigid reference seat at 14 frequencies of vibration in the range 1–20 Hz at magnitudes of vibration from 0.2 to 1.6 ms⁻² r.m.s. The ‘measured seat dynamic discomfort’ (MSDD) was calculated for each foam seat from the ratio of the vibration acceleration required to cause similar discomfort with the foam seat and with the rigid reference seat. Using the frequency weightings in current standards, the SEAT values of each seat were calculated from the ratio of overall ride values with the foam seat to the overall ride values with the rigid reference seat, and compared to the corresponding MSDD at each frequency. The SEAT values provided good predictions of how the foam seats increased vibration discomfort at frequencies around the 4-Hz resonance but reduced vibration discomfort at frequencies greater than about 6.3 Hz, with discrepancies explained by a known limitation of the frequency weightings.     

Qualitative models of seat discomfort including static and dynamic factors

Judgements of overall seating comfort in dynamic conditions sometimes correlate better with the static characteristics of a seat than with measures of the dynamic environment. This study developed qualitative models of overall seat discomfort to include both static and dynamic seat characteristics. A dynamic factor that reflected how vibration discomfort increased as vibration magnitude increased was combined with a static seat factor which reflected seating comfort without vibration. The ability of the model to predict the relative and overall importance of dynamic and static seat characteristics on comfort was tested in two experiments. A paired comparison experiment, using four polyurethane foam cushions (50, 70, 100, 120 mm thick), provided different static and dynamic comfort when 12 subjects were exposed to one-third octave band random vertical vibration with centre frequencies of 2.5 and 5.5 Hz, at magnitudes of 0.00, 0.25 and 0.50 m.s^-2 rms measured beneath the foam samples. Subject judgements of the relative discomfort of the different conditions depended on both static and dynamic characteristics in a manner consistent with the model. The effect of static and dynamic seat factors on overall seat discomfort was investigated by magnitude estimation using three foam cushions (of different hardness) and a rigid wooden seat at six vibration magnitudes with 20 subjects. Static seat factors (i.e. cushion stiffness) affected the manner in which vibration influenced the overall discomfort: cushions with lower stiffness were more comfortable and more sensitive to changes in vibration magnitude than those with higher stiffness. The experiments confirm that judgements of overall seat discomfort can be affected by both the static and dynamic characteristics of a seat, with the effect depending on vibration magnitude: when vibration magnitude was low, discomfort was dominated by static seat factors; as the vibration magnitude increased, discomfort became dominated by dynamic factors.     

Perception of fore-and-aft whole-body vibration intensity measured by two methods

This experimental study investigated the perception of fore-and-aft whole-body vibration intensity using cross-modality matching (CM) and magnitude estimation (ME) methods. Thirteen subjects were seated on a rigid seat without a backrest and exposed to sinusoidal stimuli from 0.8 to 12.5 Hz and 0.4 to 1.6 ms^? 2 r.m.s. The Stevens exponents did not significantly depend on vibration frequency or the measurement method. The ME frequency weightings depended significantly on vibration frequency, but the CM weightings did not. Using the CM and ME weightings would result in higher weighted exposures than those calculated using the ISO (2631-1, 1997) W_d. Compared with ISO W_k, the CM and ME-weighted exposures would be greater at 1.6 Hz and lesser above that frequency. The CM and ME frequency weightings based on the median ratings for the reference vibration condition did not differ significantly. The lack of a method effect for weightings and for Stevens exponents suggests that the findings from the two methods are comparable.     

Fore-and-aft and dual-axis vibration of the seated human body: Nonlinearity,cross-axis coupling,and associations between resonances in the transmissibility and apparent mass

This study examined how the apparent mass and transmissibility of the human body depend on the magnitude of fore-and-aft vibration excitation and the presence of vertical vibration. Fore-and-aft and vertical acceleration at five locations along the spine, and pitch acceleration at the pelvis, were measured in 12 seated male subjects during fore-and-aft random vibration excitation (0.25–20 Hz) at three vibration magnitudes (0.25, 0.5 and 1.0 ms⁻² r.m.s.). With the greatest magnitude of fore-and-aft excitation, vertical vibration was added at 0.25, 0.5, or 1.0 ms⁻² r.m.s. Forces in the fore-and-aft and vertical directions on the seat surface were measured to calculate apparent masses. Transmissibilities and apparent masses during fore-and-aft excitation showed a principal resonance around 1 Hz and a secondary resonance around 2–3 Hz. Increasing the magnitude of fore-and-aft excitation, or adding vertical excitation, decreased the magnitudes of the resonances. At the primary resonance frequency, the dominant mode induced by fore-and-aft excitation involved bending of the lumbar spine and the lower thoracic spine with shear deformation of tissues at the ischial tuberosities. The relative contributions to this mode from each body segment (especially the pelvis and the lower thoracic spine) varied with vibration magnitude. The nonlinearities in the apparent mass and transmissibility during dual-axis excitation indicate coupling between the principal mode of the seated human body excited by fore-and-aft excitation and the cross-axis influence of vertical excitation.Relevance to industryUnderstanding movements of the body during exposure to whole-body vibration can assist the optimisation of seating dynamics and help to control the effects of the vibration on human comfort, performance, and health. This study suggests cross-axis nonlinearity in biodynamic responses to vibration should be considered when optimising vibration environments.     

(Abstracts)Changing Industrial Demands

Abstract

The frequency dependence of discomfort caused by vertical mechanical shocks has been investigated with 20 seated males exposed to upward and downward shocks at 13 fundamental frequencies (1–16 Hz) and 18 magnitudes (±0.12 to ±8.3 ms^?2). The rate of growth of discomfort with increasing shock magnitude depended on the fundamental frequency of the shocks, so the frequency dependence of equivalent comfort contours (for both vertical acceleration and vertical force measured at the seat) varied with shock magnitude. The rate of growth of discomfort was similar for acceleration and force, upward and downward shocks, and lower and higher magnitude shocks. The frequency dependence of discomfort from shocks differs from that of sinusoidal vibrations having the same fundamental frequencies. This arises in part from the frequency content of the shock. Frequency weighting W_b in BS 6841:1987 and ISO 2631-1:1997 provided reasonable estimates of the discomfort caused by the shocks investigated in this study.

Practitioner Summary: No single frequency weighting can accurately predict the discomfort caused by mechanical shocks over wide ranges of shock magnitude, but vibration dose values with frequency weighting W_b provide reasonable estimates of discomfort caused by shocks similar to those investigated in this study with peak accelerations well below 1 g.     

Comparisons of apparent mass responses of human subjects seated on rigid and elastic seats under vertical vibration

The apparent mass (AM) responses of human body seated on elastic seat, without and with a vertical back support, are measured using a seat pressure sensing mat under three levels of vertical vibration (0.25, 0.50 and 0.75 m/s² rms acceleration) in 0.50–20 Hz frequency range. The responses were also measured with a rigid seat using the pressure mat and a force plate in order to examine the validity of the pressure mat. The pressure mat resulted in considerably lower AM magnitudes compared to the force plate. A correction function was proposed and applied, which resulted in comparable AM from both measurement systems for the rigid seat. The correction function was subsequently applied to derive AM of subjects seated on elastic seat. The responses revealed lower peak magnitude and corresponding frequency compared to those measured with rigid seat, irrespective of back support and excitation considered.     

Car driving with and without a movable back support: Effect on transmission of vibration through the trunk and on its consequences for muscle activation and spinal shrinkage

The aim of this study was to test the effect of a movable backrest on vibration transmission through the trunk during driving and on the physiological consequences thereof. Eleven healthy male subjects drove for about 1 h on normal roads with a movable and with a fixed backrest while surface electromyography (EMG) was measured at the level of the fifth lumbar vertebra (L5) and vertical accelerations were measured at the seat, backrest and at the spine at the levels of the second sacral vertebra (S2) and seventh cervical vertebra (C7). The movable backrest significantly reduced accelerations at C7 by up to 11.9% at the 5 Hz frequency band. The movable backrest also significantly reduced the coherence and transmission between S2 and C7 accelerations, but not the differential motion between these sensors. EMG at both sides of L5 was on average 28% lower when using the movable backrest. Spinal shrinkage was unaffected by backrest type. It is concluded that a movable backrest reduces the transmission of vibration through the trunk and that it reduces low back EMG. Car driving is associated with the risk of developing low back pain and this may be related to exposure to whole body vibration. This study found an effect of a simple ergonomics measure on the transmission of vibration through the trunk as well as on back muscle activation.     

The effects of vibration frequency and direction on the location of areas of discomfort caused by whole-body vibration

Although much research has been devoted to the determination of equivalent comfort contours for human response to whole-body vibration little consideration has been given to the source of the feelings that give rise to such comfort contours. This paper shows that for vertical vibration there is a distinct difference in the locations of discomfort on the body at different frequencies and that the locations are not much affected by the vibration level. For horizontal motions, feelings of discomfort predominated in the lower abdomen and buttocks irrespective of vibration frequency or direction. A semantic scaling technique indicates the maximum sensitivity to vertical vibration acceleration in the 4 to 16 Hz range, but for both fore-and aft and lateral vibration there is a decrease in sensitivity with increasing frequency above 2Hz.     

Finite element modelling of human-seat interactions: vertical in-line and fore-and-aft cross-axis apparent mass when sitting on a rigid seat without backrest and exposed to vertical vibration

Biodynamic models representing distributed human-seat interactions can assist seat design. This study sought to develop a finite element (FE) model representing the soft tissues of the body supported by seating and the vertical in-line apparent mass and the fore-and-aft cross-axis apparent mass of the seated human body during vertical vibration excitation. The model was developed with rigid parts representing the torso segments, skeletal structures (pelvis and femurs) and deformable parts representing the soft tissues of the buttocks and the thighs. The model had three vibration modes at frequencies less than 15 Hz and provided reasonable vertical in-line apparent mass and fore-and-aft cross-axis apparent mass. The model can be developed to represent dynamic interactions between the body and a seat over a seat surface (e.g. dynamic pressure distributions and variations in seat transmissibility over the seat surface).