- Email: [email protected]
A novel force transducer for the measurement of grip force
Journal of Biomechanics 34 (2001) 125}128
Technical note
A novel force transducer for the measurement of grip force E.K.J. Chadwick, A.C. Nicol* Bioengineering Unit, University of Strathclyde, Wolfson Centre, 106 Rottenrow, Glasgow G4 0NW, UK Accepted 22 July 2000
Abstract A new strain gauge transducer has been developed to measure functional grip forces. The gripping area is a cylinder of diameter 30 mm and length 150 mm and simulates the handle of a number of devices, allowing a range of activities to be studied. The device measures radial forces divided into six components and forces of up to 250 N per segment can be measured with an accuracy of $1%. The device therefore gives information about the magnitude and distribution of force around the cylinder during gripping, and has been shown to be a valuable research tool in a study of four di!erent types of grip, providing valuable input data for biomechanical models. 2000 Elsevier Science Ltd. All rights reserved. Keywords: Transducer; Grip; Hand; Force
1. Introduction The accurate measurement of external contact forces is essential for providing realistic loading data for biomechanical models. In the case of the upper limb, the measurement of grip forces during object manipulation is of prime importance. A number of devices for the measurement of grip strength have been reported in the literature. Richards and Palmiter-Thomas (1996) describe several devices falling into four groups: hydraulic, pneumatic, mechanical and strain gauge. Most of these are designed to measure maximal, not functional, forces, have only one degree of freedom and are lacking versatility. Strain gauged devices are an obvious choice for this type of study as they allow the recording of force variation with time and are sensitive and accurate (Richards and Palmiter-Thomas, 1996). Purves and Berme (1980) described the application of a six-channel pylon transducer to the measurement of "nger force. Three components of force and three moments were measured on one "nger. They measured the loads on the index "nger during tap turning and jar opening: forces up to 100 N and moments up to 5 N m were measured. An et al. (1980)
* Corresponding author. Tel.: #44-101-552-3028; fax: #44-141552-6098. E-mail address: [email protected] (A.C. Nicol).
described two strain gauged grip-measuring devices: one was uni-axial only, the other measured the radial force on each phalanx but only for one "nger. Pronk and Niesing (1981) described a uni-axial device with a range of 0}900 N and an accuracy of better than 5%. Radhakrishnan and Nagaravindra (1993) used strain gauges to measure the normal forces on all 12 "nger phalanges during maximal cylinder gripping. They found total grip forces of up to 250 N for a cylinder of 50 mm. Other devices measure distribution of loading record pressure, not force. Without information about the area of contact between the "ngers and the transducer it is not possible to calculate the joint loading from this. The limitations of existing devices have been outlined and a new device that attempts to address these has been developed.
2. Design of the new device A force transducer that has been specially designed to measure hand grip forces during activities of daily living. It consists of six active beams, mounted as cantilevers on a hexagonal core (see Fig. 1a) so as to form a cylinder of diameter 30 mm and length 150 mm, thus allowing the device to form the handle of a piece of equipment. The grip forces being measured are assumed to be radial. Six beams were incorporated to give the maximum amount of information about grip distribution, whilst ensuring
0021-9290/00/$ - see front matter 2000 Elsevier Science Ltd. All rights reserved. PII: S 0 0 2 1 - 9 2 9 0 ( 0 0 ) 0 0 1 6 8 - 8
126
E.K.J. Chadwick, A.C. Nicol / Journal of Biomechanics 34 (2001) 125}128
mises the sti!ness of the beams and thus minimises the end de#ection. Foil electrical resistance strain gauges (Showa N23MA-5-120-16) are applied to the top and bottom surfaces of the beams, at two cross-sections (see Fig. 1b), giving two channels of strain output for each beam so that the magnitude and position along the beam of the applied force may be calculated. At each of the two measurement sites are bonded four 120 ) gauges, two on the tensile surface and two on the compressive surface, arranged in a full Wheatstone bridge con"guration. The output from each bridge was then recorded through a DC ampli"er with a gain of 2000 and a bridge supply of 3 V. The gauges are environmentally protected with silicon rubber and the Wheatstone bridge arrangement ensures that the gauges are fully temperature compensated. If the measured bending moments at the two gauge sites are M and M with a and b being the respective ? @ distances from the gauge sites to the point of application of the load, F, then F and a may be calculated as follows: M "Fa and M "Fb ? @
Fig. 1. (a) Transducer assembly. (b) Detail showing application of strain gauges to tensile surface of beam. Gauges (not shown) are also applied to the compressive surface. The positions of channels 1 and 2 are also shown.
that the design was still feasible and the number of electrical channels manageable. Grip strength values reported in the literature show a wide spread. Using a mechanical dynamometer, EjeskaK r and OG rtengren (1981) found values for the #exion force produced by individual "ngers of between 78 and 104 N for men. Swanson et al. (1970) quoted an average hand grip force for 100 subjects, measured on a Jamar dynamometer, of 467 N. Using pressure-sensitive "lm, Lee and Rim (1991) found a total grip force of 600 N for a cylinder of 30 mm diameter. With these values in mind, the decision was made to measure up to a maximum force of 250 N per segment. Resolving the six force components along one axis gives an equivalent value of 500 N, similar to that reported by Swanson et al. In terms of the design of transducer, the limiting factor was the strain due to bending at the gauge sites, not the shear force, thus the maximum rated load of 250 N was considered to be applied at the ends of the beams. Under these conditions, the strain gauges would be subjected to approximately 500 le. To protect against overloading the device, the design is such that all six beams come together at a load of approximately 250 N per beam. The shape of the cross-section at the gauge sites is that of a trapezium which allows the maximum amount of material to be included in the smallest overall area, whilst allowing a #at surface for gauge mounting. This maxi-
NM !M "F(a!b) ? @ M !M @ NF" ? a!b M Na" ? . F 3. Calibration Each beam was calibrated individually up to 250 N by the application of calibrated weights in four di!erent positions along the beam. The applied bending moments were calculated and the ampli"ed electrical outputs recorded. The equipment was switched on at least 1 h in advance of the calibration procedure to allow the system to warm up and stabilise, after which time no drift was seen. The calibration was performed in loading and unloading and no drift was observed during this time of
Fig. 2. Calibration graph for beam 1 showing linearity of the two channels measured at the two sites.
E.K.J. Chadwick, A.C. Nicol / Journal of Biomechanics 34 (2001) 125}128
127
Fig. 3. Subject performing the four activities: (a) vertical power grip, (b) chuck grip, (c) horizontal power grip, (d) hook grip.
20 min. The device was checked for cross-talk e!ects by the sequential loading of pairs of beams and the simultaneous recording of outputs from all other channels. Cross-talk e!ects were found to be less than 0.1% of the main e!ect in all cases. A linear regression analysis was performed giving calibration factors for each channel of each beam. A sample output from this procedure is shown in Fig. 2. The correlation coe$cients of greater than 0.999 show that the device has good linearity and the 95% con"dence interval for the predicted force values was $1% ($2.5 N at 250 N load), after subtraction of the two channels.
4. The use of the device In use, di!erent attachments can be placed at the end of the device to alter its orientation and mass, allowing di!erent grips, or activities, to be studied. The output from the strain gauge ampli"ers is sampled through an analogue to digital converter. The equipment was switched on at least 1 h in advance of the tests to allow the system to warm up. The o!set value for the gauges was measured at the start of each test (lasting a few seconds), preventing any drift which may have occurred over the course of a session from in#uencing the results. There is also provision for the attachment
128
E.K.J. Chadwick, A.C. Nicol / Journal of Biomechanics 34 (2001) 125}128
5. Conclusions The device described here represents a signi"cant advance in the measurement of functional grip force, allowing the measurement of force variation with time during functional activities as well as providing information on the distribution of grip around the hand. Its ability to achieve this with a high degree of accuracy ensures its potential as a valuable research tool. The ability of the device to be marked for motion studies has allowed it to provide valuable force and moment data for biomechanical modelling.
Acknowledgements The authors would like to thank the Medical Research Council (UK) for funding this work.
Fig. 4. Forces (N) measured around the circumference of a cylinder during four activities: (a) chuck grip, (b) vertical power, (c) horizontal power, (d) hook.
of re#ective markers to the device, for use in motion analysis studies. The device has been used to analyse the forces encountered during four di!erent grip activities, which may be described as follows (see Fig. 3): 1. Power grip with the handle vertical 2. Chuck grip with the thumb and three "ngers 3. Power grip with the handle horizontal with additional end weight 4. Hook grip with handle horizontal with additional weight under hand. Results of this study for one subject, showing the distribution of force at an instant in time, are given in Fig. 4. This information, combined with knowledge of the position of the metacarpophalangeal (MCP) joints, allows the calculation of the moments acting on the MCP joints, and the calculation of the "nger #exor tendon forces. Chadwick and Nicol (1998) have reported values of up to 200 N in the #exor digitorum super"cialis and profundus tendons for the activities described.
References An, K.-N., Chao, E.Y.S., Askew, L.J., 1980. Hand strength measurement instruments. Archives of Physical Medicine and Rehabilitation 61, 366}368. Chadwick, E.K.J., Nicol, A.C., 1998. Forces in the structures of the forearm during activities of daily living. Journal of Biomechanics 31 (Suppl. 1), 161. EjeskaK r, A., OG rtengren, R., 1981. Isolated "nger #exion force } a methodological study. The Hand 13 (3), 223}229. Lee, J.W., Rim, K., 1991. Measurement of "nger joint angles and maximum "nger forces during cylinder grip activity. Journal of Biomedical Engineering 13, 152}161. Pronk, C.N.A., Niesing, R., 1981. Measuring hand-grip force, using a new application of strain gauges. Medical & Biological Engineering & Computing 19, 127}128. Purves, W.K., Berme, N., 1980. Resultant "nger joint loads in selected activities. Journal of Biomedical Engineering 2, 285}289. Radhakrishnan, S., Nagaravindra, M., 1993. Analysis of hand forces in health and disease during maximum isometric grasping of cylinders. Medical & Biological Engineering & Computing 31, 372}376. Richards, L., Palmiter-Thomas, P., 1996. Grip strength measurement: a critical review of tools, methods and clinical utility. Critical Reviews in Physical and Rehabilitation Medicine 8 (1&2), 87}109. Swanson, A. B., Matev, I. B., de Groot, G., 1970. The strength of the hand. Bulletin of Prosthetics Research 145}152.
Technical note
A novel force transducer for the measurement of grip force E.K.J. Chadwick, A.C. Nicol* Bioengineering Unit, University of Strathclyde, Wolfson Centre, 106 Rottenrow, Glasgow G4 0NW, UK Accepted 22 July 2000
Abstract A new strain gauge transducer has been developed to measure functional grip forces. The gripping area is a cylinder of diameter 30 mm and length 150 mm and simulates the handle of a number of devices, allowing a range of activities to be studied. The device measures radial forces divided into six components and forces of up to 250 N per segment can be measured with an accuracy of $1%. The device therefore gives information about the magnitude and distribution of force around the cylinder during gripping, and has been shown to be a valuable research tool in a study of four di!erent types of grip, providing valuable input data for biomechanical models. 2000 Elsevier Science Ltd. All rights reserved. Keywords: Transducer; Grip; Hand; Force
1. Introduction The accurate measurement of external contact forces is essential for providing realistic loading data for biomechanical models. In the case of the upper limb, the measurement of grip forces during object manipulation is of prime importance. A number of devices for the measurement of grip strength have been reported in the literature. Richards and Palmiter-Thomas (1996) describe several devices falling into four groups: hydraulic, pneumatic, mechanical and strain gauge. Most of these are designed to measure maximal, not functional, forces, have only one degree of freedom and are lacking versatility. Strain gauged devices are an obvious choice for this type of study as they allow the recording of force variation with time and are sensitive and accurate (Richards and Palmiter-Thomas, 1996). Purves and Berme (1980) described the application of a six-channel pylon transducer to the measurement of "nger force. Three components of force and three moments were measured on one "nger. They measured the loads on the index "nger during tap turning and jar opening: forces up to 100 N and moments up to 5 N m were measured. An et al. (1980)
* Corresponding author. Tel.: #44-101-552-3028; fax: #44-141552-6098. E-mail address: [email protected] (A.C. Nicol).
described two strain gauged grip-measuring devices: one was uni-axial only, the other measured the radial force on each phalanx but only for one "nger. Pronk and Niesing (1981) described a uni-axial device with a range of 0}900 N and an accuracy of better than 5%. Radhakrishnan and Nagaravindra (1993) used strain gauges to measure the normal forces on all 12 "nger phalanges during maximal cylinder gripping. They found total grip forces of up to 250 N for a cylinder of 50 mm. Other devices measure distribution of loading record pressure, not force. Without information about the area of contact between the "ngers and the transducer it is not possible to calculate the joint loading from this. The limitations of existing devices have been outlined and a new device that attempts to address these has been developed.
2. Design of the new device A force transducer that has been specially designed to measure hand grip forces during activities of daily living. It consists of six active beams, mounted as cantilevers on a hexagonal core (see Fig. 1a) so as to form a cylinder of diameter 30 mm and length 150 mm, thus allowing the device to form the handle of a piece of equipment. The grip forces being measured are assumed to be radial. Six beams were incorporated to give the maximum amount of information about grip distribution, whilst ensuring
0021-9290/00/$ - see front matter 2000 Elsevier Science Ltd. All rights reserved. PII: S 0 0 2 1 - 9 2 9 0 ( 0 0 ) 0 0 1 6 8 - 8
126
E.K.J. Chadwick, A.C. Nicol / Journal of Biomechanics 34 (2001) 125}128
mises the sti!ness of the beams and thus minimises the end de#ection. Foil electrical resistance strain gauges (Showa N23MA-5-120-16) are applied to the top and bottom surfaces of the beams, at two cross-sections (see Fig. 1b), giving two channels of strain output for each beam so that the magnitude and position along the beam of the applied force may be calculated. At each of the two measurement sites are bonded four 120 ) gauges, two on the tensile surface and two on the compressive surface, arranged in a full Wheatstone bridge con"guration. The output from each bridge was then recorded through a DC ampli"er with a gain of 2000 and a bridge supply of 3 V. The gauges are environmentally protected with silicon rubber and the Wheatstone bridge arrangement ensures that the gauges are fully temperature compensated. If the measured bending moments at the two gauge sites are M and M with a and b being the respective ? @ distances from the gauge sites to the point of application of the load, F, then F and a may be calculated as follows: M "Fa and M "Fb ? @
Fig. 1. (a) Transducer assembly. (b) Detail showing application of strain gauges to tensile surface of beam. Gauges (not shown) are also applied to the compressive surface. The positions of channels 1 and 2 are also shown.
that the design was still feasible and the number of electrical channels manageable. Grip strength values reported in the literature show a wide spread. Using a mechanical dynamometer, EjeskaK r and OG rtengren (1981) found values for the #exion force produced by individual "ngers of between 78 and 104 N for men. Swanson et al. (1970) quoted an average hand grip force for 100 subjects, measured on a Jamar dynamometer, of 467 N. Using pressure-sensitive "lm, Lee and Rim (1991) found a total grip force of 600 N for a cylinder of 30 mm diameter. With these values in mind, the decision was made to measure up to a maximum force of 250 N per segment. Resolving the six force components along one axis gives an equivalent value of 500 N, similar to that reported by Swanson et al. In terms of the design of transducer, the limiting factor was the strain due to bending at the gauge sites, not the shear force, thus the maximum rated load of 250 N was considered to be applied at the ends of the beams. Under these conditions, the strain gauges would be subjected to approximately 500 le. To protect against overloading the device, the design is such that all six beams come together at a load of approximately 250 N per beam. The shape of the cross-section at the gauge sites is that of a trapezium which allows the maximum amount of material to be included in the smallest overall area, whilst allowing a #at surface for gauge mounting. This maxi-
NM !M "F(a!b) ? @ M !M @ NF" ? a!b M Na" ? . F 3. Calibration Each beam was calibrated individually up to 250 N by the application of calibrated weights in four di!erent positions along the beam. The applied bending moments were calculated and the ampli"ed electrical outputs recorded. The equipment was switched on at least 1 h in advance of the calibration procedure to allow the system to warm up and stabilise, after which time no drift was seen. The calibration was performed in loading and unloading and no drift was observed during this time of
Fig. 2. Calibration graph for beam 1 showing linearity of the two channels measured at the two sites.
E.K.J. Chadwick, A.C. Nicol / Journal of Biomechanics 34 (2001) 125}128
127
Fig. 3. Subject performing the four activities: (a) vertical power grip, (b) chuck grip, (c) horizontal power grip, (d) hook grip.
20 min. The device was checked for cross-talk e!ects by the sequential loading of pairs of beams and the simultaneous recording of outputs from all other channels. Cross-talk e!ects were found to be less than 0.1% of the main e!ect in all cases. A linear regression analysis was performed giving calibration factors for each channel of each beam. A sample output from this procedure is shown in Fig. 2. The correlation coe$cients of greater than 0.999 show that the device has good linearity and the 95% con"dence interval for the predicted force values was $1% ($2.5 N at 250 N load), after subtraction of the two channels.
4. The use of the device In use, di!erent attachments can be placed at the end of the device to alter its orientation and mass, allowing di!erent grips, or activities, to be studied. The output from the strain gauge ampli"ers is sampled through an analogue to digital converter. The equipment was switched on at least 1 h in advance of the tests to allow the system to warm up. The o!set value for the gauges was measured at the start of each test (lasting a few seconds), preventing any drift which may have occurred over the course of a session from in#uencing the results. There is also provision for the attachment
128
E.K.J. Chadwick, A.C. Nicol / Journal of Biomechanics 34 (2001) 125}128
5. Conclusions The device described here represents a signi"cant advance in the measurement of functional grip force, allowing the measurement of force variation with time during functional activities as well as providing information on the distribution of grip around the hand. Its ability to achieve this with a high degree of accuracy ensures its potential as a valuable research tool. The ability of the device to be marked for motion studies has allowed it to provide valuable force and moment data for biomechanical modelling.
Acknowledgements The authors would like to thank the Medical Research Council (UK) for funding this work.
Fig. 4. Forces (N) measured around the circumference of a cylinder during four activities: (a) chuck grip, (b) vertical power, (c) horizontal power, (d) hook.
of re#ective markers to the device, for use in motion analysis studies. The device has been used to analyse the forces encountered during four di!erent grip activities, which may be described as follows (see Fig. 3): 1. Power grip with the handle vertical 2. Chuck grip with the thumb and three "ngers 3. Power grip with the handle horizontal with additional end weight 4. Hook grip with handle horizontal with additional weight under hand. Results of this study for one subject, showing the distribution of force at an instant in time, are given in Fig. 4. This information, combined with knowledge of the position of the metacarpophalangeal (MCP) joints, allows the calculation of the moments acting on the MCP joints, and the calculation of the "nger #exor tendon forces. Chadwick and Nicol (1998) have reported values of up to 200 N in the #exor digitorum super"cialis and profundus tendons for the activities described.
References An, K.-N., Chao, E.Y.S., Askew, L.J., 1980. Hand strength measurement instruments. Archives of Physical Medicine and Rehabilitation 61, 366}368. Chadwick, E.K.J., Nicol, A.C., 1998. Forces in the structures of the forearm during activities of daily living. Journal of Biomechanics 31 (Suppl. 1), 161. EjeskaK r, A., OG rtengren, R., 1981. Isolated "nger #exion force } a methodological study. The Hand 13 (3), 223}229. Lee, J.W., Rim, K., 1991. Measurement of "nger joint angles and maximum "nger forces during cylinder grip activity. Journal of Biomedical Engineering 13, 152}161. Pronk, C.N.A., Niesing, R., 1981. Measuring hand-grip force, using a new application of strain gauges. Medical & Biological Engineering & Computing 19, 127}128. Purves, W.K., Berme, N., 1980. Resultant "nger joint loads in selected activities. Journal of Biomedical Engineering 2, 285}289. Radhakrishnan, S., Nagaravindra, M., 1993. Analysis of hand forces in health and disease during maximum isometric grasping of cylinders. Medical & Biological Engineering & Computing 31, 372}376. Richards, L., Palmiter-Thomas, P., 1996. Grip strength measurement: a critical review of tools, methods and clinical utility. Critical Reviews in Physical and Rehabilitation Medicine 8 (1&2), 87}109. Swanson, A. B., Matev, I. B., de Groot, G., 1970. The strength of the hand. Bulletin of Prosthetics Research 145}152.