EFFECT OF GAFF DESIGN ON LINEMEN'S PERCEPTION OF THE CLIMBABILITY OF RED PINE UTILITY POLES OF PRE-SELECTED WOOD HARDNESS.
The aim of this study was to investigate the effectiveness of a gaff adapted for chromated copper arsenate (CCA) in order to improve the linemen's psychophysical perception of pole climbability compared to the standard gaff used by Hydro-Quebec's staff. Three red pine poles were selected in a range of wood hardness where the linemen's acceptability level is around 30 to 50 percent (hard poles) and three others in a range showing generally over 80 percent of acceptance (soft poles). In a first series of tests, psychophysical data collected from 16 line workers showed that the CCA-adapted gaff provided a significant but limited improvement (10% to 15%) in the perception for the hard poles. Conversely, the use of the CCA-adapted design in conjunction with the soft poles led to a slight but significant deterioration in the linemen's appreciation. These soft pre-selected poles, which met total acceptance with the standard gaff, were found to be unacceptable by about one-third of the linemen when the CCA-adapted g aff was used. In a second series of tests, the same workers were sequentially equipped with an instrumented climber from each gaff design to record gaff penetration and gaff impact values. The results confirmed the possibility of using a combination of these physical parameters for estimating the linemen's perception when assessing the hard poles. However, the situation reported by the linemen for the soft poles cannot be followed with the same level of discrimination. In a third series of tests, a mechanical device that directly provides a climbability index was used to assess the poles. As expected, the measured values showed trends very close to those observed with the instrumented climbers. From these results, it is obvious that a third parameter such as gaff withdrawal force should be measured to use the setup for discriminating the gaff effect over these pre-selected soft poles.
A recent psychophysical evaluation of pole climbability (1) revealed that linemen's acceptance falls drastically when the outer-shell wood hardness, as estimated with a Pilodyn of 6 Joules, shows striker-pin penetration depths in the range of 10 to 12 mm. The evaluated poles were made of red pine species treated with a waterborne chromated copper arsenate (CCA) solution amended with a polyethylene glycol of average molecular weight of 1000 dalton (CCA PEG 1000). For example, the acceptability level as established by 24 experienced line workers was 79 percent for a pole rated 12.8 mm by the Pilodyn, but fell to 37 percent for another pole showing a penetration depth of 10.8 mm. Another study  indicated that CCA-PEG-treated red pine poles show Pilodyn values of less than 11 mm over a very short period of time when exposed to weathering conditions. This increase in pole hardness over time was attributed to the leaching of the PEG additive. The leaching of the additive was tentatively corrected by using an improved CCA-PEG formulation [4,10] based on a polyethylene glycol with an average molecular weight of 8000 dalton (CCA-PEGPLUS). Less soluble in water and less hygroscopic than the former additive, it was expected that use of this product would more consistently retain the climbability characteristics obser ved for freshly treated poles. However, Pilodyn readings performed every 6 months on a series of CCA-PEGPLUS-treated red pine poles submitted to weathering conditions indicated that the poles were becoming harder over time just as for poles treated with the former PEG-amended GGA solution (2).
Rather than looking at CCA-amended treatments, other investigators approached the GCA hardness issue by optimizing the design of the gaff used by the line workers. The climbing gaff is known to be one of the main components that determines pole climbing comfort and safety . By measuring the forces required to penetrate into and withdraw from CCA-treated southern yellow pine (SYP) specimens, they found a gaff design that substantially reduces the force required when using a standard gaff . Subsequent field trials showed that the modified design effectively improves the climbing characteristics of CCA-treated poles by providing better penetration and support without increasing the withdrawal force. The gaff design is characterized by a standard spur shape and tip with smoothly machined, semi-circular grooves running along the entire longitudinal section (see description of prototype gaff design no. 6 in reference ). After having concluded that the currently available CCA-amended formulations were inad equate for maintaining a permanent pole hardness similar to that conferred by the pentachlorophenol-oil treatment (PCP-oil) , researchers at the Institut de recherche d'Hydro-Quebec (IREQ) decided to evaluate the effectiveness of this CCA-adapted gaff for modifying the linemen's perception with regard to waterborne-treated products. The CCA-adapted gaff was thus compared with the standard gaff used by Hydro-Quebec linemen on red pine poles pre-selected in two wood-hardness ranges.
In a first series of tests, data on the linemen's perception were collected using a procedure described in the literature . In a second series of tests, the linemen were equipped with instrumented climbers for the purpose of recording gaff penetration and gaff impact values. It was demonstrated in a previous study  that a relationship exists between the psychophysical perception of hardness - and, by extension, of pole climbability - and gaff penetration and gaff impact data measured during the insertion of the linemen's gaff into the wood. The index obtained by combining the two parameters mathematically showed a good correlation with the hardness as perceived by the linemen. In a third series of tests, a portable device that directly provides a climbability index based on penetration and impact recordings obtained through a gaff welded to the rod of a pneumatic jack was used to assess pre-selected poles . By eliminating the variability associated with humans, this test concentrates on differences in the design characteristics of the gaff fixed to the portable assembly and the test poles, such as wood species, variation of wood density along the pole axis, combination of wood species and chemical treatment, etc. The results presented and discussed in this paper will be used to establish the effectiveness of the CCA-adapted gaff for improving the linemen's perception. These results will also be used to further validate the usefulness of a device for obtaining a correlation with the human appreciation.
Eight teams, each consisting of 2 experienced linemen were recruited from Hydro-Quebec's staff for a total of 16 linemen. Upon arriving at the site, each team received verbal instructions on the general test procedure. One lineman from each team was then asked to participate in the perception evaluation while the other was required to wear the instrumented climbers for measuring gaff penetration and impact. After the first series of tests was completed, i.e., measurement with the instrumented climbers by one subject and measurement of perception by the other, each lineman in each team was invited to switch to the other type of measurement. By doing so, each team member was considered to have performed the different measurements on an individual basis. The characteristics of the 16 linemen are summarized in Table 1. Once the tests were completed, the experimenters were able to confirm that the body weights of the sixteen participants followed a normal distribution. Then to further investigate the results, the sixteen participants were split into two distinct weight categories, i.e. heavyweight and lightweight. The characteristics of these two sub-categories are also given in Table 1.
MEASUREMENT OF LINEMEN'S PERCEPTION
The purpose of the experimental design was to study the effect of three determining factors on the dependent variable (the linemen's psychophysical perception of pole climbability): 1) the weight category; 2) the pole class; and 3) the gaff design. Overall, 24 trials were performed by each lineman: 2 gaff designs (CCA-adapted gaff Model 9106, Buckingham Manufacturing Co., Binghamton, New York and standard gaff, Model BD-16 BC, Bashlin's Linemen's Equipment, Grove City, Pennsylvania), 6 poles, and 1 replicate. The sequence of all 24 trials was completely randomized for each lineman and required about 1 hour to complete (including rest allowance). A total of 384 trials were completed by all the linemen. One trial consisted of climbing one of the six poles to a sufficient height so that the subject could establish his perception, then climbing down. After each trial, the subject used an "X" to mark the desired location on a 137-mm visual analog scale (VAS). Each subject was told during the verbal instructions t hat there were no right or wrong answers, and was asked to be as objective as possible and to try to use the entire range of the perception scale. The corresponding value was obtained using a ruler to measure the distance between the position at the X mark and the origin (extreme left). This distance was then converted into a percentage and the corresponding psychophysical perception value was placed in a range of 0 to 100 percent. This last value was used as a dependent variable for statistical analysis. At the end of each climbing session, the subject was asked to give an overall assessment by rating each pole as acceptable or unacceptable for climbing with a specific gaff design. It took 3 days to conduct the entire test (2 teams in the morning and 1 in the afternoon for the first 2 days and 2 teams in the morning on the final day).
MEASUREMENT OF GAFF PENETRATION AND GAFF IMPACT
The sensors of the instrumented climbers were attached to the linemen's right foot support according to a setup detailed elsewhere . Briefly, an aluminum slider superimposed over the gaff and connected to a 0-to-SO-mm displacement sensor (Measurements Group Inc., Model 0-2 inches, Raleigh, North Carolina) measured gaff penetration. An accelerometer (Bruel & Kjaer, Model 4370, Montreal, Canada) measured the impact generated by the penetration of the gaff along the axis of the impact. The measurements were taken continuously at a frequency of 1000 Hz. The signals were treated using a data-acquisition system consisting of Labview 5.0 software and a data-acquisition card (Model DAQ700, National Instruments, Austin, Texas). Each subject was asked to execute two climbs (1 replicate) of about 8 seconds on each pole with a given instrumented climber and after completing the six poles (pole 1 towards pole 6 or pole 6 towards pole 1), to repeat the climbs with the other instrumented climber. The data-acquisition sy stem was used to record each climb for a total of 384 recordings (2 gaff designs, 6 poles, 1 replicate, and 16 linemen). The penetration and impact values of each climb were averaged over five to seven individual strikes recorded as the lineman ascended the pole.
The climbability testing device consisted of a pneumatic jack (Bimba Manufacturing Co., Model SR-243-DP-#OJ, Monee, Illinois) 45 mm in diameter, with a 75-mm travel attached to a support through a joint and equipped with an external displacement sensor (Model SE373 150 inches also from Measurements Group), a piezoelectric charge cell (Piezotronics Co., Model PCB 202B, Depew, New York), and an accelerometer (Model Q353-1302 also from Piezotronics). The cylinder rod on which the gaff is attached at the upper edge was interchangeable so that the two gaff models could be compared. The jack was controlled by a bidirectional valve (MAC Co., Model 82A-OA-CAA, Wixon, Michigan) equipped with a choke (Schader Bellows Co., Model 3250-0219, available at Cowper, Montreal, Canada). Force was applied at a 20-degree angle with a 30-degree angle between the upper edge of the gaff and the surface of the pole. The acceleration of the device was set to a maximum value of 10 G with a Gerk of 3 km/[sec.sup.3] and an impact load o f about 1600 Newton for a mass of 1 kg. The system operates with an air pressure of 2.76 x [10.sup.2] kPa. The conditions were fixed in order to reproduce as closely as possible the penetration and impact values observed with the linemen. The force, penetration, and velocity were recorded using the Labview data-acquisition system. Three measurements about 1.1 m from the groundline were taken each day at each cardinal point for a total of 12 readings per pole.
OUTDOOR EXPERIMENTAL SITE
Six red pine poles evenly distributed among two wood hardness classes were used. The characteristics of these poles are given in Table 2. The first three poles were chosen in a wood hardness range where the linemen's acceptability level is around 30 to 50 percent (hard poles), and the three remaining ones in a range where the acceptability is generally over 80 percent (soft poles). These two hardness classes were selected to determine any improvement or decline in the linemen's perception that could be attributed to gaff design. By choosing poles of pre-selected hardness, it is obvious that the linemen's satisfaction generated by the full range of wood hardnesses that can be obtained from the application of CCA-PEGPLUS and PCP-oil treatments to red pine species could not be estimated from this study. The poles were initially selected based on surface hardness measurements performed with a Pilodyn of 6 Joules (Proceq SA, Zurich, Sweden). The surface hardness of each pole was monitored each day by taking 3 Pil odyn readings at each cardinal point 1.1 m from the groundline for a total of 12 readings per day. The first day, the averaged Pilodyn values showed a penetration range of 11.3 to 12.1 mm for the hard poles and of 15.6 to 17.3 mm for the soft poles (Table 2). After the tests, a variance check based on Bartlett's test performed over the Pilodyn data showed that the standard deviations within each of the 3-day results were the same at the 95 percent confidence level. The stability of pole conditions over the 3-day testing was then assessed by performing an analysis of variance (ANOVA) over the 3 sets of 12 individual Pilodyn data. No significant difference between the means was observed for pole 1 (p = 0.23), pole 2 (p = 0.83), and pole 3 (p = 0.08), while a significant difference was noted for pole 4 (p = 0.04), pole 5 (p = 0.006), and pole 6 (p 0.02) at the 95 percent confidence level. For the cases where p [less than] 0.05, a multiple comparison test showed that the Pilodyn data of each pole belong to the sa me homogeneous group the first 2 days, and become a different group on the third day. These poles were cut off at the groundline following the second day of testing in order to remove the section damaged by the first 12 participants. Considering that weather conditions proved to be very stable during the 3-day testing (minimum effect on pole conditions), the difference noted on the third day may reflect the fact that readings were performed at a different longitudinal pole location. From a practical standpoint, it is expected that these small variations in pole conditions would have a negligible effect on the linemen's psychophysical perception.
EFFECT OF GAFF DESIGN AND POLE HARDNESS ON THE LINEMEN'S PERCEPTION
Analysis of variance was used to evaluate the 384 results collected from the 16 linemen. Considering that the measurements tend to be correlated with each other when taken from the same experimental unit, a repeated measures ANOVA was used to take into account this correlation . The factors that appear in all combinations are weight category (two levels), gaff design (two levels), and pole class (two levels). The following experimental units were also considered: 1) subjects, which is random and nested within weight category (eight subjects drawn from a heavyweight group and eight subjects drawn from a lightweight group); 2) poles, which is also random and nested within pole class (three poles drawn from hard poles and three poles drawn from soft poles); and 3) replicates. Each combination of weight category, subject (weight category), gaff design, pole class, pole (pole class) was replicated twice. The goal was to study the changes in the measurements that were taken on each subject; in other words, it is of interest to test the hypotheses about the within subject effects and the within subject by-between-subject interactions. Since only three poles were taken in each class, a first level of simplification was applied by taking the mean of the associated observations obtained under the same combination of gaff design x pole class x replicate. For example, the first replicate refers to the mean of the three measures taken on the three poles of a specific class by a subject who used a given gaff design on a first trial, while the second replicate refers to the mean of the three measures taken under the same experimental condition on the second trial. A first repeated measures ANOVA showed that the replicate factor was not statistically significant at the 95 percent confidence level. To further simplify the analysis, it was then decided to ignore this factor by considering the mean of the six observations (three poles and two trials under the same experimental condition) for each fixed level of gaff design and pole class.
Table 3 shows the results of the repeated measures ANOVA for the "perception of climbability (%)" dependent variable with a confidence level of 95 percent to identify the significant factors. The results showed that the weight category has no statistically significant effect on the climbability perception (p = 0.4855). On the other hand, the pole class and the gaff design X pole class have a statistically significant effect on climbability (p [less than]0.0001 and p = 0.0023, respectively). The pole class X gaff design interaction curves are shown in Figure 1a. Use of the CCA-adapted design provided a significant improvement in the perception of climbability for the hard poles (poles 1, 2, and 3) when compared with the results of the standard gaff. However, the perception was improved by only 10 to 15 percent for poles rated 11.0 to 12.2 mm by the Pilodyn (Table 2). On the other hand, the same CCA-adapted gaff had a detrimental effect on perception for the soft poles (poles 4, 5, and 6) when compared with th e results of the standard gaff. The perception declined by nearly the same percentage as the increase observed for the hard poles. The magnitude of the improvement or decline in the perception is about the same for the two weight categories considered.
At the end of each climbing session, the subjects were asked to attribute an overall assessment by rating each pole as acceptable or unacceptable for climbing with a specific gaff design. Figure 2 presents the frequency of occurrence of a response for each pole X gaff design combination. As seen in this figure, use of the CCA-adapted gaff resulted in a slight increase in the linemen's acceptance of hard poles when compared with the standard gaff. On the other hand, the standard gaff resulted in a total acceptance for the three soft poles tested, while about one-third of the linemen rated these poles as unacceptable when the CCA-adapted gaff was used. It is interesting to note that the data of this overall assessment test confirm the trends shown by the application of the ANOVA over the 384 perception data collected with the VAS (Fig. 1a).
EFFECT OF GAFF DESIGN AND POLE HARDNESS ON INSTRUMENTED-CLIMBER MEASUREMENTS
The averaged gaff penetration and impact values obtained by weight categories, pole classes, and gaff designs are presented in Table 4 (database constructed using five to seven ascending strikes per climb). For the lightweight category, the use of a CCA-adapted gaff improved hard pole wood penetration by 16 percent (poles 1, 2, and 3), while for the same poles, an improvement of about 5 percent was noted for the heavyweight category. When considering the soft poles (poles 4, 5, and 6), it is no longer obvious that the CCA-adapted gaff can improve wood penetration regardless of the weight category considered (penetration improved by less than 1%). For both pole classes, the gaff of the heavyweight subjects penetrated more deeply into the wood than the lightweight subjects regardless of the design used. On the other hand, use of a CCA-adapted gaff increased by 9 percent the impact felt by the heavyweight subjects when climbing hard poles (poles 1, 2, and 3), while essentially no variation was noted for the lig htweight category on the same poles. For soft poles (poles 4,5, and 6), an increase of 3.2 percent of the impact was noted for the heavyweight subjects, while an increase of 15 percent was experienced by the lightweight participants. The observed impact variation with body weight is somewhat inconsistent with some previously published results from poles rated [less than] 12.5 mm by the Pilodyn (Table 3 in reference (8)). In light of the present study, the high impact value previously measured for the 108-kg lineman could no longer be considered as an exception but rather should indicate a normal trend of this parameter with the subjects' increasing weight. Furthermore, the trends shown by the averaged values (Table 4) do not necessarily represent what a given subject is experiencing in real life. Indeed, the analysis of individual results reveals that some participants made a lower penetration with the CCA-adapted gaff on all the poles tested, others made a deeper penetration with the CCA-adapted gaff than fo r the standard gaff, and for some others, there was a more or less equivalent penetration for both gaffs on all the poles.
A variance analysis was done to test the capability of the climbability index to achieve a correlation with the linemen's perception of climbability. For this statistical analysis, various mathematical combinations of the gaff penetration and impact data were tested as the dependent variable. The combination that allowed a better discrimination in the results was obtained by dividing the impact value by the corresponding penetration value of each strike (L/P). To be self-consistent with the previous section where a lineman performed a series of strikes aimed at positioning an X along the 137-mm VAS, all the I/P data of a specific climb (five to seven ascending strikes used per climb) were averaged to give a unique value for each trial; 9 data were missing over the 384 trials due to field recording problems. The results of the repeated measures ANOVA are presented in Table 5. It is shown that the weight category has no significant effect on the climbability index (expressed in G/mm) at the 95 percent confiden ce level as for the perception measurements (p = 0.3445). The results also showed that pole class and gaff design x pole class have a statistically significant effect on the climbability index (p [less than]0.0001 and p = 0.0066, respectively). The pole class x gaff design interaction curves shown in Figure 1b exhibit a pattern very similar to the corresponding interaction for the linemen's perception of climbability. In this figure, a lower index value means a higher level of satisfaction. This combination of the penetration and impact data collected with the instrumented climbers was successful in generating significant differences between the two gaff designs for the pre-selected hard poles, similar to the linemen's perception. This becomes a strong indication that gaff penetration and gaff impact are the prevailing physical parameters used by line workers for establishing perception when evaluating poles in this range of outer-shell wood hardness. On the other hand, the interaction curves in Figure 1b show for soft poles (poles 4, 5, and 6) a lack of significant difference in the gaff design effect, contrary to what was observed for the linemen's perception (Fig. 1a). The results of Table 5 also show a significant interaction between the gaff design, the pole class, and the weight category (p = 0.0051), which is depicted in Figure 3. As discussed previously for the results of Table 4, the use of the CCA-adapted gaff by the lig htweight subjects improved gaff penetration into the hard poles without modifying the impact, and consequently, reduced the 1/P values as shown in Figure 3. The same gaff increased substantially the impact value of the lightweight subjects for the soft poles without modifying the penetration, with the net result of an increase of the I/P values. This situation was not observed for the heavyweight subjects.
EFFECT OF GAFF DESIGN AND POLE HARDNESS ON THE TESTER MEASUREMENTS
The average gaff-penetration and gaff-impact values over the 3-day testing are presented in Table 6. These results show the same trends towards the pole classes as for the collected data with the instrumented climbers (Table 4). Use of the tester with a welded CCA-adapted gaff or a standard gaff did not generate any significant difference in wood penetration, as noted previously for the overall subjects using the instrumented climbers. This is an indication that the linemen did not modify their individual climbing technique when they used the CCA-adapted gaff. On the other hand, the average penetration corresponding to each pole class is somewhat lower than that measured with the instrumented climbers. For example, the penetration achieved with the tester for poles 1 to 6 was 22.7 mm and 22.9 mm for the standard gaff and the CCA-adapted gaff, respectively, compared to 26.5 mm and 27.7 mm for the subjects using the respective instrumented gaffs. Conversely, the average impacts resulting from the tester were s omewhat higher than the corresponding instrumented gaff results. In this case, use of a CCA-adapted gaff reduced the impact by 17 percent for hard poles (poles 1, 2, and 3) and by 10 percent for soft poles (poles 4, 5, and 6). This reduction in the impact value should be specifically attributed to differences in the design characteristics of the two spurs.
To check the capacity to identify a gaff design effect over the "climb-ability" as estimated by the tester, a variance analysis was done on the readings collected during the 3 days. The 12 I/P values obtained each day for a specific gaff on a given pole were averaged to give a unique value for a total of 36 entries (3 days, 2 gaff designs, 6 poles). The climbability index based on the I/P averaged values was used as the dependent variable, while the pole class and gaff design were the independent variables. The ANOVA results presented in Table 7 show that the pole class and gaff design have a statistically significant effect on the dependent variable at the 95 percent confidence level (p = 0.0000 in both cases). The pole class x gaff design interaction curves are shown in Figure 1c. Once again, a lower index value means a higher level of satisfaction. The graph again exhibits a pattern similar to the one observed for the corresponding interaction with the linemen's perception of climbability (Fig. 1a). This means that the tester is able to at least detect a difference in climbability associated with the wood preservative used for treating these two sets of red pine poles. Furthermore, the index generated by the tester is also able to reveal a gaff design effect for the hard poles tested (poles 1, 2, and 3). This constitutes an additional indication that gaff penetration and impact are the main factors influencing human perception in this particular range of outer-shell wood hardness. For the soft poles, the tester failed to reveal the trend shown by the human perception as expected from the previous instrumented-climber results.
The psychophysical measurements performed in this study showed that the CCA-adapted gaff has a beneficial effect on the linemen's perception of both weight categories in the hardness range of the pre-selected hard poles tested. At the end of each climbing session, the subjects reported that the CCA gaff penetrated more deeply into the wood of these poles. This observation was also supported by the averaged gaff penetration values collected with the same line workers equipped with the instrumented climbers (Table 4). The same behavior was reported some years ago from measurements performed with a dynamic test setup (9). These tests, which were performed without any human variables, showed that the CCA Model 9106 penetrated 23 percent deeper than the Bashlin Model DB-12 polished gaff in a red pine CCA-PEG specimen rated 10.5 mm by the Pilodyn. Furthermore, our line workers reported that they had no difficulty withdrawing the CCA-adapted gaff when using it to climb poles with Pilodyn readings between 11.0 and 1 2.2 mm. These overall findings confirm the observations made by the CCA-adapted gaff developers based on field tests with CCA-treated southern yellow pine poles . Notwithstanding improved gaff penetration and the lack of any gaff withdrawal problems, our results show that the linemen's perception improved by only 10 to 15 percent in this range of pole hardness. It is obvious that this improvement does not makeup for the very poor appreciation given by line workers to the CCA-treated red pine poles rated [less than] 11 mm by the Pilodyn, as encountered in real life .
Conversely, use of the CCA-adapted gaff in conjunction with the pre-selected soft poles (Pilodyn readings between 15.6 and 18.2 mm), such as those normally found on Hydro-Quebec's distribution system, led to a significant deterioration in the linemen's assessment, for both weight categories. The deterioration in perception seems to be associated with the difficulty experienced by the subjects in withdrawing the gaff from the wood poles. The detrimental effect noted is of the same magnitude as the positive effect for the pre-selected hard poles. These soft poles, which meet total acceptance with the standard gaff, were found to be unacceptable by about one-third of the linemen for climbing with the CCA-adapted gaff. The creosote-treated southern yellow pine poles assessed during development of the prototype gaff did not have such a detrimental effect on perception (5); unfortunately, the authors did not provide any information on the wood hardness of these poles. Our measurements performed with the instrumente d climbers showed that the two gaffs reach about the same penetration depth in the wood of these pre-selected soft poles. The same behavior was also observed during the previously cited dynamic tests . The author showed that the improved penetration of the CCA-adapted gaff on hard poles was totally lost when penetrating softer poles. The previously mentioned 23 percent improvement dropped to 9.4 percent for a specimen rated 13 mm by the Pilodyn, and was virtually non-existent for a PCP-oil-treated specimen rated 14 mm by the Pilodyn. The same author  also reported the force required for withdrawing the gaff after each dynamic penetration test. The results indicated systematically higher values for the CCA-adapted gaff; for example, the withdrawal force required was at least twice the normal value for a PCP-oil-treated specimen rated 14 mm by the Pilodyn. These results support the evaluation done by the Hydro-Quebec linemen who had a negative perception of the CCA-adapted gaff notwithstanding the fact th at the penetration in the soft poles was not necessarily deeper or lower than for the standard gaff.
For hard poles such as those evaluated in this study, the measurements performed with the instrumented climbers confirmed the possibility of using gaff penetration and gaff impact data to achieve a correlation with the linemen's perception of climbability. The pole class x gaff design interaction curves of both the linemen's perception of climbability and the climbability index exhibit distinctive gaff-design data in the area of hard poles (Figs. 1a and 1b). However, the situation reported by the subjects for the soft poles cannot be followed with the same level of discrimination by simply combining the impact and penetration data. This lack of significant difference in the gaff design effect is due to the fact that a third factor prevailing over the previous ones is used by the subjects for establishing their perception in this range of wood-surface hardness (Pilodyn striker-pin penetration ranging from 15.6 to 18.2 mm). This point was confirmed by doing a survey of the linemen's comments collected at the en d of each climbing session. The problems faced by the subjects during gaff retrieval were at the source of their negative assessment of soft poles. The index is based on a mathematical relation (I/P) that does not integrate any parameter on this aspect, which explains the fact that it failed to strongly discriminate between the two gaffs for soft poles.
Contrary to the results obtained with the instrumented climbers, the data collected by the portable tester are unaffected by the random variable characterizing the participating subjects. In such a case, the variability in the data should then reflect differences in the wood species, wood density along the pole axis, wood species/chemical treatment combination, and, hopefully, the design characteristics of the gaff attached to the pneumatic jack. The climbability index generated by the tester shows results very close to those collected by the instrumented climbers (Figs. 1a and 1b). However, a better discrimination is obtained between the gaff designs for the hard poles, which is probably due to the elimination of the variability associated with the body weight and the climbing technique of each subject. The tester, which also measures the penetration and impact values, cannot take into account the gaff withdrawal problem reported by the linemen. The original index was developed by measuring the linemen's per ception over a wider range of pole hardness (Pilodyn readings between 8.6 and 21.3 mm, see reference ), but all the participants were using the same standard gaff for which a withdrawal problem was never reported.
As discussed in a previous publication , each line worker may consider gaff penetration, gaff impact, and the withdrawal force differently when estimating pole climbability. For example, some subjects may give more importance to gaff penetration or gaff impact while others may give greater weighting to the force required for gaff withdrawal. The weighting attributed to each of these physical parameters by one subject may also vary along the wood hardness scale of the poles tested. The individual climbing technique used by each subject could be one of the elements contributing to the variations observed in the weighting attributed to each parameter. These human-weighting variations are clearly depicted in Figure 4 where the subject x pole interaction curves are shown for the linemen's perception of pole climbability. This figure displays at least 1 lineman out of the 16 (subject 4) who attributed about the same level of assessment regardless of the pole class tested (poles 1, 2, and 3 versus poles 4, 5, an d 6). It is interesting to see that gaff withdrawal had no negative impact on perception even though the subject reported a withdrawal problem at the end of the session.
The linemen's climbability perception of hard poles could be significantly improved by using a CCA-adapted gaff such as the commercial model tested in this study. However, the detrimental effect on the perception generated with this gaff when climbing poles of softer wood hardness hinders its use for climbing purposes in a distribution system composed of poles with a large range of wood hardnesses. This study also showed some limitations in using a portable tester for measuring gaff effects over the two wood-hardness ranges of poles tested. In order to generate a significant difference between the two gaffs for the soft poles, such as was observed with the human evaluation, a new climb-ability index combining gaff penetration, gaff impact, and withdrawal force will have to be defined.
The authors are, Senior Research Scientist and Research Scientist, Institut de recherche d'Hydro-Quebec (IREQ), 1800, boulevard Lionel-Boulet, Varennes, Quebec, Canada J3X 1S1; Professor, Ecole de technologie superieure (ETS), 1100, rue Notre-Dame Quest, Montreal, Quebec, Canada H3C 1K3; and Research Scientist, Centre de recherche industrielle du Quebe (CRIQ), 8475, avenue Christophe-Colomb, Montreal, Quebec, Canada H2M 2N9. The authors would like to thank Hydro-Quebec's Distribution Dept. fonts substantial financial contribution to this study. The authors also gratefully acknowledge P. Lachapelle and J. Germain for their assistance in organizing the Hydro-Quebec line crews at the site. Thanks also go to F. Vachon from ETS and S. Pelletier from CRIQ for their support in collecting data on the test site. The authors gratefully acknowledge the linemen who were willing to perform the field evaluation of the pole gaffs in spite of the labor problems prevailing during the test period. They also acknowledge Profes sor B. Abdous from Universite du Quebec a Trois-Rivieres for the guidance provided during the revision of the statistics. Last but not least, the authors would like to thank J. Jalbert from IREQ for his assistance in the use of the statistical software. This paper was received for publication in April 2000. Reprint No. 9111.
(*.)Forest Products Society Member.
[C] Forest Products Society 2001.
Forest Prod. J. 51(5):63-72.
(1.) In this paper, the word team does not apply to the tests performed with the subjects, but reflects the fact that the subjects came to the outdoor experimental site both at the same time.
(1.) Beauchamp, Y., M. Thomas, J. Arteau, and D. Marchand. 1999. Assessment of wood utility pole climbability using psychophysical and mechanical measurements. Inter. J. of Occupational Safety and Ergonomics 5(1):3-28.
(2.) Besner, A., R. Gilbert, P. Tetreault, and J.-F. Labrecque. 1999. Effect of weathering exposure on the preservative and softening agent retentions, wood hardness and climbability of CCA-treated red pine utility poles. Forest Prod. J. 49(10):59-68.
(3.) Brudermann, G.E. 1994. Climbability of CCA-treated poles. CEA Rept. SD-169A. Technology Dept., Canadian Electricity Assoc., Montreal, Quebec.
(4.) Cooper, P.A., Y.T. Ung, F.M.S. Ma, and W.E. Zirk. 1995. Performance of a new polyethylene glycol (PEG) additive for CCA treated red pine poles. In: Proc. 16th Conf. of the Canadian Wood Preserv. Assoc., Vancouver, BC, Canada. pp. 151-166.
(5.) Fox, R.F., H.J. Demers, T.O. Geiser, and J.E. Pennefeather. 1987. Improved climbability of CCA-treated poles through modification of gaff design. In: Proc. 83:191-213. Am. Wood Preservers' Assoc., Granbury, TX.
(6.) Gilbert, R., A. Besner, and P. Octeau. 1997. Effect of aging and temperature on the resistance to spur penetration in red pine poles treated with CCA-PEG 1000. Forest Prod. J. 47(3):81-88.
(7.) _____, _____, C. Roy, and P. Octeau. 1998. Wood hardness of utility poles. Part 1. Study of physical parameters in correlation with the linemen's psychophysical perception of wood pole hardness. Forest Prod. J. 48(10):49-58.
(8.) _____, _____, _____, and _____, 1998. Wood hardness of utility poles. Part 2. Transposition of a wood hardness indicator into a measuring instrument in correlation with linemen's perception. Forest Prod. 3.48(11/12):91-97.
(9.) Hanrahan, R.C. 1993. Factors influencing the climbability of CCA-PEG wood poles. Tech. Rept. No. HSD-SD-93-5. Ontario Hydro, Res. Div., Toronto, Canada.
(10.) Ma, F.M.S. and M. Ayotte. 1995. Compositions for the preservation of timber products. U.S. Patent 5,460,751.
(11.) SAS Institute. SAS/STAT Software. User's Guide. Vol. 2, Vers. 6, 4th ed. p.951. Cary, NC.
Characteristics of the participating linemen. Average SD [a] Maximum Minimum All subjects (n 16) Age (yr.) 40 6.5 Height (cm) 169.4 10.9 Body weight (kg) [b] 99.3 11.3 120.2 83.7 Climbing experience (yr.) 16.5 8.3 Heavyweight category (n = 8) Age (yr.) 39.4 5.1 Height (cm) 167.6 13.2 Body weight (kg) 109.1 6.1 120.2 102.1 Climbing experience (yr.) 17 5.4 Lightweight category (n = 8) Age (yr.) 40.6 8.0 Height (cm) 171.1 8.6 Body weight (kg) 89.4 3.8 94.3 83.7 Climbing experience (yr.) 15.9 10.9 Median All subjects (n 16) Age (yr.) 41 Height (cm) 172.7 Body weight (kg) [b] 98.2 Climbing experience (yr.) 17.0 Heavyweight category (n = 8) Age (yr.) 39 Height (cm) 172.7 Body weight (kg) 107.7 Climbing experience (yr.) 19 Lightweight category (n = 8) Age (yr.) 42 Height (cm) 172.7 Body weight (kg) 88.7 Climbing experience (yr.) 15.5 (a.)SD standard deviation. (b.)Including weight of the equipment. Characteristics of the red pine poles selected for the study. Pilodyn 6J striker-pin penetration [a] Pole Pole Length (ft.)/ Day 1 class no. Treatment class Average (mm) Hard 1 CCA-PEGPLUS 35/4 11.4 Hard 2 CCA-PEGPLUS 35/4 12.1 Hard 3 CCA-PEGPLUS 35/4 11.3 Soft 4 PCP-oil 40/4 16.8 Soft 5 PCP-oil 40/4 15.6 Soft 6 PCP-oil 40/4 17.3 Pole Day 2 Day 3 class SD [b] Average SD Average SD Hard 1.0 12.0 0.9 11.4 0.7 Hard 1.3 12.2 1.3 11.9 0.9 Hard 0.9 12.1 1.3 11.0 1.1 Soft 0.9 17.8 1.1 16.9 0.8 Soft 0.9 15.6 0.8 16.6 0.6 Soft 1.3 17.2 0.6 18.2 0.9 (a.)Calculatcd from 12 values. (b.)SD = standard deviation. Repeated measures ANOVA results for evaluating weight category, pole class, and gaff design effects with the linemen's perception of climbability as the dependent variable. Source of variation Df Sum of squares Tests of hypochcses for between-subject effects Weight category 1 403.54 Error 14 11004.49 Univariate tests of hypotheses for within-subject effects Gaff design 1 3.27 Gaff design x weight category 1 55.75 Error (gaff design) 14 1284.51 Pole class 1 27661.31 Pole class x weight category 1 245.81 Error (pole class) 14 6484.56 Gaff design x pole class 1 1724.72 Gaff design x pole class x weight category 1 0.86 Error (gaff design x pole class) 14 1745.16 Source of variation Mean square Tests of hypochcses for between-subject effects Weight category 403.54 Error 786.03 Univariate tests of hypotheses for within-subject effects Gaff design 3.27 Gaff design x weight category 55.75 Error (gaff design) 91.75 Pole class 27661.31 Pole class x weight category 245.81 Error (pole class) 463.18 Gaff design x pole class 1724.72 Gaff design x pole class x weight category 0.86 Error (gaff design x pole class) 124.65 Source of variation F-ratio Tests of hypochcses for between-subject effects Weight category 0.51 Error Univariate tests of hypotheses for within-subject effects Gaff design 0.04 Gaff design x weight category 0.61 Error (gaff design) Pole class 59.72 Pole class x weight category 0.53 Error (pole class) Gaff design x pole class 13.84 Gaff design x pole class x weight category 0.01 Error (gaff design x pole class) Source of variation p-value Tests of hypochcses for between-subject effects Weight category 0.4855 Error Univariate tests of hypotheses for within-subject effects Gaff design 0.8529 Gaff design x weight category 0.4487 Error (gaff design) Pole class [less than]0.0001 Pole class x weight category 0.4783 Error (pole class) Gaff design x pole class 0.0023 Gaff design x pole class x weight category 0.9350 Error (gaff design x pole class)
Averaged values of gaff penetration and impact obtained with instrumented climbers. [a] Gaff penetration Standard gaff Average SD (mm) All subjects All poles (poles 1 to 6) 26.5 6.5 Hard poles (poles 1,2, and 3) 23.7 5.9 Soft poles (poles 4, 5, and 6) 29.2 5.8 Heavyweight category All poles (poles 1 to 6) 28.3 5.9 Hard poles (poles 1, 2, and 3) 26.0 5.4 Soft poles (poles 4, 5, and 6) 30.5 5.4 Lightweight category All poles (poles 1 to 6) 24.5 6.5 Hard poles (poles 1,2, and 3) 21.2 5.4 Soft poles (poles 4,5, and 6) 27.7 5.8 Gaff impact CCA-adapted gaff Standard gaff Average SD Average (G) All subjects All poles (poles 1 to 6) 27.7 5.5 22.9 Hard poles (poles 1,2, and 3) 26.0 5.2 25.6 Soft poles (poles 4, 5, and 6) 29.3 5.2 20.2 Heavyweight category All poles (poles 1 to 6) 29.0 5.4 23.5 Hard poles (poles 1, 2, and 3) 27.2 5.1 25.5 Soft poles (poles 4, 5, and 6) 30.7 5.1 21.6 Lightweight category All poles (poles 1 to 6) 26.3 5.2 22.2 Hard poles (poles 1,2, and 3) 24.6 4.9 25.7 Soft poles (poles 4,5, and 6) 28.0 4.9 18.7 CCA-adapted gaff SD Average SD All subjects All poles (poles 1 to 6) 11.0 24.3 11.8 Hard poles (poles 1,2, and 3) 11.3 26.8 11.8 Soft poles (poles 4, 5, and 6) 10.0 21.9 11.3 Heavyweight category All poles (poles 1 to 6) 11.8 25.0 12.7 Hard poles (poles 1, 2, and 3) 11.8 27.8 12.7 Soft poles (poles 4, 5, and 6) 11.5 22.3 12.1 Lightweight category All poles (poles 1 to 6) 9.9 23.6 10.7 Hard poles (poles 1,2, and 3) 10.6 25.8 10.5 Soft poles (poles 4,5, and 6) 7.6 21.5 10.4 (a.)Values averaged over the 3-day testing period. Repeated measures ANOVA results for evaluating weight category, pole class, and gaff design effects with the climbability index obtained by the instrumented climbers as the dependent variable. Source of variation Df Sum of Squares Mean Square Tests of hypotheses for between subject effects Weight category 1 0.1335 0.1335 Error 14 1.9528 0.1394 Univariate tests of hypotheses for within subject effects Gaff design 1 0.023 0.023 Gaff design x weight category 1 0.0335 0.0335 Error (gaff design) 14 0.917 0.0655 Pole class 1 2.30 2.30 Pole class x weight category 1 0.056 0.056 Error (pole class) 14 0.36 0.026 Gaff design x pole class 1 0.072 0.072 Gaff design x pole class x weight 1 0.078 0.078 category Error (gaff design x pole class) 14 0.099 0.0071 Source of variation F-ratio p-value Tests of hypotheses for between subject effects Weight category 0.96 0.3445 Error Univariate tests of hypotheses for within subject effects Gaff design 0.36 0.5607 Gaff design x weight category 0.51 0.4865 Error (gaff design) Pole class 89.02 [less than]0.0001 Pole class x weight category 2.15 0.1646 Error (pole class) Gaff design x pole class 10.13 0.0066 Gaff design x pole class x weight 11.00 0.0051 category Error (gaff design x pole class) Averaged valves of gaff penetration and impact obtained with the mechanical device [a] Gaff penetration Standard gaff Pole groupings Average SD (mm) All poles (poles 1 to 6) 22.7 5.3 Hard poles (poles 1, 2, and 3) 20.3 4.8 Soft poles (poles 4, 5, and 6) 24.9 4.6 Gaff impact CCA-adapted gaff Standard gaff Pole groupings Average SD Average (G) All poles (poles 1 to 6) 22.9 4.7 31.3 Hard poles (poles 1, 2, and 3) 21.2 4.4 37.4 Soft poles (poles 4, 5, and 6) 24.3 4.4 25.3 CCA-adapted gaff Pole groupings SD Average SD All poles (poles 1 to 6) 7.7 26.6 6.6 Hard poles (poles 1, 2, and 3) 5.9 31.0 5.6 Soft poles (poles 4, 5, and 6) 3.4 22.6 4.7 (a.)Values averaged over the 3-day testing period. ANOVA results for evaluating pole class and gaff design effects with the climbability index obtained by the mechanical device as the dependent variable Sum of Mean Source of variation Df squares square F-ratio p-valve Main effects Pole class 1 4.87 4.87 197.59 0.0000 Gaff design 1 0.58 0.58 23.64 0.0000 Pole class x gaff design 1 0.21 0.21 8.34 0.0069 Residual 32 0.79 0.22 Total (corrected) 35 6.45
|Printer friendly Cite/link Email Feedback|
|Author:||GILBERT, ROLAND; BESNER, ANDRE; BEAUCHAMP, YVES; OCTEAU, PASCAL|
|Publication:||Forest Products Journal|
|Date:||May 1, 2001|
|Previous Article:||INADEQUATE REDRYING LINKED TO DIMENSIONAL INSTABILITY OF CCA-TREATED SOUTHERN PINE LUMBER.|
|Next Article:||A CIRCULAR INTERNAL BOND TEST FOR PREDICTING THE HOLDING STRENGTH OF FURNITURE FASTENERS.|