Longitudinal Assessment of Vibrations During Manual and Electric Powered Wheelchair
Driving Over Selected Sidewalk Surfaces

Departments of Rehabilitation Science & Technology1, and Bioengineering2
University of Pittsburgh, Pittsburgh, PA 15261

HUMAN ENGINEERING RESEARCH LABORATORIES3
VA Rehabilitation Research and Development Center
VA Pittsburgh Healthcare System
Pittsburgh, PA, 15206

ADDRESS CORRESPONDENCE TO:

Rory A. Cooper, Ph.D.
Human Engineering Research Laboratories (151R-1H)
VA Pittsburgh Healthcare System
7180 Highland Drive
Pittsburgh, PA 15206

TEL: (412) 365-4850
FAX: (412) 365-4858
e-mail: rcooper+@pitt.edu

Acknowledgements

This study was partially funded by a consortium of the Interlocking Concrete Pavement Institute (ICPI), Brick Industry Association (BIA) and the National Concrete Masonry Association (NCMA). In addition, funding was provided by the VA Rehabilitation Research and Development Service, Veterans Health Administration, U.S. Department of Veterans Affairs (F2181C), and the U.S. Department of Education, National Institute on Disability and Rehabilitation Research (NIDRR) Rehabilitation Engineering Research Center on Wheeled Mobility (H133E990001), and a National Science Foundation Graduate Research Fellowship.

ABSTRACT

Wheelchair users rely on their wheelchairs for mobility for extended periods of time every day. According to the International Standards Organization 2631-1 standard on human vibration, individuals in a seated position when exposed to whole-body vibration, are at risk of injury. The goals of this research were to evaluate vibration exposure during powered and manual wheelchair driving over nine sidewalk surfaces and differences in vibration exposure over three years. Ten unimpaired subjects were asked to drive a manual at 1 m/s and electric powered wheelchair at 1 m/s and 2 m/s over nine sidewalk surfaces, while whole-body vibrations were measured at the seat and the footrest of the wheelchair. At 1 m/s significant differences existed between the surfaces and between years at both the seat and the footrest for the manual and power wheelchair. At 2 m/s there were significant differences between both surfaces and years at the seat and the footrest. Based on the results of this study, both manual and power wheelchair users may be at risk for injuries due to whole-body vibration while traveling over certain surfaces.

Keywords: Accessible Surface, Injury, Sidewalks, Vibration, Wheelchairs

INTRODUCTION

Wheelchair users, both manual and powered, use their wheelchairs for mobility for extended periods of time each day [1]. This extensive use, combined with the bumps, uneven driving surfaces, and other obstacles, can expose wheelchair users to harmful whole-body vibrations (WBVs) which can lead to secondary injuries such as low-back and neck pain, muscle ache and fatigue, and other harmful effects [2]. Few studies have reported the levels of vibration that are experienced by manual wheelchair users, and even fewer have reported on powered wheelchairs or their users. This study will examine different sidewalk surfaces and the resulting whole-body vibrations that are transmitted during manual and power wheelchair driving. Differences in vibration exposure over years will also be examined.

A review of the literature revealed that little research has been conducted on exposure to whole-body vibration over various surfaces during wheeled mobility, including bicycle riding, in-line skating, baby stroller use, and scooters. Thompson et al [3] looked at vibration during in-line skating; however it was only over a standard paved road and did not examine any other surfaces. A study done in Italy by Frendo et al [4] examined vibration during motorscooter driving over different street surfaces including heavy paved brick, light paved brick, and cobblestone. Results revealed differences between the surfaces, with the light paved brick resulting in the lowest transmitted vibration and the cobblestone surface producing the highest.

Multiple studies have shown the negative effects associated with the exposure to whole-body vibrations on humans in the seated positions [5-8]. Occupations where whole-body vibrations are a concern include heavy machinery operation, bus and truck driving, and helicopter piloting. These industries, as well as the automotive industry, have taken measures to reduce the amounts of whole-body vibrations transmitted to their users while operating these vehicles [9-11].

This problem has been recognized by the wheelchair community and efforts have been made to quantify the amounts of vibrations transmitted to wheelchair users during propulsion. VanSickle et al [12] showed that manual wheelchair users traveling over a simulated road course experience levels of vibrations that exceed the “fatigue-decreased performance boundary” and that may cause fatigue and injury. Wolf et al [13] evaluated vibration exposure to wheelchair users while traveling over sidewalk surfaces. They showed differences between interlocking concrete pavement (ICPI) surfaces and a standard poured concrete surface. In some cases, the ICPI surfaces caused lower whole-body vibrations than the standard poured concrete surface. Maeda et al [14] issued questionnaires to 33 wheelchair users and tested 10 wheelchair users on a vibration platform. Results from the questionnaire revealed that wheelchair users did feel whole-body vibrations at the neck, back, and buttocks during propulsion, and that users sensed differences while traveling over different surfaces and obstacles.

Wheelchair companies have attempted to address this problem by adding suspension to manual and powered wheelchairs, however studies have demonstrated that these additions do not necessarily reduce the amounts of oscillatory and shock whole-body vibrations. Additionally, in the case of manual wheelchairs, titanium rigid framed wheelchairs without suspension performed better than some wheelchairs with suspension [15-17].

The ISO 2631-1 – Evaluation of Human Exposure to Whole-Body Vibration was established to define the methods of collection, the effects, and the health concerns associated with whole-body vibration [18]. The standard defines a health guidance caution zone (Figure 1) which characterizes the amount of whole-body vibration that is considered unsafe. When evaluating exposure of whole-body vibrations over long periods of time, lower cumulative levels of whole-body vibration are considered harmful (less than 1 m/s2 of weighted Root Mean Squared (RMS) acceleration at eight hours of exposure). The harmful effects of whole-body vibrations can be negated by an eight hour rest period; however this is extremely rare during an ordinary day of a manual or powered wheelchair user, and through days, months and years the cumulative exposure to whole-body vibrations can result in secondary injuries.

Figure 1 - Limit boundaries of vibration exposure as defined by the ISO-2631 Standard

The goal of this research was to evaluate the whole-body vibrations experienced while driving a manual and electric powered wheelchair over selected sidewalk surfaces, and differences in exposure over years. We hypothesized that different surfaces would induce significantly different whole body vibrations, suggesting that some are less likely to cause secondary injuries to wheelchair users. Furthermore, we hypothesized that over time, due to weather-related wear, the surfaces would become smoother and thus induce significantly lower vibrations.

METHODS

Six different sidewalk surfaces were tested in three consecutive years (May 2002, July 2003, and June 2004). All of the sidewalk surfaces were approximately 1.2 meters wide and 7.6 meters long. Surface 1 was a poured concrete sidewalk with a brush finish which acted as the control surface. Surfaces 2, 3, and 4 were made from interlocking concrete pavement installed to industry specifications [19], and were installed with a 90-degree herringbone pattern. The interlocking concrete blocks used to construct Surface 2 had no bevel, blocks used for Surface 3 had 2 mm beveled edges, and blocks for Surface 4 had 8 mm beveled edges. Sidewalk surfaces 5 and 6 were constructed of fired clay bricks, and were constructed using a 45-degree herringbone pattern. Blocks used for Surface 5 had 4 mm beveled edges and blocks used for Surface 6 had no bevel. In year three, three additional concrete surfaces were added. Surfaces 7 and 8 both had a 6mm bevel and were installed using a 90 degree and 45 degree herringbone pattern respectively. Surface 9 had a 4 mm bevel and was installed using a 90 degree herringbone pattern.

Figure 2 – The surfaces that were tested

Surfaces 1 (poured concrete), 2 (concrete, no bevel), and 3 (concrete, 2mm bevel)

Surfaces 4 (concrete, 8mm bevel), 5 (brick, 4 mm bevel), and 6 (brick, no bevel)

Surfaces 7 (concrete, 6mm bevel), 8 (concrete, 6mm bevel), and 9 (concrete, 4mm bevel)

The specifications of the surfaces can be seen in Table 1. An Interlocking Concrete Pavement Institute (ICPI) certified contractor installed all of the sidewalks.

Table 1 - Specifications of Surfaces Tested

Ten able-bodied subjects were recruited in each of the three testing years. Efforts were made to recruit either the same subject each year or subjects with matching weights and heights in order to make sure variability in subject population was accounted for. Subject demographics, averages and standard deviations of heights and weights, can be seen in Table 2. A repeated measures ANOVA showed that there were no significant differences between subjects heights (p=0.7548) and weights (p=0.3962) over years.

Table 2 - Subject Means and Standard Deviations Across Years

This study was approved by the Veterans Affairs Pittsburgh Healthcare Systems Institutional Review Board. Study requirements stated that subjects be between the ages of 18-65, free of any shoulder pain that would prevent them from propelling, no history of cardiopulmonary disease, and free of a physical disability. After giving their written informed consent, subjects were asked to propel a manual wheelchair (at 1 m/s) (Figure 2) and drive an electric powered wheelchair (at 1 m/s and 2 m/s) over six sidewalk surfaces a total of three times each.

Figure 2 - Setup of the Quickie GP manual wheelchair

The manual wheelchair (Quickie GP, Sunrise Medical Ltd.) was a rigid frame design with 127 mm diameter polyurethane tires, and standard 610 mm diameter rear wheels. The seat width was 406 mm, the seat depth was 458 mm, and the backrest height was 410 mm. The rear axles were placed 45 mm in front of the backrest tubes. The SMARTWheels were used as the rear wheels during this study [20]. They were used in the first year of testing to evaluate that there were no differences in work during propulsion over all of the surfaces [13], and therefore used each of the following years for congruency. SMARTWheels use solid foam inserts. The approximate mass of the manual wheelchair was 15.5 kg with the SMARTWheels attached.

The electric powered wheelchair (Quickie P200, Sunrise Medical Ltd.) had a rigid frame with 203 mm front casters, and 254 mm diameter rear wheels (Figure 3).

Figure 3 - Setup of the Quickie P200 electric powered wheelchair

The seat width was 406 mm, the seat depth was 415 mm, and the backrest height was 435 mm. A standard position-sensing joystick was mounted to the right side armrest, and the manufacturer default controller settings were used. All tires were properly inflated to the rated air pressure (248.2 kPa for the caster, and 344.7 kPa for the rear wheels). The approximate mass of the electric powered wheelchair with batteries was 89 kg. The frame of the electric powered wheelchair was made from aircraft quality aluminum. All subjects sat on a 50 mm thick linear polyurethane foam cushion during all testing. Both wheelchairs were not used in between years of testing to ensure minimal deterioration (specifically to the tires, and frames) or other changes over time.

A tri-axial accelerometer was used to collect vibrations in three orthogonal axes at the seat and the footrest. Acceleration data were collected at 200 Hz. The ISO Standards describe the minimum collection rate for accelerations as 160 Hz. The seat accelerometer was attached to a 40.64 cm x 40.64 cm x .64 cm aluminum plate. The footrest accelerometer was attached to a 7.62 cm x 15.24 cm x .95 cm aluminum plate which in turn was attached to the wheelchair footplates. Based upon the ISO 2631-1 Standard [18], the whole-body vibrations defined along the z-axis (vertical, along the spine of a seated subject and along the legs transmitted through the footrest) were analyzed using the Root Mean Square method (Equation 1). The choice to only measure the z-axis acceleration direction was based on the ISO 2631-1 standard which states that the results of the measurements should be made on the direction which presents the highest vibrations. Once acceleration data were collected at the seat and the footrest for each trial, frequency weightings, as described by the ISO 2631-1 standard were applied. The frequency weighted accelerations in the vertical direction are given as awz and the time of the trial is T. The result is the root mean squared acceleration in the vertical direction (RMSz).

[1]

The acceleration data were calibrated and converted for analyses in custom software written using Matlab (The Mathworks Inc., Natick, MA).

Statistical analyses were done using SAS (SAS Institute, Inc., Cary, NC). Data were analyzed using a mixed model to evaluate the differences in RMS vertical vibrations at the seat and the footrests between surfaces and between years. Analyses between years only included surfaces 1 through 6 (poured concrete, concrete of 0 mm, 2 mm, and 8 mm bevels, and brick of 4 mm and 0 mm bevels, respectively) because surfaces 7, 8 and, 9 (concrete of 6 mm, 6 mm (45o pattern) and 4 mm bevels, respectively) were only tested in the third year. Post-hoc analysis was completed using a Tukey pairwise comparison.

RESULTS

Data were analyzed for normality. Outliers were removed and data were found to be normally distributed.

Table 3 - Average Seat Root Mean Square acceleration. Surfaces significantly lower (p=0.05) than surface 1 are denoted by *. Surfaces significantly higher (p=0.05) than surface 1 are denoted by #.

Table 4 - Average Footrest Root Mean Square acceleration. Surfaces significantly lower (p=0.05) than surface 1 are denoted by *. Surfaces significantly higher (p=0.05) than surface 1 are denoted by #.

Table 5 - Average Seat Root Mean Square vibrations over three years. Years not significantly different are denoted by *.

Table 6 - Average Footrest Root Mean Square vibrations over three years. Years not significantly different are denoted by *.

1. Manual Wheelchair

A. Surfaces Vibrations at the seat and the footrest for the manual wheelchair there were significant differences between the surfaces (p<.0001). Post-hoc analysis revealed that the standard poured concrete surface was significantly higher than surfaces 2 and 3 (concrete of 0 mm and 2 mm bevels), significantly lower than 4, 7, and 8 (concrete of 8 mm, 6mm and 6 mm (45o pattern) bevels), and not significantly different from surfaces 5 and 6 (brick of 4 mm and 0 mm (45o pattern)).

B. Years For RMS vibrations at the seat and the footrest there were significant differences between the three years (p < 0.0001). Post-hoc analysis revealed that for RMS vibrations at the seat, year 1 was significantly lower than years 2 and 3. Year 2 was not significantly different (p = 0.3257) than year 3. For RMS vibrations at the footrest all three years were significantly different (p < 0.0001).

2. Powered Wheelchair

A. Surfaces At 1 m/s, for vibrations at the seat and the footrest significant differences were found between surfaces (p<.0001). Post-hoc analysis of the 1 m/s speed revealed that at the seat surface 2 (concrete of 0mm bevel) was significantly lower than surface 1. Surfaces 3, 5, 6, 8, and 9 (concrete with 2 mm bevel, brick with 4 mm bevel, brick with 0 mm bevel (45o pattern), concrete with 6 mm bevel, and concrete of 4 mm bevel) were not significantly different than the standard poured concrete surface. Surface 4 (8 mm bevel) and surface 7 (6 mm bevel) were significantly higher than surface 1 (poured concrete). At the footrest, surfaces 2, 3, and, 9 were significantly lower than surface 1, surfaces 5, 6, 7, and, 8 were not significantly different, and surface 4 was significantly higher.

At 2 m/s, for vibrations at the seat and the footrest significant differences were found between surfaces (p<.0001). Post-hoc analysis of the 2 m/s speed revealed that at the seat, all surfaces were significantly lower than the standard poured concrete surface. At the footrest, all surfaces were significantly lower than the standard poured concrete surface except Surface 4, which was not significantly different.

B. Years For RMS vibrations at the seat and the footrest there were significant differences in years at 1 m/s (p=0.0008 and p=0.0005) and at 2 m/s (p<0.0001 and p<.0001). Post-hoc analysis revealed that year 1 was not significantly different from year 3 (p=0.1756) at the footrest at 1 m/s.

DISCUSSION

Based on its nature, a wheelchair represents a system where its user will be subjected to whole-body vibrations in a seated position for long durations. Because the ISO 2631-1 standard requires an eight hour rest period to negate the damaging effects of any transmitted vibration [18], and because powered wheelchair users typically rely on their wheelchair for all of their mobility, they characterize a population that is potentially at high risk of secondary injuries due to WBVs.

Wheelchair users experience pain over the course of the day for multiple reasons. Sitting for long periods of time, and exposure to whole-body vibration can cause discomfort. Maeda et al [14] issued a questionnaire to wheelchair users to determine if vibrations experienced during wheelchair propulsion caused discomfort. Results showed that the vibration experienced during propulsion caused discomfort, specifically at the neck, lower back, and buttocks. This research, as well as others measuring levels of vibration during wheelchair use [12, 15-17] shows that whole-body vibration exposure may cause discomfort and eventually pain in wheelchair users.

The results of this study showed that surfaces other than poured concrete should be considered for pedestrian access routes. Interlocking concrete and brick surfaces that have small bevels may decrease the amount of whole-body vibrations that are transmitted to wheelchair users during propulsion, especially at higher speeds.

The results showing differences in the surfaces were expected based on previous studies [13] as well as the physical properties of the surfaces. Surface 4 has the largest bevel (8 mm) and would expectantly cause the most vibration. Surface 2 has the smallest bevel (0 mm) and resulted in the lowest RMS vibration.

1. Manual Wheelchair

Results for the manual wheelchair showed that surfaces 2 and 3 (concrete with 0 mm bevel and 2 mm bevel) produce significantly lower RMS vibrations than the poured concrete surface. These surfaces present a good alternative to the standard poured concrete because they transmit lower amounts of whole-body vibration to wheelchair users. This result contradicts the statement from the Rights-of-Way Advisory Board [21] that claim that surfaces comprised of individual units are undesirable due to the vibrations they cause. Results showed that there were differences in RMS vibration over surfaces between years for the manual wheelchair at both the seat and the footrest with RMS vibrations trending to increase over time.

2. Power Wheelchair

Results for the power wheelchair also showed promising results for the use of alternative surfaces to reduce the amount of whole-body vibrations transmitted to wheelchair users. At 1 m/s surface 2 was significantly lower at the seat than the standard poured concrete surface and surfaces 2, 3, and 9 were significantly lower at the footrest. At 2 m/s all surfaces were significantly lower than the standard concrete surface at the seat and only surface 4 was not significantly different at the footrest while all other surfaces were significantly lower. The results from the data collected at 2 m/s most likely are caused by the breaks in the poured concrete surface. The higher speed of the wheelchair causes larger transmission of shocks from the breaks in the sidewalk. Results also showed that there were significant differences in years at the seat and the footrest for both speeds of the power wheelchair. There appeared to be no trend for increase or decrease of RMS vibrations over years. The results from the change in the surfaces over years suggests that data has not been collected for a long enough period of time to show either an increase or decrease in vibration. Data should continue to be collected to more appropriately analyze the trend of change over time.

We hypothesized that due to weather conditions and use, one might expect that the wear on the bevels would reduce the amounts of vibrations experienced by the wheelchair. The results however show an increase in RMS vibration at the seat and the footrest for the manual wheelchair and the power wheelchair at 2 m/s. There are multiple reasons that this result may have occurred. The nine tested surfaces are isolated and do not see normal wear due to travel and use, which would result in similar results over multiple years. Heaving and settling of the interlocking concrete and brick pavers may occur over time and would cause sharper transitions resulting in higher RMS vibrations. Finally, there may be no significant difference over time if results are followed for longer periods..Limitations of this study include no use of a standard non-vibration surface such as smooth tile to compare surfaces to a baseline control. However the standard poured concrete was used as the control surface because it is the most common outdoor pedestrian surface. There was no additional pedestrian wear on any of the surfaces. In a real world situation there would be wear on surfaces from normal use. However keeping the surfaces in a controlled environment allows the comparison to be more accurate since the wear on all surfaces is equal. Only one manual and power wheelchair was used for this study. It is understood that certain wheelchairs are capable of reducing the amounts of whole-body vibration transmitted to wheelchair users, however since differences in surfaces are being examined, rigid frame manual and power wheelchairs were selected for this study.

The use of able-bodied subjects as well as using different subject over the years is a further limitation to this study. There could be differences in propulsion style during the manual wheelchair driving. Although there are differences in the sitting biomechanics of wheelchair users and able-bodied subjects, this subject population was used for ease of recruitment over multiple years and because the metric of interest was the differences between a standard poured concrete surface and interlocking concrete or brick surfaces. Additionally, matching the height and weight of the replacement subjects is assumed to control for important factors of whole-body vibration.

These results do demonstrate that some interlocking pavement surfaces should be considered for wheelchair access routes and may reduce the amount of WBVs that are transmitted to wheelchair users, specifically the surfaces with the smallest bevels. The results clearly show that many of the ICPI surfaces are just as good if not better than the standard poured concrete surface at reducing the amounts of WBV transmitted to wheelchair users. Additionally some surfaces may produce levels of WBV exposure that could cause secondary injuries to both manual and power wheelchair users over time.

REFERENCES

1. Cooper RA, Thorman T, Cooper R, Dvorznak MJ, Fitzgerald SG, Ammer W, Songfeng G, Boninger ML. Driving characteristics of electric powered wheelchair users: How far, fast and often do people drive? Arch Phys Med Rehabil. 2002; 83: 250-255.

2. Griffin MJ. Handbook of Human Vibrations. San Diego CA: Academic Press Inc; 1990: 173-186.

3. Thompson C, Belanger M. Effects of vibration in inline skating on the Hoffman reflex, force and proprioception. Med Sci Sports Exerc. 2002; 34 (12): 2037-2044

4. Frendo F, Sgueglia M, Vitale E, Hippoliti R. Analysis of motorscooter ride comfort. Conference of Automotive and Transportation Technology. SAE: Paris, France; 2002

5. Pope M, Wilder DG, Magnusson ML. A review of studies on seated whole body vibration and low back pain, Proc Inst Mech Eng. 1999; 213: 435-446.

6. Mansfield NJ, Griffin MJ. Effect of magnitude of vertical whole-body vibration on absorbed power for the seated human. Journal of Sound and Vibration. 1998; 215(4): 813-825.

7. Bovenzi M, Hulshof, CTJ. An updated review of epidemiologic studies on the relationship between exposure to whole-body vibration and low back pain. Journal of Sound and Vibration. 1998; 215(4): 595-611.

8. Griffin MJ, Vertical vibration of seated subjects: Effects of posture, vibration level, and frequency. Aviat Space Environ Med. 1975; 46 (3): 269-276.

9. Wan Y, Schimmels JM. Optimal seat suspension design based on minimum “simulated subjective response”. J Biomed Eng. 1997; 119: 409-416.

10. Nishiyama K, Taoda K, Kitahara T. A decade of improvement in whole-body vibration and low back pain for freight container tractor drivers. Journal of Sound and Vibration. 1998; 215(4): 635-642.

11. Wilder D, Magnusson ML, Fenwick J, Pope M. The effect of posture and seat suspension design on discomfort and back muscle fatigue during simulated truck driving. Appl Ergon. 1994; 25(2): 66-76.

12. VanSickle D, Cooper R, Boninger M, DiGiovine C. Analysis of vibrations induced during wheelchair propulsion. J Rehabil Res Dev. 2001; 38: 409-421.

13. Wolf EJ, Pearlman J, Cooper RA. Vibration exposure of individuals using wheelchairs over sidewalk surfaces. Disabil Rehabil. 2005; 27(23): 1443-1449.

14. Maeda S, Futatsuka M, Yonesaki J, Ikeda M, Relationship between questionnaire survey results of vibration complaints of wheelchair users and vibration transmissibility of manual wheelchair. Environmental Health Prevention Medicine. 2003; 8: 82-89.

15. Cooper RA, Wolf EJ, Fitzgerald SG, Boninger ML, Ulerich R, Ammer WA. Seat and footrest shocks and vibrations in manual wheelchairs with and without suspension, Arch Phys Med Rehabil. 2003; 84: 96-102.

16. Kwarciak AM, Cooper RA, Wolf EJ. Effectiveness of rear suspension in reducing shock exposure to manual wheelchair users during curb descents. Proceedings of the 25th Annual RESNA Conference. Minneapolis, MN; 2002.

17. Wolf, E.J., Cooper, R.A., and Kwarciak, A.M. Analysis of whole-body vibrations of suspension manual wheelchairs: utilization of the absorbed power method. Proceedings of the 25th Annual RESNA Conference. Minneapolis, MN; 2002.

18. ISO 2631-1 - Evaluation of Human Exposure to Whole-Body Vibration - Part 1: General Requirements. Geneva, Switzerland. International Standards Organization; 1995.

19. The problem and measures of interlocking block pavement for eyesight or walking disabled people. Interlocking Block Association of Japan; 1993: 6-24.

20. Asato K, Cooper RA, Robertson RN, Ster JF. SMARTWheels: Development and testing of a system for measuring manual wheelchair propulsion dynamics. IEEE Trans Biomed Eng. 1993; 40: 1320-1324.

21. Building a True Community: Final Report, Public Rights-of-Way Access Advisory Committee [Online]. 2001 January [cited 2006 Nov 7]; Available from: URL: http://www.access-board.gov/prowac/commrept/index.htm