The Ergonomic Design of Bent-Handled Wok for Female Cooks ’ Wok Flipping Task

Conventional straight-handled woks and the ergonomic bent-handled woks under different wok sizes were evaluated using the psychophysical approach. Twelve female subjects were tested using 15 different woks in a random order. The independent variables were the wok handle angle (with two angles of 25 ̊, 10 ̊ for traditional wok, and three angles of -5 ̊, -20 ̊ and -35 ̊ for bent-handled wok) and the wok size (with three diameters of 33 cm, 36 cm and 39 cm). The criterion measures included maximum acceptable weight of flipping (MAWF), wrist angle and subjective rating. The results showed both the handle angle and wok size significantly affected the maximum acceptable weight of flipping, wrist angle and subjective rating. The bent-handled wok is superior to the straight-handled wok, and the size-weight illusion effect is significant. In general, a small wok (33 cm diameter) with an ergonomic bent handle (-35 ̊) is ideal for females to decrease the risk of upper extremity musculoskeletal disorders when performing the flipping task.


Introduction
The prevalence of upper extremity cumulative trauma disorders (CTDs) has been found related to work activity in varied tasks (Chang et al., 1999;Li, 2002;Roquelaure et al., 2004;Lintula & Nevala, 2006;McGorry et al., 2014;You et al., 2014).It has been reported hand tools account for a large proportion of cumulative trauma disorders (Aghazadeh & Mital, 1987;Hsu & Chen, 1999), such as carpal tunnel syndrome and tendonitis (Waters, 2004).Posture, force, frequency and duration were recognized as key ergonomic risk factors of non-powered hand tools for cumulative trauma disorders (Armstrong et al., 1986;Putz Anderson, 1988;Malchaire et al., 1996).You et al. (2014) verified that prolonged exposure to non-neutral wrist postures is associated with a twofold increased risk for CTS compared with low hoers of exposure to non-neutral wrist posture.To prevent workers from becoming injured due to unnatural postures and repetitive forceful exertion of hand/wrist movements, an ergonomic bent handle idea for hand tool design was introduced (Tichauer & Gage, 1977;Tichauer, 1978), and verified by many scholars (e.g.Armstrong et al., 1982;Konz, 1986;Lewis & Narayan, 1993;Hsu & Chen, 1999;Li, 2002;Wu et al., 2015).
The wok and culinary spatula (turning shovel) are the most indispensable cooking utensils in Oriental kitchens.These two utensils have been used to cook various dishes.For most dishes the cooks need to manipulate both the wok and culinary spatula simultaneously by two hands.In general, the cooks are used to standing still with the preferred hand grasping the handles of the culinary spatula, and the other hand grasping the handle of the wok.To help the food be evenly stirred and heated in the wok, the cooks usually flip the wok with one hand and hold it with a strong grip.The operation of flipping the wok requires repetitive forceful exertion and wrist movements, combined with a static load on the upper arm and shoulder.This action increases the risk of upper extremity musculoskeletal disorders (Silverstein et al., 1986(Silverstein et al., , 1987;;Okunribibo & Haslegrave, 1999;Mital et al., 1999;Buckle & Devereux, 2002;Waters, 2004;Eksioglu, 2006;McGorry & Lin, 2007).
The wok and culinary spatula are frequently used and for a prolonged period by many cooks.However, their ergonomic properties have not been thoroughly studied.Though the culinary spatula has been studied (Hsu et al., 1994;Wu & Hsieh, 2002;Wu et al., 2015) the conventional wok with a straight handle has seldom been ergonomically studied (Wu et al., 2011).Wu et al. (2011) investigated the effects of wok size and handle angle on the task of flipping by male cooks.The results showed that both the wok size and handle angle had a significant effect on the maximum acceptable weight of wok flipping, the subjective rating and the subjective ranking.They suggested a small wok (36cm diameter with an ergonomically bent handle (-20°± 15°) is the optimal design for male cooks.However, it is not clear if these results apply to female cooks.In addition, the wok handle designed by Wu et al. (2011) is to bend down the handles purchased in the market at different angles.Because the handle is bent too low, the handle would touch the ground surface, and the user hand might easily get burned by the stove.The wok handle in the study was redesigned by ourselves.To raise the handle, it is bent upward at a right angle and then bent down at various angles.When the users hold the handle, their hand won't touch the stove fire.It is a patented, innovative design.Therefore, the aim of the present study was to assess the effects of the conventional straight-handled wok, and the ergonomic bent-handled wok with different wok sizes, on the maximum acceptable weight, wrist angle and the subjective rating in an attempt to propose an ergonomic wok to decrease the awkward posture and reduce the physical load on the female cooks.

Subjects
Twelve females with at least one year of cooking experience participated in these experiments.The mean (±std) age, height and body weight were: 19.6 (±0.67) years, 159.84 (±5.07) cm, and 58.08 (±9.90) kg, respectively.All subjects were free from any musculoskeletal and cardiovascular disorders, they were asked to sign an informed consent.Each subject was told to avoid participating in strenuous exercise and sleep normally one day before the experiment.The subjects were compensated for their participation in the experiments.

Apparatus
The wrist angle in the palmar flexion/extension and the radial/ulnar deviation planes was measured using a twin-axis Penny and Giles Biometrics electrogoniometer (model XM65).To eliminate the nonnatural postures and repetitive forceful exertion of the wrist, the upward handle of the straight-handled wok was modified to be downward bent handle.Based upon Hsu and Chen (1999), the mean of the natural angle between forearm and grasp center is about 67˚, with a standard deviation of 3.8˚.This natural angle corresponds to 23˚ downwards (between the horizontal line and the handle).In this experiment, beside two conventional straight-handled woks (upward 10 ˚ and 25˚)were used, three bent-handled woks (downward -5˚, -20˚ and -35˚) were modified from the conventional wok.In addition, to prevent the stove fire burn the hand when holding the downward handle, the base of wok handle were bent upward vertically as shown in Fig 1 .Therefore, the experimental woks used in this study were different from those of used in Wu et al. (2011), and had obtained a prize of utility model patent.Finally, to test the effect of the wok size and size-weight illusion on the maximum acceptable weight of the flipping task, according to a simple survey among housewives, the three most preferred wok sizes (33 cm, 36 cm and 39 cm) were used.The weight of the three woks were 905 g, 1,094 g and 1,194 g, for the 33 cm, 36 cm and 39 cm diameter woks, respectively.Here, a total of 15 woks were used in this study.Fig 1 depicts the size and shape of the woks used in this experiment.

Experimental Design
The experiment used a two-factor, within-subject design.The independent variables were the wok size (with three diameters of 33 cm, 36 cm and 39 cm) and the wok handle angle (with two angles of 25°, 10° for straight-handled wok; and three angles of -5°, -20° and -35° for bent-handled wok).Fifteen size and angle combinations were used in the experimental conditions.The experimental task was the same as Wu et al. (2011).The criteria measures were the maximum acceptable weight of flipping (MAWF), wrist angle(palmar flexion/extension and radial/ulnar deviation), and the subjective rating under various flipping task conditions.These criteria measures represented psychophysical, biomechanical and physiological, respectively.To prevent the effects of cooking utensils, the size of the cooking utensils was controlled as follows: the handle of culinary turning shovel was 25 cm length with 25° for the lifting angle, the height of the gas stove was 9 cm, and the container size was 24ψcm×12 cm height.The working platform height was adjusted according to the subject's knuckle height in order to standardije the distance between shoulder and stove.The temperature and relative humidity in the laboratory were about 24°C and 50 %, respectively.Figure 2 demonstrates the laboratory setting.In the first training session the subjects were asked to use the woks with the straight-handle angle of 10° in three wok sizes (33 cm, 36 cm and 39 cm) to practice the flipping task.This was done to get them familiarize with the flipping task and the muscle group used according to the instructions given by the speaker.In the second to fourth training sessions, the subjects were asked to practice flipping five woks in a random order.They practiced flipping each wok for 2 minutes, including "adjusting the weight" and "flipping the wok after determining the weight".They practiced adjusting the weight of the food in the wok using the psychophysical approach to determine the maximum acceptable weight.
After training, each subject performed the flipping task using all 15 woks in random order.On any given day, data for only one experimental combination was collected for each subject.The procedures for collecting each data set included: (1) reading the psychophysical instructions (Ciriello et al., 1983); (2) performing the warming-up exercises, the experimenter led the subjects in stretching/rotary exercises, for the shoulder, arm, wrist and palm, four times; (3) wearing the electrogoniometer on the non-dominant hand; (4) adjusting the weight in the wok and performing the simulated flipping the wok tasks; and (5) filling the subjective rating scale.

Simulated Wok Flipping
When the subjects formally participated in the wok flipping experiment, they were required to read the psychophysical instructions first and then perform the warming-up exercise for ten minutes.Then, the subjects wore the electrogoniometer to measure the angles of their wrist.The experimental stage was then begun.
The experimental stage was divided into three steps.The first step involved "the subjects adjusting the starting weight according to their perceived exertion", lasting one minute.The subject grasped the handle of the wok with her non-dominant hand, and adjusted the amount of soybeans in the wok with her dominant hand holding the measuring glass according to her perceived exertion until she could accept the weight in the wok while performing the flipping task.
The second step followed, which was "flipping the wok and adjusting the weight", lasting three minutes.The subject was instructed to lift the wok and flip the wok with her non-dominant hand, based upon the speaker sound, to redistribute the soybeans and evenly heat them in the wok.The subject should flip the wok with the help of the culinary turning shovel lest the soybeans drop out of the wok.If the subject thought the wok was too heavy or too light, she could then adjust the amount of soybeans in the wok with her dominant hand holding the measuring glass according to her subjective perceived exertion until it represented the maximum acceptable weight she could flip for a formula without straining herself or becoming unusually tired, weakened, overheated, or out of breath (Ciriello et al., 1983).The frequency of flipping the wok was once per second.At the end of this step, the speaker said, "Please hold the culinary turning shovel".The third step, which was "simulating the flipping wok task", lasted two minutes.Beginning with the adjusted and determined weight in the second step, the subject performs the task according to the instructions given by the speaker.Each working cycle lasted 15 seconds.The subject should lift the wok and shake it three times (one time per second) with her left hand, and then put it down.At this time, the subject should lift the culinary turning shovel with her right hand to perform the simulated food stir-frying task, first from right to left, then from left to right, and from front to back, altogether three times.After evenly stirring the soybeans in the wok, she then stops the task, waiting for the next cycle.After the subject does the task eight times, the experimental task is finished.
At the end of the flipping task, the experimenter weighs the weight of wok and soybeans, and asks the subject to report her fatigue and awareness of the various parts of her body on the subjective rating scale to collect the psychophysical perceived stress data.Figure 3 shows the experimental steps of the simulated wok flipping.Table 1.Summary of the maximum acceptable weight of the flipping wok for the subjects

Maximum Acceptable Weight of Flipping
The subjects performed the experimental task for five minutes to determine the maximum acceptable weight of flipping (MAWF).Table 1 shows the means and standard deviations of the MAWF.
As expected, the results of the analysis of variance (Table 2) show the wok size and the handle angle of the wok significantly affect the MAWF (p < 0.05).As for the wok size, the Duncan multiple range test in Table 3 shows the MAWF produced by the 39 cm wok was significantly heavier than the 33 cm wok.There was no significant difference between the 39 cm and 36 cm wok.As for the handle angle, the MAWF produced by the -35° handle was significantly heavier than the -5°, 10° and 25° handles.There was no significant difference between the -35° and -20° handles.In addition, the difference between the -20°, -5° and 10° handles was not statistically significant.The 25° handle was the worst, but not significantly different from the 10° handle.

Wrist Angle
Table 4 shows the means and standard deviations of the wrist angles for each deviation type.

Ulnar Deviation
Both the wok size and the handle angle significant affected the ulnar deviation (Table 5).As for the wok size, the Duncan multiple range test shows the 39 cm wok had the greatest ulnar deviation, while 36 cm and 33 cm woks did not differ between themselves.As to the handle angle, the -35° handle had the least ulnar deviation compared to the -20°, -5°, 10° and 25° handles; the -20° handle came next; the -5° handle came third; and the 10° and 25° handles had the greatest ulnar deviation.

Radial Deviation
Data analysis of the wrist angle shows there was no significant difference in the radial deviation of the subjects for each wok.

Palmar Flexion
The analysis of variance in Table 5 indicates both the wok size and handle angle significantly affected the wrist angle of palmar flexion.As for the wok size, the Duncan test results in Table 6 reveal the 33 cm wok had the least palmar flexion, the 36 cm wok came next; and the 39 cm wok had the greatest palmar flexion.For the handle angle, the -35° handle had the least palmar flexion compared to the -20°, -5°, 10° and 25° handle, but did not significantly differ from the -20° handle, and the 25° handle had the greatest palmar flexion.

Palmar Extension (Dorsiflexion)
The ANOVA in Table 5 shows both wok size and handle angle significantly affected the palmar extension.As for the wok size, the subsequent Duncan test (Table 6) indicated the 39 cm wok had the least palmar extension, but did not significantly differ from the 36 cm wok, and the 33 cm wok had the greatest palmar extension.As to the handle angle, the 25° handle had the least palmar extension compared to the 10°, -5°, -20° and -35° handles, but did not significantly differ from the 10° handle.The -5°, -20° and -35° handle were significantly different from one another, and the -35° handle had the greatest palmar extension.

Subjective Rating
To rate the subject's subjective feeling and her perceived exertion, after each test, the subject was asked to fill out a semantic differential scale.The scale contained five adjective pairs: wrist aching -wrist not aching, arm aching -arm not aching, shoulder aching -shoulder not aching, gripping easily -gripping difficulty, and exerting easily -exerting difficulty.Each pair had an unmarked scale from one (aching or difficulty) to nine (not aching or easy).The higher the scores on the scale, the less fatigued the subject felt, or the better the design.Table 7 summarizes the means and standard deviations of the subject's rating.Table 8 and Table 9 show the summary of the ANOVA results and Duncan multiple range test of the subjective rating, respectively.Both the wok size and the handle angle significantly affected the subjective rating for the subject's wrist aching (Table 8).With the wok size, significantly less ache in the wrist was reported in the 33 cm wok than the 36 cm and 39 cm wok, and significantly less ache in the wrist was reported in the 36 cm wok than the 39 cm wok.With the handle angle, significantly less ache in the wrist was reported in the -35° handle than the -5°, 10° and 25° handles, while there no significant reported difference from the -20°handle (Table 9).
The wok size significantly affected the subjective rating for the subjects' shoulder aching (Table 5).As shown in Table 8, significantly less ache in the shoulder was reported in the 33 cm wok than the 39 cm wok, while there was no significant reported difference from the 36 cm wok.There was no significant difference reported in shoulder ache between the handle angles (Table 9).

Ease of Grip and Exertion
Both the wok size and the handle angle significantly affected the subjective rating of the subject's ease of grip (Table 8).With the wok size, significantly more ease of grip was reported in the 33 cm wok than the 36 cm and 39 cm woks, and significantly more ease of grip was also reported in the 36 cm wok than the 39 cm wok.With the handle angle, significantly more ease of grip was reported in the -20° handle than the 25° handle, and there was no difference between the -20°, -35°, 10° and -5° handles (Table 9).
Both the wok size and handle angle significantly affected the subjective rating of exertion easiness (Table 8).
With the wok size, significantly more exertion ease was reported in the 33 cm wok than the 36 cm and 39 cm woks, and significantly more exertion ease was reported in the 36 cm wok than the 39 cm wok.With the handle angle, significantly more exertion ease was reported in the -35° handle than the -5°, 10° and 25° handles, and there was no difference between the -20°, -5° and 10° handles.Significantly less ease of exertion was reported in the 25° handle (Table 9).

Discussion
There are numerous factors influencing the comfort and performance of the cooks.This study focused attention on the effect of the wok size and handle angle on the MAWF, wrist angle and subjective rating.The results show both wok size and handle angle significantly affect these criteria.The interaction effects among all responses for the wok size and handle angle were not significant.Therefore, the ideal wok size and handle angle can be selected independently from each other without considering the combined effect of these two factors.As shown in Figure 4, the subject's MAWF increased with the wok size.This finding was consistent with Wu et al. (2011).Based upon biomechanical mechanisms, the center of gravity of a larger wok would be relatively far away from the exertion point of wrist which in turn caused low maximum acceptable weight of flipping.These conflicting results may be caused by the effect of the size-weight illusion (Luczak & Ge, 1989).In other words, the subjects perceive larger woks to be lighter than smaller woks of the same mass, and lift greater masses.This phenomenon is very important because any decrease in perceived heaviness accompanying larger woks could lead the subjects exceeding the limits of safety (Naylor & Amazeen, 2004).

Maximum Acceptable Weight of Flipping
Figure 5.Effect of wok size on the mean wrist angle for all deviation types 4.1.2Wrist Angle Figure 5 show the effect of the wok size on the overall mean wrist angle for all three deviation types for the subjects' wrist angle.The 39 cm wok caused a greater wrist angle than the 33 cm and 36 cm woks.In general, the 33 cm wok decreased the wrist angle significantly, especially the ulnar deviation and palmar flexion.This decrese in wrist angle for small woks may be primarily due to increased MAWF of larger woks.As mentioned by McGorry et al. (2014), carpal tunnel pressure (CTP) increase during wrist extention when employing a pinch or power grip.Therefore, with respect to wrist angle, the small wok is superior to the larger wok.

Subjective Rating
A wok with a bigger diameter is usually quite heavy and its center of gravity is far from the hand.When a subject does the flipping task, he will be apt to feel fatigue in his wrists, arms and shoulders.It will also be difficult for him to hold, and exert with the wok.Because of his difficulty in exertion with the wok, some uncoordinated postures and movements will appear in his hands, causing low wok assessment scores.As shown in Table 7, the 33 cm wok was reported the highest in overall subjective rating (6.00), while the 39 cm wok was reported the lowest (4.37).This may be caused by the overexertion of the larger MAWF produced by the size-weight illusion.

Overall Comparison
Figure 6 shows the data for the three criteria used in this study.The criterion scales in Figure 6 were based upon relative percentage and converted into fairly arbitrary values respectively.Only "desirable" and "undesirable" attributes were indicated (Hsu & Wu, 1991, Wu & Hsieh, 2002).It can be seen the 33 cm wok was the best in the wrist angle and subjective rating.However, when considering the size-weight illusion effect, the 33 cm, wok was better than the 39 cm wok.Further comprehensive comparison of the three criteria in the  user can keep his wrist neutral and avoid lifting his arm and shoulder.This improves the subject's MAWF.This finding was consistent with previous studies (Wu et al., 2011;Hsu & Chen, 1999).

Figure 4 .
Figure 4. Effect of wok size on the MAWF

Figure
Figure 6.C

Figure 7 .
Figure 7. Effect of the handle angle of the wok on the MAWF le in this study et al. (2011

Table 2 .
Table 5 and Table6show the results of the analysis of variance and Duncan multiple range test, respectively.ANOVA for subject's maximum acceptable weight of the flipping wok

Table 3 .
The Duncan multiple range test of the MAWF

Table 4 .
Summary of the mean wrist angle of the subjects for each deviation type

Table 5 .
Summary of the ANOVA results of the wrist angle for each deviation type

Table 6 .
The Duncan multiple range test of the subject's wrist angle for each deviation type

Table 7 .
Summary of the means of the subjective rating*

Table 8 .
Summary of the ANOVA results of the subjective rating

Table 9 .
Summary of the Duncan Test of the subjective rating