Ann. occup. Hyg., Vol. 46, No. 2, pp. 157-163, 2002
© 2002 British Occupational Hygiene Society
Published by Oxford University Press
Article |
Temperature Limit Values for Gripping Cold Surfaces
1Université catholique de Louvain, Clos Chapelle-aux-Champs 30–38, 1200 Brussels, Belgium; 2TNO Human Factors Research Institute, Soesterberg, The Netherlands; 3Loughborough University, Loughborough, UK; 4National Institute for Working Life, Solna, Sweden; 5Oulu Regional Institute of Occupational Health, Oulu, Finland
Received 11 July 2001; in final form 22 October 2001.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Objectives. At the request of the European Commission and in the framework of the European Machinery Directive, research was conducted jointly in five different laboratories to develop specifications for surface temperature limit values for the gripping and handling of cold items. Methods. Four hundred and fourteen experiments were run where male and female subjects were invited to grip for up to 20 min cold bars of different contact coefficients, i.e. polished wood, nylon, stone, steel and aluminium. The air temperature and the bars’ initial surface temperatures ranged between 0 and –30°C for the various experiments. While gripping the bars, either only the hand or the whole body was exposed to cold. Results. The data were used to develop a prediction formula and a graph of the surface temperature limit values in order for the skin contact temperature not to reach <15°C. This duration is shown to offer a significant degree of safety with respect to the minimal surface temperature spontaneously tolerated by the subjects. Conclusions. Experiments and modelling must be pursued to extend these data to other conditions of exposure.
Keywords: cold stress; gripping; contact temperature; machine safety
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
In a number of activities, workers or people in general have to handle and grip with the hand, tools or objects in cold environments. In some other circumstances, they might come in contact with a cold surface accidentally. In this case, the contact area might be smaller, such as a fingertip. In both cases, contact between the skin and the cold material can induce discomfort, pain or frostbite. Furthermore, exposure to cold may negatively influence manual dexterity and sensitivity of the subject (Enander, 1984; Havenith et al., 1995; Powell et al., 2000). Contact cooling while handling metal objects has been studied (Daniels, 1956; Havenith et al., 1992; Chen et al., 1994a,b), but a systematic comparison of different materials at different temperatures is lacking.
In the context of the essential requirements for safety of machinery under the European Machinery Directive, it is therefore necessary, to prescribe surface temperature limit values for cold elements which might be touched, and also to prescribe the duration for which an element can be handled safely at a certain temperature. The Directive necessitates a standard to provide manufacturers with data so that they can fulfil these essential requirements. Such a standard exists concerning touching of hot surfaces (European Standard EN 563, 1994), but data have been missing for the development of a parallel standard for cold surfaces (Holmér et al., 2001).
The European Commission requested research to collect the data needed for the development of such a standard. A concerted research project was conducted with the contribution of five laboratories from five different European countries, using the same methodology. A companion paper will report on the results of touch experiments where limit values were obtained for short duration contact with the tip of the forefinger. The present paper will concentrate on the conditions where a person is led to grip a cold part such as a tool or a handle for up to 1000 s.
Apart from temperature, the risks associated with gripping cold objects are also determined by various other factors, such as material characteristics (e.g. conductivity, specific heat, specific weight, surface texture), as well as subject characteristics (e.g. contact area, skin structure, blood perfusion of the skin, sensitivity to the cold) and contact itself (e.g. contact pressure and type of grip) (Jay and Havenith, 2000; Piette et al., 2000; Powell et al., 2000; Rissanen et al., 2000).
In the context of a standard, the intrinsic characteristics of the material must be taken into account. Other conditions linked to the subject or the material surface texture must be averaged, as the only information available in practice will be the type of material. Furthermore, the exposure condition (environment) can vary from neutral to the same temperature as the material. For the purpose of making a standard, all conditions are therefore to be averaged.
The objective of the research was to determine, for whatever exposure context and for any adult subject, the maximum allowable duration of gripping as a function of the properties of the material and its initial surface temperature.
The concept of maximum allowable duration was understood in two different ways, as follows:
The spontaneous duration that 75% of the subjects, in laboratory controlled conditions, were prepared to hold a cold object.
The duration after which 75% of these subjects were expected to have an average temperature of the palm of the hand >15°C. This limit of 15°C was adopted as it was shown (Havenith et al., 1992, 1995) that this temperature corresponds, on average, to the sensation of ‘slightly painful’ in a neutral environment.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Materials
Five materials were chosen to be representative of the whole range of contact factor values. The contact factor is defined as the square root of the product of the thermal conductivity
(W/mK), the density 
(kg/m3) and the specific heat C (J/kg K). These characteristics are given in Table 1 for the five chosen materials.
|
Bars of the five materials were prepared, 30 or 40 cm in length and 4 cm in diameter. Their surfaces were polished. They were hung in a temperature-controlled cold environment with a counter weight to keep the holding force constant at 0.5 kg for all the materials.
The temperature of the bars was measured continuously beneath the location of the hand during the experiments: this is called the contact temperature Tc (°C), influenced by the temperature of the bar and the cooling of the hand. The temperature was also recorded on the top of the bar, away from the hand: the surface temperature Ts (°C), reflecting the initial temperature of the mass of the bar.
Two different sets of cold environments were investigated: cold boxes where only the hand was exposed to cold air and cold rooms where the subject as a whole was exposed to cold.
The cold boxes were used in two laboratories. They made it possible to regulate temperatures between –35 and +5°C with an accuracy of ±0.5°C. The cold boxes had an opening to allow the subjects to insert their dominant hand and half of the forearm, in order to lift and grip the bar. Cold boxes had windows and light so that the grip could be controlled.
The cold boxes themselves were placed inside a climatic room, air conditioned at a temperature of 20°C and a relative humidity of 50%, in order for the PMV index (ISO 7730, 1984) to be equal to –1 (slightly cold). The clothing insulation of the subjects was 0.6 clo.
In the other three laboratories, the experiments were carried out in climatic rooms so that the whole subject was exposed to the required temperature as mentioned above. The clothes were adapted to the climatic chamber temperature so that the PMV index was also equal to about –1. At ambient temperatures >–20°C, the clothing was: long underwear, long trousers and insulated winter jacket. At ambient temperatures <–20°C, an additional middle layer was used. Then the total insulation was ~2.2 clo. Hat, winter boots and mitten (on non-contacting hand) were also used.
Subjects
Twenty-five women and 24 men gave their free informed consent to participate in the research. The conditions of selections were: not having worked recently in, nor being acclimatized to cold; having used vibrating tools only very occasionally; not having any history of musculoskeletal, vascular or neurological complaints or disorders; and being able and willing to grip the test bars for a maximum of 20 min in the different temperature conditions.
For each person, the following information was collected: age, weight, height, gender and characteristics of the dominant hand as follows:
Hand volume measured by the difference in weight of a water basket with and without the hand of the subject immersed up to the wrist.
Hand surface and contact surface: the contour of the hand was outlined when pressing flat on a sheet of white paper, or when gripping the bar surrounded by a sheet of paper. In each case, the pencil was held perpendicular to the paper. The contours were then cut out of the sheet and placed on black paper in a scanner. The size of the contour was counted as the number of pixels inside. Comparisons of the results from three different operators measuring both surfaces on three different occasions for five subjects proved the method to be reproducible.
Table 2 gives the characteristics of the 49 persons who participated in the study. They were essentially young college students. Attention must be drawn to the ratios between men and women for the surfaces of the hand in contact with the bar, the total surface and the volume.
|
Measurement of skin temperature
In four laboratories, contact temperatures were measured with very small thermocouples (K-type, diameter 0.2 mm) fixed on three sites on the palm (near index finger, thumb and little finger) of the dominant hand. Thermocouples were attached by a small piece of surgical tape (Blenderm, 3M), while the tip of the thermocouple was left uncovered.
The contact temperature for the palm was computed as the average of the three measurements.
In the fifth laboratory, the contact temperature during the grip was measured at the level of the metacarpo-phalangian articulation of the fourth finger at the centre of the palm, using a Craftemp probe (model 6202). This thermistor was very small and light, so that it made it possible to record the contact temperature between the skin and the bar with very fast response. In addition, at regular time intervals (2 min) during the experiments, the subjects were invited to withdraw their hands from the cool box and to place them, as fast as possible and for 3–4 s, at a reference point indicated by a laser beam, 15 cm in front of a infra-red thermometer (Ultrakust Thermopil M202 with probe T1051) in order to record the mean temperature on the palm which was in contact with the cold bar.
Protocol
As shown in Table 3, experiments were distributed between the different partners (10 subjects per laboratory) in 13 conditions in order to cover the broadest range of exposure in terms of surface temperature (Ts, °C) and to obtain replications of conditions with subjects from different countries.
|
The protocol was as follows. The subject rested comfortably for at least 15 min, during which the course of the experiment was explained and the individual data were collected. After 15 min, the gripping experiment was started. The experiment lasted 20 min or less if the subject decided to stop for any reason.
The first parameter recorded was the spontaneous grip duration (DSP, in seconds).
From the curve of evolution of the contact temperature during the experiment, the time needed to reach a contact temperature of 15°C (D15, s) was determined by linear interpolation or extrapolation. When the cooling curve was such that the contact temperature of 15°C would never be reached, a value of 9999 was encoded, indicating that the datum was missing, but that the experiment had lasted 20 min.
Many other parameters were recorded, but the present paper will concentrate on the surface temperature, the contact temperature and the two durations measured as indicated above.
The experimental data obtained from 414 experiments were pooled in a database for statistical analysis.
Statistical analysis
As the results indicate, most of the distributions of durations in a given condition were non-Gaussian and not symmetrical, especially for DSP, limited to 1200 s when the subject ended the 20 min exposure period. Therefore, the means and standard deviations were not computed and medians and quartiles were used for statistical analyses.
As more data are missing (values at 1200 s and 9999) for some conditions than for others, an analysis of variance (ANOVA) to determine the ‘experiment’ effects would be biased. However, this type of analysis can be used to highlight a ‘gender’ or an ‘environment’ effect, if the assumption is made that this effect is a constant additive or multiplicative value. With this assumption, indeed, the missing values are less likely to influence the results. In order to improve the validity of the analysis of variance and as the durations DSP and D15 appeared to vary exponentially as a function of the material, these data were logarithmically transformed before the ANOVA.
A non-linear regression model was then developed to obtain the best prediction of these quartile values of DSP and D15 as a function of the surface temperature (TS) and the contact factor (CF). In order to make this prediction model as universal as possible, a single model was preferred over having different models for each condition. This choice will affect the predictive quality of the equation.
The best non-linear model was found to have the following form:
The coefficients a, b, c and d were estimated by an iterative non-linear regression procedure and, for each model, the first general expression was simplified when some of the coefficients were not significant.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Table 3 provides the descriptive statistics for the spontaneous grip duration in the 13 exposure conditions. Of these 414 durations, 302 were below the limit value of 20 min.
Table 4 gives the descriptive statistics for the time to reach a contact temperature of 15°C, which could be evaluated in 338 cases out of the 414 (76 values set at 9999).
|
The multifactor analysis of variance of log(DSP), taking into account the ‘experiment’ and ‘gender’ effects, showed that the results for men and women did not differ significantly (P > 0.1).
On the other hand, the ‘gender’ effect was highly significant (P < 0.001) with respect to log(D15), with women reaching a contact temperature of 15°C in ~70% of the time needed for men.
The analysis demonstrated statistically significant differences between the exposures with the hand only (and inevitably half of the forearm) in the cold versus the whole body. However, the differences were in opposite directions for the spontaneous grip durations (DSP on average 3.3 times greater when the subject stayed entirely in the cold room) and for the time to reach 15°C (D15 on average 1.6 times greater when a cold box was used).
Tables 3 and 4 show that, for some combinations of material temperature, the number of points available to determine the lower percentile P25 was very limited. Furthermore, while the P25 values for DSP show a consistent pattern of increase with the material surface temperature, the P25 values for D15 show a strongly inconsistent pattern for the three materials with the largest contact coefficients (stone, steel and aluminium). This may be related to an unequal distribution of the different experimental set-ups over the temperatures.
As, moreover, gripping conditions with deep cooling of the hand are likely to occur only for grip durations >100 s (i.e. for wood and nylon), it was decided to limit the modelling of D15 to situations with low contact factors and to use solely the P25 values for wood and nylon.
For shorter durations, the stricter limits derived from the touch experiments (to be reported elsewhere) have to be applied.
The final model for the spontaneous grip duration DSP was for wood, nylon, stone, steel and aluminium (R2 = 0.94):
The final prediction model for the time to reach 15°C, derived only from data for wood and nylon was (R2 = 0.99):
All the models were quite accurate, as indicated by correlation coefficients close to 1.
Figures 1 and 2 show the predicted values from the models for each parameter. The observed values are also indicated.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
The contact temperature was measured in two different ways. Both had advantages and drawbacks. The thermocouples taped to the skin recorded the temperature continuously but punctually at different places on the hand. They could change position if the subjects moved their hands along the bar. The infra-red thermometer recorded the true mean skin temperature on the total palmar surface of the hand, but these recordings were intermittent. In addition, the hand was withdrawn from the cold box. A previous study (Malchaire et al., 1986) had shown that the skin does not rewarm instantaneously, so that the palm skin temperature does not vary significantly in the first few seconds. Comparison experiments were run to verify that both methods were providing equivalent results.
As the objective of the research was to develop limit values applicable to the general population and as the group of subjects included as many women as men, the data were not corrected for this gender effect.
However, if applied to a strictly male or female population, the limit for D15 derived below can be multiplied or divided by a factor 1.2 for males and females, respectively. Clearly, the contact temperature decreases more rapidly for women than for men. Although the data do not permit quantification of this effect, it appears that it is mainly associated with the smaller mass of the hand (Jay and Havenith, 2000), even though these hands usually have also a smaller surface and therefore are less in contact with the cold source.
As for the data for the two genders, data from the two different sets of exposure environments were deliberately mixed. The first set refers to situations such as cold rooms where heavily clad subjects must handle cold and dry items. The experiments did not concern conditions where the objects were wet or iced. The other exposure conditions referred to environments where only the hand is exposed to cold. Such conditions are more likely to be met in industry, where workers come into contact with refrigerated pipes or pieces of machinery. Again, in many conditions, these parts might be iced or wet: the consideration of these situations was outside the scope of the present research.
Spontaneous durations were smaller when using the cold boxes; it can be suggested that, in these conditions, greater discomfort resulted from greater skin temperature heterogeneity over the whole body and that, subjectively, the subjects were less inclined to endure low local cooling. On the other hand, the time to reach a contact temperature of 15°C was shorter when the subject was in the cold chamber. This suggests that, despite the heavy clothing, total peripheral cooling was faster. Furthermore, the initial temperature of the palm of the hand was slightly lower with the experiments in the cold chamber (28.6 versus 26.7°C).
Comparison between the lower quartile values of Tables 3 and 4 (spontaneous grip duration DSP versus time to reach 15°C) shows that the subjects tolerated contact temperatures <15°C and therefore pain levels much higher than ‘slight pain’. These subjects were obviously out of an industrial context, motivated by their participation in the research and reassured by the fact that the test conditions were carefully and continuously controlled and that the experiment would be stopped in case of a problem. They knew that they would, at the most, incur temporary discomfort with no possible permanent effects. Some of them might therefore have been prepared to exceed what they would tolerate in normal and unexpected conditions. It is reassuring that D15 appears to offers a significant degree of protection with respect to DSP.
The reverse is true for wood, the reason being that the experiments at –30 and –20°C were actually stopped not because of cooling of the palm, as the contact factor of wood is very small, but because of cold pain in the little finger tip exposed to the cold air.
Due to this fact and because D15 was often <100 s in the conditions investigated, a valid algorithm for the prediction of D15 could not be developed, in particular for stone, steel and aluminium.
The DSP and D15 values from each experiment were compared when available and the analysis indicated that the ratio D15/DSP was, on average, equal to 0.60 ± 0.78. Therefore, at the present stage of knowledge, since additional data are lacking, the proposition can be made to adopt, whatever the gender and the exposure conditions, duration limit values, DL, defined as being 60% of the values predicted for DSP. These limit values would then be close but not equal to D15. They can be predicted from the following expression:
In order to present the results in a way comparable to the existing standard for hot surfaces (European Standard EN 563, 1994), Fig. 3 gives the surface temperature limit values as a function of the duration limit values DL. As indicated earlier, the model was limited to the range 100–1000 s and, as not enough data are available for surface temperatures >0°C, to the range from –40 to 0°C.
|
Figure 3 shows clearly that the range of surface temperatures concerned by gripping metallic objects is poorly covered by the study.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
The risk of cold injury or discomfort associated with gripping a cold object varies as a function of many parameters associated with the subject and the exposure conditions. The database reported in this paper is the largest ever assembled and still does not make it possible to define safe limits for everyone in all conditions and does not cover in a satisfactory manner the gripping of metallic objects.
Nevertheless, it can be concluded that, in the great majority of the conditions, the time taken to reach 15°C is lower than the spontaneous duration volunteers are prepared to grip cold bars without any negative effects and therefore offers a safety factor.
However, D15 could not be predicted accurately due to too many missing values in the conditions that were investigated. Therefore, it is proposed to derive an approximation, DL, of these D15 values by reducing the predicted DSP values by a factor of 0.6. An expression is proposed to predict, as a function of the surface temperature and the contact factor, this allowable duration limit value, DL, while gripping a cylindrical plain and solid bar of a given material. This DL should be close to the duration after which 25% of the subjects are expected to reach a contact temperature of 15°C and experience slight pain. This expression assumes that the material is and remains dry. It concerns also a mixed male–female population and mixed exposure conditions with the hand or whole body exposed to the low temperature.
Research is pursued to validate a physiological model that would make it possible to understand how the hand, with certain characteristics of size, mass, shape and vascularization, reacts when in contact with different materials at different temperatures in given conditions. This would allow the extension of the range of the proposed standard drastically and in particular for metals.
Acknowledgement—The project was supported by the Standards, Measurement and Testing Programme of the European Commission (research contract No. SMT4-CT97–2149).
|
|
FOOTNOTES |
|---|
* Author to whom correspondence should be addressed. Tel: +32-2-764-32-29; fax: +32-2-764-39-54; e-mail: malchaire{at}hytr.ucl.ac.be
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Chen F, Nilsson H, Holmér I. (1994a) Finger cooling by contacting cold aluminium surfaces—effect of pressure, mass and whole body thermal balance. Eur J Appl Physiol Occup Physiol; 69: 55–60.
Chen F, Nilsson H, Holmér I. (1994b) Cooling responses of finger in contact with an aluminium surface. Am Ind Hyg Assoc J; 55: 218–22.
Daniels F, Jr. (1956) Contact cooling of the hand at –20°F, Technical Report EP-22. Quartermaster Research and Development Center, US Army Natic, MA.
Enander A. (1984) Performance and sensory aspects of work in cold environments: a review. Ergonomics; 27: 365–78.[Medline]
European Standard EN 563. (1994) Safety of machinery—temperature of touchable surfaces—ergonomics data to establish temperature limit values for hot surfaces. European Committee for Standardization (CEN), Brussels.
Havenith G, van de Linde EJG, Heus R. (1992) Pain, thermal sensation and cooling rates of hands while touching cold materials. Eur J Appl Physiol; 65: 43–51.
Havenith G, Heus R, Daanen HAM. (1995) The hand in the cold, performance and risk. Arctic Med Res; 54 (suppl. 2): 1–11.
Holmér I, Geng Q, Cold Surfaces Research Group. (2001) Temperature limit values for cold touchable surfaces, Final Report of EU Project SMT4-CT97-2149. Arbetslivsinstitutet, Stockholm.
ISO 7730. (1984) Ergonomics of the thermal environment—analytical determination and interpretation of thermal comfort using calculation of the PMV and PPD indices and local thermal comfort effects. International Standard Organization, Geneva.
Jay OE, Havenith G. (2000) Skin contact with cold materials—a comparison between male and female responses to short term exposure. Proceedings of the 9th ICEE Ruhr 2000 Conference, Ruhr-University Bochum, Germany, July 30–August 4. pp. 185–8.
Malchaire J, Tibackx B, Rens G. (1986) Validité du test de réchauffement cutané dans le diagnostic du syndrome de Raynaud d’origine vibratoire. Arch Mal Prof; 47: 79–83.
Piette A, Malchaire J, Cold Surface Research Group. (2000) Duration limit after cold grip exposure with several materials. Proceedings of the 9th ICEE Ruhr 2000 Conference, Ruhr-Univerisity Bochum, Germany, July 30–August 4. pp. 193–6.
Powell S, Havenith G, Cold Surface Research Group. (2000) The effects of contact cooling on manual dexterity and cooling of the hand. Proceedings of the 9th ICEE Ruhr 2000 Conference, Ruhr-University Bochum, Germany July 30–August 4. pp. 205–8.
Rissanen S, Rintamäki H, Cold Surface Research Group. (2000) Individual variation during slow and rapid contact cooling. Proceedings of the 9th ICEE Ruhr 2000 Conference, Ruhr-University Bochum, Germany, July 30–August 4. pp. 189–91.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  T. OGDEN Annals of Occupational Hygiene at Volume 50: Many Achievements, a Few Mistakes, and an Interesting Future Ann. Hyg., November 1, 2006; 50(8): 751 - 764. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. Jay and G. Havenith Differences in finger skin contact cooling response between an arterial occlusion and a vasodilated condition J Appl Physiol, May 1, 2006; 100(5): 1596 - 1601. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
O. JAY and G. HAVENITH Skin Cooling on Contact with Cold Materials: The Effect of Blood Flow During Short-term Exposures Ann. Hyg., March 1, 2004; 48(2): 129 - 137. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||