Evaluation of Seat Comfort of Office Armchairs : an Impact of Articulated Seat Support and Gas Spring

This paper describes the application of an alternative seating system. The aim of this alternative approach was to determine the comfort of offi ce armchairs equipped with new construction solutions ensuring articulated support of the seat as well as articulated mounting of the gas spring. An offi ce armchair with a different seat support and gas spring was selected. Operational loads were applied to the seat surface. The following parameters were measured and calculated in the course of the performed experiments: contact area, average contact pressure and coeffi cient of seat pressure distribution (SPD). A new discomfort coeffi cient D expressing seat quality was elaborated. Preliminary data suggests that the prototypes provided greater sitting comfort than did the conventional chair. It was demonstrated that the new construction solution of the gas spring support guaranteed the highest comfort of the use of the examined armchairs.


INTRODUCTION 1. UVOD
Offi ce armchairs, depending on the purpose of their utilization, can be used for several minutes or several hours.Despite advanced industrial developments, many workers are still required to adapt to the machines and thus accept less than ideal working conditions.Computer dominated jobs and industrial automation have created more sedentary tasks often chara cterized by constrained postures, high frequency (repetitive work), monotonous work requiring good eyesight, and precision work with repetitive movements in the arms, hands, and fi ngers.As a result of these limitations, a variety of musculoskeletal conditions, involving the entire upper limb, neck and back, have approached the forefront of work related disorders (Fernandez et al., 1999).The authors concluded that in light assembly and computer work tasks, an arm support system would be recommended to minimize the effort and RPE, and to maximize comfort.A study by Zhu and Shin (2012) has shown that forearm support can help computer users lessen physical stress in typing, but only when the supports are positioned at resting elbow height.Also, evaluation of a dynamic arm support for seated and standing tasks suggested that a dynamic forearm support may improve subjective comfort and reduce static muscle loads in the upper extremity for tasks that involve horizontal movement of the arms (Odell et al., 2007).
The research on sitting comfort demonstrates a particularly pronounced relationship between seat pressure and comfort.De Looze et al. (2003) concluded that the most consistent predictor of seat comfort was related to seat pressure distribution and that this relationship was considerably more straightforward than with the research that measures muscle activity or spinal profi les.Using a specially designed seat fi xture, Goossens (1998) varied pressures and found a strong correlation between the amount of pressure applied to the buttocks and discomfort.The values of maximum contact pressure and the seat contact area are most commonly employed as measures of their utilization comfort (Adler, 2007; Ebe and Griffi n, 2001; Milvojevich et al., 2000; Tewari and Prasad, 2000; Uenishi et al., 2000).According to Dhingra et al. (2003), the distribution of contact pressure is more uniform on a soft seat than on a hard one.Ebe and Griffi n (2001) confi rmed that the values of contact pressure under ischium bones can be applied as the principal criterion of foam hardness and seat comfort.There is, therefore, a close correlation between the contact area and the value of contact pressure.Reswick and Rogers (1976) described the relationship between contact pressure, time of their action and the degree of soft tissue damages.Kosiak (1961Kosiak ( , 1959) ) reported that microscopic pathological changes of soft tissues appeared already after one hour of pressure action of values not exceeding 8 kPa.On the other hand, no such changes were observed when the applied pressure had the value of 4.7 kPa.According to Hostensa et al. (2001), Landis (1930), Takahashi et al. (2010) pressures ranging from 2.7 to 4 kPa can close capillary blood vessels and cause discomfort during sitting.That is why the contact pressure of 4-8 kPa was employed as a criterion of comfort in many investigations dealing with designing of various seats and beds (Butcher andThompson, 2010, 2009;Hamanami et al., 2004;Seigler and Ahmadian, 2003;Smardzewski, 2009;Smardzewski et al., 2010aSmardzewski et al., , 2010b;;Tewari and Prasad, 2000;Wang and Lakes, 2004;Wang et al., 2004).Rasmussen and Zee (2009) made an attempt at a numerical parameter optimization of an airplane armchair.Their conclusion was that, although there were some general characteristics of seats, numerous additional factors had to be taken into consideration during the modelling process before experimental results could be used in practice.Paoliello et al. (2008) made an analysis of armchairs loading during their daily use.Vlaović et al. (2008) proposed a questionnaire method for assessing seat comfort of offi ce armchairs.Nero et al. (2011) described the application of an alternative seating concept for surgeons that refl ects the research of Zen sitting postures, which require Zazen meditators to maintain fi xed postures for long durations.The aim of this alternative approach was to provide sitters with a seat pan with sacral support that provides a more even distribution of seat pressures, induces forward pelvic rotation and improves lumbar, buttock and thigh support.The authors concluded that the sacral support of the prototype chair prevents backward pelvic rotation.Preliminary data suggests that the prototype provided greater sitting comfort and support for constrained operating postures than did the conventional chair.These fi ndings support the selective application of concave-shaped seat pans that conform to users' buttocks and refl ect Zen sitting principles.However, this solution is characterized by rigid support of seats and columns.What is lacking, however, is a wider discussion concerning the effect of the new construction of seat support on the comfort of utilization of offi ce armchairs as well as ways of assessment of this comfort.
The principal goal of the present investigations was to determine the comfort of offi ce armchairs with novel design solutions ensuring articulated support of the seat as well as articulated mounting of the gas spring.Another objective was to select the most advantageous construction design that would exert the strongest infl uence on armchair comfort.

METHODS AND MATERIALS 2. METODE I MATERIJALI
The armchair selected for the investigations was an offi ce armchair with the backrest manufactured from 34 mm thick Atria® foam and 7.85 kPa hardness (PN-EN ISO 2439), whereas the seat was made from 47 mm thick and 8.9 kPa stiffness Event® foam (Fig. 1).The column of the pneumatic spring was mounted in a fi ve-arm base using three different methods.The fi rst of them was a stiff linkage typical for majority of offi ce armchair constructions (Fig. 1a, 2a,d -type S).
The second method ensured articulated support with a possibility of regulation of the defl ection angle of the column from the perpendicular (Fig. 1b, 2b,e -type P), whereas the third one was also an articulated coupling but with no possibility of defl ection regulation (Fig. 1c, 2c -type NP).The seat together with the backrest was fi xed to the column of the pneumatic spring using two methods.The fi rst method was a stiff connection with no possibility of seat defl ection in the horizontal plane (Fig. 2, type A).The second solution consisted in the application of a VMS (Vertical Moving System) mechanism making it possible for the seat to be defl ected in horizontal plane (Fig. 2, type B).In total, fi ve different designs were investigated.
The examined armchairs were tested in accordance with the standard (PN-EN 1728:2012) (used only in relation to the load) (Fig. 3).An FSA Clinical, Vista Medical Ltd., sensor mat (sensing area 465 mm x 465 mm, poly thickness 2.5 mm, sensor size 11.1125 mm, sensor gap 3 mm, sensor arrangement 32 x 32, cover size 565 mm x 565 mm, number of sensors 1024, sensor surface 211 mm 2 , standard calibration range 13.3 kPa) was placed on the seat surface.The mat was fi rst calibrated and then connected to the computer (In- tel® Core™ i5 CPU, 2.53 GHz, RAM 4 GB, Windows 7®).Loads were applied only to the seat.The force of 300 N was imposed vertically downward in places indicated in Figure 3. Consecutive load schemes were designated from 1 to 5. Each loading lasted 60 seconds, during which values of the contact stresses between the indenter and the seat were measured with 10 Hz frequency and 0.01 kPa accuracy.Direct measurement results were recorded in a text fi le and presented graphically as distribution maps of contact pressure.
Indirect experimental results were collated in the form of diagrams comparing the following values: A (m 2 ) -contact area, p m (kPa) -average contact pressure, SPD (%) -coeffi cient (Seat Pressure Distribution, Ahmadian et al., 2002).
(1) where: n -number of sensors in which contact pressure has non-zero values, p i -contact pressure in any mat sensor, p m -average contact pressure for n sensors.
Since the comfort of sitting depends directly on the contact area, values of contact pressure as well as on the above-mentioned SPD coeffi cient, a decision was also taken to defi ne and calculate the value of the discomfort coeffi cient D (daN/m 4 ) determined on the basis of the following formula: (2) In the case of uniform distribution of contact pressure on the seat surface, the p i pressure at any sensor should be equal to the average pressure p m .In such case, the SPD coeffi cient should equal zero.Therefore, a seat characterized by low SPD values may indicate a more uniform support of the user's body in comparison with seats characterized by high SPD values.However, this does not rule out that the developing stresses will be too high for the sitting comfort.In the case of D coeffi cient, it should be expected that high discomfort of the user will be achieved at high p m pressure as well as at low values of A and SPD.In such case, low values of the D coeffi cient will speak in favor of a high comfort of seat utilization.Figure 5 presents differences between the contact area, SPD coeffi cient and average contact pressure of seats loaded with the force of 300 N in accordance with diagram 1 developed as a result of comparison of individual design solutions.It is evident that in the case of the NP construction, the contact area was the largest and amounted to 443 cm 2 , whereas for BS and BP constructions, it was the smallest and amounted to 371 and 373 cm 2 , respectively.In addition, average contact pressure on the NP seat reached the lowest value of  A change in the method of the gas spring column support from fi xed to articulated (AP type) results in a discernible improvement of discomfort coeffi cient D. In such case, for type of load 1 and 3, the D coeffi cient reached the value of 69.3 daN/m 4 , for the load type 2 D=88.1 daN/m 4 and for load type 4 and 5 -162.9 and 110.5 daN/m 4 , respectively.Moreover, it can also be noticed here that the articulated seat support, as in the case of BS and BP constructions, did not cause a signifi cant comfort improvement.In the case of the BP construction for the load type 1 and 2, coeffi cient D reached the values 112.5 and 100.7 daN/m 4 , for the load type 3 -D=95.6daN/m 4 , whereas for the type of load 4 and 5 -178.7 and 147.5 daN/m 4 , respectively.In order to arrange the examined seats into groups characterized by similar properties, statistical analysis of clusters was performed and Figure 8 presents the results of this analysis.The analysis was performed with all types of construction, all loading types as well as four above-mentioned criteria of seat quality assessment.It is apparent from this Figure that two basic construction clusters were formed.The fi rst cluster, characterized by large linkage distances, was formed by the following types: AS4, BP5, AP4, BP4, BS5 and BS4.Large link-6.13kPa.These stresses for the BP type seat were the highest reaching 10.28 kPa, while for AS and BS constructions -9.33 and 9.41 kPa.Despite favorable sizes of contact areas and average contact pressure for seats of NP design, in this case the SPD coeffi cient reached the highest value of 4.0 %, while its values for AP, AS, BS and BP constructions were determined at: 3.0, 2.7, 2.7 and 2.4 %, respectively.In this situation, this means that despite attractive values with respect to the contact area and contact pressure, the seat in the NP construction revealed the highest unevenness of pressure distribution.Therefore, Figure 6 illustrates the impact of load schemes on the quality of these seats.It is clear from Figure 6 that for load schemes 2 and 3 causing defl ection of the gas spring column to the right or left, respectively, changes of the contact area, values of average contact pressure as well as of the SPD coeffi cient were small.On the other hand, for load schemes 4 and 5 causing defl ection of the gas spring column forward, respectively, to the right or left, the contact area decreased, average value of contact pressure increased and the SPD coeffi cient decreased.In comparison with the load schemes 2 and 3, the value of the contact area, SPD coeffi cient and average contact pressure for the load schemes 4 and 5 changed by: -19 %, -21 %, +19 % as well as by -15 %, -9 % and +10 %.

REZULTATI
This regularity appears to indicate that the NP construction favors comfortable sitting since it supports better the user's body during different positions adopted while working.

RASPRAVA
The above observations illustrate well the calculated values of the discomfort coeffi cient D.   age distances indicate greater differences between individual types in the cluster.The second cluster was formed by the remaining types.In this group, linkage distances were considerably smaller.On this basis, the second cluster analysis was conducted and four basic groups of similarity regarding comfort were distinguished (Fig. 9).It is evident from Figure 9 that the most important and decisive factor affecting the allocation into individual clusters was the discomfort coeffi cient D. The fi rst cluster comprised the following construction types: AS1, AS2, AS3, BP3, AP1, AP2, AP3, BS1 and BS3; types NP1, NP2, NP3, NP4 and NP5 were allocated to cluster two; cluster three included the following types: AS5, BP1, BP2, AP5 and BS2 and the last, fourth cluster consisted of the following types AS4, BP4, BP5, AP4, BS4 and BS5.The fi rst cluster is dominated by constructions with a fi xed linkage of the seat and/or fi xed linkage of the gas spring column.The second group is made exclusively of NP type construction characterized by a fi xed support of the seat and articulated support of the gas spring.The third group is formed by miscellaneous types of construction but within the range of loads 1, 2 and 5.The last, fourth cluster is made of AS, BP, AP and BS with loads of type 4 and 5. On the basis of the present analysis of aggregations, it can be concluded that NP seats from the second cluster turned out to be the most uniform with respect to sitting comfort.This similarity refers to all load schemes.

Figure 1 Figure 2 Figure 3 3 .
Figure 1 Examples of offi ce armchairs: a) a seat with articulated support; gas spring with rigid support -type BS; b) a seat with rigid support, gas spring with articulated support and with possibility of regulation of the defl ection angletype AP; c) a seat with rigid support; gas spring with articulated support but without possibility of regulation of the defl ection angle -type NP Slika 1. Uzorci uredskih stolaca: a) sjedalo s gibljivim postoljem; zračna opruga s krutim postoljem -tip BS; b) sjedalo s krutim postoljem, zračna opruga s gibljivim postoljem i s mogućnošću regulacije nagibnog kuta-tip AP; c) sjedalo s krutim postoljem; zračna opruga s gibljivim postoljem, ali bez mogućnosti regulacije nagibnog kuta -tip NP

Figure 4
Figure 4 presents the distribution of seat contact pressure, when the seats are loaded in accordance with diagrams 1, 2 and 4. It is evident from this Figure that the smallest pressure occurred on seat surfaces of NP type design.At the same time, it should be observed that in the case of diagrams 2 and 4 causing defl ection of the column and seat to the right, greater pressure developed on the left side of the seat.In addition, the pressure area exerted by the right thigh was larger than the area of pressure exerted by the left thigh.Figure5presents differences between the contact area, SPD coeffi cient and average contact pressure of seats loaded with the force of 300 N in accordance with diagram 1 developed as a result of comparison of individual design solutions.It is evident that in the case of the NP construction, the contact area was the largest and amounted to 443 cm 2 , whereas for BS and BP constructions, it was the smallest and amounted to 371 and 373 cm 2 , respectively.In addition, average contact pressure on the NP seat reached the lowest value of

Figure 4 4 . 4 Figure 5 5 .
Figure 4 Distribution of contact pressure on the seat surface under loading of type 1, 2 and 4 Slika 4. Raspodjela kontaktnog pritiska na površini sjedala pod opterećenjem tipa 1, 2 i 4 Figure 7 presents the infl uence of the seat type construction and load on the D coeffi cient values.It is quite apparent from this Figure that the NP construction ensured the lowest values of the D coeffi cient and, hence, the highest comfort for the user.For the type of load 1-3, the D coeffi cient attained values ranging from 33.1 to 34.4 daN/m 4 , whereas for the type of load 4 and 5 -58.1 and