Transcript
WBV v1.0 Technical Specifications Introduction Long term exposure to high amplitude whole-body vibration is associated with the subsequent development of back pain1-3. ISO2631.1 provides guidance regarding the measurement of wholebody vibration and interpretation of the results4. However, the cost and complexity of commercially available devices for measuring whole-body vibration is a barrier to the collection of the systematic data required for the management of occupational whole-body vibration exposures in dynamic workplaces. The aim of the WBV application is to allow the use of an iPod Touch to estimate whole-body vibration amplitude as part of a whole-body vibration management program.
Accelerometer specifications The 5th generation iPod Touch (Apple Inc., Cupertino, CA) (123 x 59 x 6 mm, 88g) incorporates a factory calibrated LIS331DLH (MEMS type) accelerometer (STMicroelectronics, Geneva, Switzerland) providing three dimensional 16 bit data output configured to a range of +/- 2g. The maximum sampling rate which may be obtained within the iOS operating system is restricted to a nominal 100 Hz (inter-sample interval of 0.01s).
Sampling rate & variability 50
Selecting a 100 Hz sampling rate results in a bi-modal distribution of inter-sample intervals.
Proportion (%)
40
30
20
10
0 0
0.002
0.004
0.006
0.008
0.010
0.012
0.014
0.016
0.018
0.020
0.022
0.024
Inter-sample interval (s)
A nominal 99 Hz sampling rate results in an actual sampling rate of the order of 89 Hz. Typical inter-sample interval variability is illustrated below. Sampling rate limitations are likely to be the principal source of any inaccuracies in the estimation of whole body vibration in comparison to gold standard devices which sample at rates of the order of 8 kHz.
Robin Burgess-‐Limerick, 2014 WBV_tech_spec_v1.1 ergonomics.uq.edu.au/WBV
WBV v1.0 Technical Specifications Calibration data An example of Z direction accelerometer data gathered from an iPod Touch placed on the SV111 calibrator is provided below. These data provide an approximation of the 1 m/s2 r.m.s. vibration at 15.9 Hz provided by the calibrator within the limitations of the sampling rate and noise inherent to the device. The average amplitude (r.m.s.) of the acceleration data collected by five iPod touch devices was 0.962 m/s2, an average constant error of -0.038 m/s2.
Acceleration (g)
−0.9
−1.0
−1.1
0.2
0.4
0.6
0.8
1.0
1.2
The spectral density of data obtained from an iPod Touch placed on the calibrator at 15.92Hz is also illustrated. While the dominant frequency is faithfully reproduced, there is evidence of aliasing (small peaks at 3Hz and 29Hz) that arises as a consequence of the relatively low sampling rate.
1.4 1.6 Time (s)
1.8
2.0
2.2
2.4
2.6
2.8
3.0
3.0
0.5 Hz resolution
2.5
Spectral density ([m/s/s]2/Hz)
0
2.0 1.5 1.0 0.5 0 0
5
10
15
20 25 Frequency (Hz)
30
35
40
Frequency weighting ISO2631.1 provides frequency weightings Wd and Wk to be applied to horizontal (X & Y) and vertical (Z) accelerations respectively. WBV applies these frequency weightings to the raw accelerometer data, adapting Matlab code provided by Irvine5. ISO2631.1 Frequency weightings 1.0
0.8
Weighting
Wd (X&Y)
Wk (Z)
0.6
0.4
0.2
Robin Burgess-‐Limerick, 2014 WBV_tech_spec_v1.1 ergonomics.uq.edu.au/WBV 0.0 0.125
0.25
0.5
1
2
4
Frequency
8
16
32
64
where
aw(t) is the instantaneous frequenc
WBV v1.0 Technical Specifications
z
is the integration time for run
theraw time t traceis is The consequences of this frequency weighting are illustrated below. The lower Z (integration variab is the Hz) timewhile of observation (in direction accelerometer data collected at a nominal 99 Hz sample rate (actualt0 rate 89.5 driving a light vehicle. The results of applying the Wk frequency weighting are illustrated in the This formula definingofa the linear integration upper trace. The weighting acts as a high pass filter in that very low frequency components accelerations are removed. + 7 [a,(r)]* ---oo
exp[y]
&
The d ifference in result is very smal I fo larger (up to 30 %) when a pplied to s ho
The maximum transient vibration value, MlW = max [a,&)]
i.e. the highest magnitude of a,,&) read
IS0 2631=1:1997(E)
Accessed by UNIVERSITY OF QUEENSLAND on 12 Jan 2009
It is recommended to use z= 1 s in m sound level meters).
The weighted r.m.s. acceleration radians per second squared (rad/s 6.3.2 The fourth power vibration dos accordance with the following equ
The fourth power vibration dose metho fourth power instead of the1 second po power vibration dose value 1(VDV) in me ISO2631.1 defines two primary measures of whole body vibration amplitude. power I,75 (rad/s 1,759,is defined as: 2
RMS & VDV calculation
The root mean squared amplitude (aw, units m/s ) is defined by:
The Vibration Dose Value (VDV, units m/s1.75) is a fourth power measure which is consequently more sensitive to high amplitude values (jolts/jars). where VDV= VDV is defined as: a,,&) is the weighted acceler The VDV measure is cumulative and increases with the duration of the second squared (m/s*) measurement. To allow comparison between trials of different duration the VDV is expressed as T is the duration of the m VDV(n) where n refers to an n hour exposure, eg., VDV(8) normalises the trial to the instantaneous frequen a, an (t9 8ishour duration. VDV(n) = VDV x (n / trial duration in hours)1/4 is the duration curves of measurem Frequency-weighting recom tables 1 and 2 and discussed in Accuracy in comparison with Gold Standards weighting curves are given in table The accuracy of the WBV application used in conjunction with an iPod Touch in comparison to gold standard devices is the subject of ongoing investigation. However, preliminary indications are promising6. Two investigations have been completed in which raw data were gathered from iPod Touch devices simultaneously with measurements by gold standard devices and the iPod 6.2 Applicability of the basic data were subsequently analysed via Matlab. Light vehicle accuracy data
6.2.1
Definition
of crest factor
For the purposes of this part of IS Three dimensional accelerometer data were gathered simultaneously from 5th generation iPod instantaneous peak value of the fr touch devices placed under the seat pad accelerometer of a gold standard vibration determined over the measurement device placed on the driver’s seat while four different light vehicles were driven in aduration of m value (see 6.1). range of environments (highway, suburban streets, gravel roads & off road). Two gold standard measurement devices were employed, SV106 (Svantek Sp., Warsaw, Poland) and a Type 4447 The crest factor does not n NOTE Human Vibration Analyzer (Brüel & Kjær Sound & Vibration Measurement A/S, Nærum, Denmark). Forty-two trials ranging in duration from 10 to 55 minutes (median = 15 minutes) were recorded.
6.2.2
Applicability
of the basic e
Robin Burgess-‐Limerick, 2014 WBV_tech_spec_v1.1 ergonomics.uq.edu.au/WBV
The crest factor may be used to in the vibration in relation to its effe basic evaluation method is normall
WBV v1.0 Technical Specifications Outcome measures derived from frequency weighted accelerometer data for vertical direction gathered simultaneously from an ipod touch and gold standard vibration measurement devices are illustrated below. The mean absolute error across the 42 trials was 0.019 m/s2 for the vertical direction and slightly higher for fore-aft and side-to-side accelerations (0.02 m/s2 & 0.015 m/s2 respectively).
Heavy vehicle accuracy data
Whole-body vibration amplitudes (r.m.s) are illustrated for X, Y & Z directions derived from frequency weighted iPod Touch accelerometer data as a function of simultaneous measurements taken via the gold standard device. The Bland-Altman plot does not indicate any consistent relationship between constant error and acceleration amplitude.
X Y Z
iPod r.m.s (m/s2)
Fifty-eight pairs of measurements of measurements were obtained. Twenty-six pairs of measurements were simultaneously obtained via Larson Davis Human Vibration Meter 100 (PCB Piezotronics, Inc, Depew, New York, USA) and a 5th generation iPod Touch during operation of a range of heavy equipment (Dozers, Graders, Excavators, Loaders, Haul trucks) at two surface coal mines. A further thirty-two pairs of measurements were obtained from SV106 (Svantek Sp., Warsaw, Poland) and a 5th generation iPod Touch from a range of heavy equipment (Haul trucks, Excavator, Loaders) in operation at a third surface coal mine. Measurement duration ranged from 12 minutes to 54 minutes (median = 29 minutes).
1.5
1.0
0.5
0.0 0.0
0.5
Gold Standard r.m.s
1.0
1.5
(m/s2)
0.2
X Y Z Constant error r.m.s (m/s2)
Three dimensional accelerometer data were gathered simultaneously from 5th generation iPod touch devices placed under the seat pad accelerometer of a gold standard vibration measurement device placed on the driver’s seat of heavy mining equipment whilst operators performed their normal duties.
0.1
0.0
0.5
1.0
-0.1
-0.2
Robin Burgess-‐Limerick, 2014 WBV_tech_spec_v1.1 ergonomics.uq.edu.au/WBV
1.5
UEENSLAND on 12 Jan 2009
WBV v1.0 Technical Specifications The table below provides the mean constant error (CE) for the fifty-eight vibration measurements made in each direction, the standard deviation (SD) of these measurements, and the resulting 95% confidence limits of agreement. The results suggest that accelerometer data gathered from an iPod Touch are able to be used to measure whole-body vibration amplitude with 95% confidence of +/- 0.09 m/s2 r.m.s. or better, depending on the direction of interest. The accuracy was better in the vertical direction (usually the dominant vibration direction) and the limits of agreement were calculated to be +/- 0.063 m/s2 r.m.s.
IS0 2631=1:1997(E) Mean constant error
X r.m.s. (m/s 222)
Y r.m.s. (m/s2)
Z r.m.s. (m/s2)
0.033
0.021
0.005
@ IS0
Standard deviation of constant error 0.038 0.046 0.032 For exposures below the zone, health effects have not been clearly documented and/or objectively observed; in the Lower 95%respect limit of agreement -0.107 -0.111 -0.068 risks are likely. This above the zone health is indicated and to potential health risks with zone, caution the shading in figure B.I. by indicated as h 8 to h 4 of range the in exposures on based mainly is recommendation Upper 95% limit of agreement 0.042 0.068 0.058 Shorter durations should be treated with extreme caution.
Other studies indicate a time dependence according to the following relationship:
Interpretation of results with respect to HGCZ
awl
.T1/4 .T'/4 1 -a,2 ISO2631.1 provides
. I .
b3.2)
the figure below in Appendix B. It is suggested that “health effects are likely” for 8 hour exposures greater than 0.9 m/s2 r.m.s., and that “no health effects caution zones for health in figure dotted linesbelow byexposures zone is indicated guidance This health 2 r.m.s. have been caution documented” for 8 hour 0.5B.1. m/s(The Theguidance standard equations (B.1) and (B.2) are the same for durations from 4 h to 8 h for which most occupational observations suggests that “caution with respect to potential health risks” is indicated for intermediate exist .) values - hence the so called Health Guidance Caution Zone (HGCZ). 2
10
iFi G= g
!
6,3
G ki m w 4 aJ 2 5
! I I
I
\ . .
\
----LLy---
LqUdl -0..
._
2,s
‘-..
\_
\
-._
lull
\D.
II
1 I
\
I
--._
I I
1.6 1
0.9 m/s2
0,63
0.5 m/s2 0,315 025
0,16
10 min
0,s
1
2
4
8
24
Exposure duration, h
zones caution the guidance B.l - in Health No explicit guidance isFigure provided ISO 2631.1 regarding evaluation of VDV, although it has been generally inferred that the values referred to for eVDV in note 2 of clause B.3.1 figure B.l at the shown the17zone with of acceleration the frequency-weighted value The r.m.s.may beofutilised, that is a lower value of 8.5 can andbea compared higher value for the VDVinHGCZ. duration of the expected daily exposure.
To characterize daily occupational vibration exposure, the 8 h frequency-weighted acceleration a, can be measured h as the time period 8 BV_tech_spec_v1.1 to the 2formula according or calculated Robin Burgess-‐Limerick, 014 in 6.1 with W T. ergonomics.uq.edu.au/WBV NOTES 1 When the vibration exposure consists of two or more periods of exposure to different magnitudes and durations, the
WBV v1.0 Technical Specifications As well as calculating and reporting the RMS and VDV amplitude measures describing each trial, the RMS and VDV(8) values are presented graphically with respect to the HGCZ, by default for an 8 hour exposure. The application also calculates the the exposure duration required to reach the lower limit of the RMS HGCZ for each direction. The WBV application allows other exposure durations to be nominated (1, 2, 4, 8, 10 or 12 hours). When a new duration (n) is selected, the VDV(n) is recalculated. The lower and upper bounds of the RMS HGCZ are also adjusted according to ISO2631.1 Appendix B, equation B1 which results in the boundaries listed below. Exposure duration 1 hr 2 hr 4 hr 8 hr 10 hr 12 hr
HGCZ lower bound 1.4 m/s2 1 m/s2 0.71 m/s2 0.5 m/s2 0.45 m/s2 0.41 m/s2
HGCZ upper bound 2.5 m/s2 1.8 m/s2 1.2 m/s2 0.9 m/s2 0.8 m/s2 0.73 m/s2
A note about “k” ISO2631.1 refers to “multiplying factors” or “k”, which require the horizontal acceleration values to be increased by 40% (kx & ky = 1.4) when calculating the vector sum (VTV, see clause 7.2.2). However, ambiguity exists in that the multiplying factors are not referred to, nor included in the equations, in Annex B which provides additional guidance regarding the health effects of whole-body vibration. Indeed the only equation in ISO 2631.1 in which the “multiplying factors” appear explicitly in relation to health effects is in the definition of VTV, and “k” is not defined in clause 4 “Symbols and subscripts”. Given this ambiguity, and the lack of any rationale for the use of such multiplying factors, all acceleration values provided by WBV do not include any additional weighting for X or Y directions.
Robin Burgess-‐Limerick, 2014 WBV_tech_spec_v1.1 ergonomics.uq.edu.au/WBV
WBV v1.0 Technical Specifications REFERENCES 1. Bernard, B.P. (Ed): Musculoskeletal disorders and workplace factors: a critical review of epidemiologic evidence for work-related disorders of the neck, upper extremities, and low back. DHHS (NIOSH) Publication No. 97-141. US Department of Health and Human Services, National Institute of Occupational Safety and Health, 1997. 2. Bovenzi, M. and C.T.J. Hulshof: An updated review of epidemiologic studies on the relationship between exposure to whole- body vibration and low back pain. Journal of Sound and vibration, 215: 595–611, 1998. 3. Sandover, J.: Dynamic loading as a potential source of low back disorder. Spine, 8, 652-658, 1983. 4. Burgess-Limerick, R: How on earth moving equipment can ISO2631 be used to evaluate WBV exposure? Journal of Safety and Health Research and Practice. 4(2): 13-21, 2012. 5. Irvine, T. : ISO2631 matlab scripts. http://vibrationdata.wordpress.com/2012/10/21/iso-2631-matlabscripts/, 2012. (Retrieved 24/7/13). 6. Wolfgang, R. & Burgess-Limerick, R. Using consumer electronic devices to estimate whole-body vibration exposure. Journal of Occupational and Environmental Hygiene. (in press)
Robin Burgess-‐Limerick, 2014 WBV_tech_spec_v1.1 ergonomics.uq.edu.au/WBV
