The Relationship Between Seat Pressure and Comfort

SAE TECHNICALPAPER SERIES

2003-01-2213

The Relationship Between Seat

Pressure and Comfort

Aernout Oudenhuijzen and Koen Tan

TNO Human Factors

Femke Morsch

Haagse Hogeschool, University for Professional Education

Digital Human Modeling for Design andEngineering Conference and Exposition

Montreal, CanadaJune 16-19, 2003

400 Commonwealth Drive, Warrendale, PA 15096-0001 U.S.A. Tel: (724) 776-4841 Fax: (724) 776-5760 Web: www.sae.org

Downloaded from SAE International by Chongqing University, Sunday, April 10, 2016

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, ortransmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise,without the prior written permission of SAE.

For permission and licensing requests contact:SAE Permissions

400 Commonwealth Drive

Warrendale, PA 15096-0001-USAEmail: permissions@sae.orgFax: 724-772-4891Tel: 724-772-4028

For multiple print copies contact:

SAE Customer ServiceTel: 877-606-7323 (inside USA and Canada)Tel: 724-776-4970 (outside USA)Fax: 724-776-1615

Email: CustomerService@sae.org

ISSN 0148-7191

Copyright © 2003 SAE International

Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SAE.The author is solely responsible for the content of the paper. A process is available by which discussionswill be printed with the paper if it is published in SAE Transactions.

Persons wishing to submit papers to be considered for presentation or publication by SAE should send themanuscript or a 300 word abstract of a proposed manuscript to: Secretary, Engineering Meetings Board, SAE.Printed in USA

Downloaded from SAE International by Chongqing University, Sunday, April 10, 2016

2003-01-2213

The Relationship Between Seat Pressure and Comfort

Copyright © 2003 SAE International

ABSTRACT

With rising customer expectations, driver comfort will become more and more important for car manufacturers in distinguishing themselves from others. This is a challenge, since it is difficult to predict comfort, especially in early design stages. Today, comfort can only be assessed and tested very late in the design a construction process (often using prototypes). Potentially, biomechanic software solutions provide a solution for early comfort testing. Hence, the application of this solution will decrease cost and development time. However, before one is able to create such a software solution, one must have insight into the relationship between mechanical parameters and comfort. This relationship was investigated through a series of experiments in a driving simulator. Both mechanical parameters and comfort were measured. The results showed that seat pressure measurements can be used to quantify driver comfort.

INTRODUCTION

Comfort covers many aspects and is, therefore, difficult to predict. The aim of TNO is to enable early automotive virtual comfort testing using digital human models. Today, anthropometric digital human models support comfort testing. However, most underlying comfort models are limited to postural comfort (Oudenhuijzen et al., 2000). The focus of TNO is to add seat pressure comfort to the existing postural comfort models. In order to be able to model these aspects insight has to be gained into the relationship between seat pressure and comfort. This relationship was investigated through an experiment. This paper will describe the relationship between comfort and seat pressure focusing on:

1. The relationship itself, does it exist?

2. What part of the seat contributes more strongly

to seat comfort, the seatpan or the backrest?

3. The best pressure distribution needed for

comfort.

Aernout Oudenhuijzen and Koen Tan

TNO Human Factors

Femke Morsch

Haagse Hogeschool, University for Professional Education

Once the relationship between comfort and seat pressure has been established, it is possible to predict seat comfort. Secondly, the seat pressure data resulting from the experiment can be used for validation purposes of biomechanic human modeling systems.

The relationship between comfort and seat pressure: In automotive research, seat pressure measurements are often used in relation with seat comfort. Most authors agree on the existence of a relationship between comfort and seat pressure (see Table 1) except for Lee, Ferraiuolo (1993) and Reed et al. (1991).

Table 1 An overview of seat comfort experiments involving seat pressure and the relationship found between comfort and seat pressure (+ = a relationship does exist, ? = the article does not mention a solid relationship between comfort and seat pressure).

A relationship between seat pressure and comfort was found Lee and Ferraiuolo (1993) ? Park et al. (1998) + Kamaijo et al.(1982) + Milivojevich et al. (2000) + Uenishi et al. (2000) + Reed et al. (1991) ? Zhao et al. (1994) +

Michida et al. (2001) + Demontis and Giacoletto (2002) + Inagaki et al. (2000) +

Lee and Ferraiuolo (1993) found a correlation which proved to be to weak to be the basis for seat design activities. However, they concluded that the results looked promising for further studies. Reed et al. (1991) concluded that higher pressures recorded were a likely

Downloaded from SAE International by Chongqing University, Sunday, April 10, 2016

cause for discomfort. However, a solid relationship between seat pressure and comfort was not found. The seatpan versus the backrest The backrest is often referred to as being of great importance to enhance comfort and to reduce discomfort. Most authors (see Table 2) only looked for the relationship between seat pressure and comfort for the complete seat. They did not look for a specific contribution of the backrest or the seatpan to seat comfort. Lee, Ferraiuolo (1993) and Uenishi et al. (2000) did not reveal any preference for the seatpan or the backrest. Kamaijo et al. (1982), Zhao et al. (1994) and Michida et al. (2001) revealed no stronger contribution to comfort for the seatpan or the backrest. Only Milivojevich et al. (2000), Park et al. (1998), Demontis, Giacoletto (2002) and Inagaki et al. (2000) revealed comfort differences between the seatpan and the backrest. The authors differed in opinion about the contribution of the backrest or the seatpan to comfort. Milivojevich et al. (2000) found a correlation between comfort scores of seats and the mean lumbar pressure. However, an even stronger correlation was found between the ischial pressure and comfort ratings. Demontis and Giacoletto (2002) created a model to estimate various aspects of driver comfort. A linear regressive model was developed for the factor static comfort. This model consisted of three aspects: postural comfort, stiffness and wrapping (the ratio between the peak and mean pressure in the anterior popliteal zone). The cushion has a higher weight in the regressive models, indicating a higher importance of the cushion compared to the backrest. Park et al. (1998) and Inagaki et al. (2000) showed a stronger influence of the backrest on seat comfort compared to the seatpan.

Table 2 An overview of seat comfort experiments showing a stronger contribution to seat comfort for the backrest compared with the seatpan (? = result is not clear, + = preference for the backrest, - = preference for the seatpan, +/- = no preference for the seatpan or backrest).

The backrest contributes stronger to seat comfort compared with the seatpan

Lee and Ferraiuolo (1993) ? Park et al. (1998) + Kamaijo et al. (1982) +/- Milivojevich et al. (2000) - Uenishi et al. (2000) ? Reed et al. (1991) + Zhao et al. (1994) +/- Michida et al. (2001) +/- Demontis and Giacoletto (2002) - Inagaki et al. (2000) +

Park et al. (1998) concluded that seats rated uncomfortable did not support the hip and the lumbar region properly. Inagaki et al. (2000) found that thigh support and backrest firmness contributed strongly to the comfort assessment.

What is the best pressure distribution? Treaster, Marras (1987) and Tichauer (1978) defined several guidelines, which are to be considered to provide the best fit for a car seat. These guidelines are: • Avoid tissue compression and pressure concentrations;

• Avoid flattening of the lumbar spine by providing support for the lower back;

•

Distribute the load equally on the back, the pelvis and upper legs.

Once these guidelines are being taken into consideration an even distribution will result in a good fit and a high comfort assessment. The question, however, is how to avoid (if possible) pressure concentrations, how to distribute the load equally and how to avoid flattening of the lumbar spine, and how to distribute the load equally on areas of the human body? Milivojevich et al. (2000) support the guidelines mentioned above, they concluded that a uniform pressure generated a higher comfort score. Park et al. (1998) found that seats rated uncomfortable did not support the hip and the lumbar region properly. They also found that the seat pressure distribution for the uncomfortable seats was asymmetric, contrary to the comfortable seats. Kamaijo et al. (1982) concluded the following: “…the quality of the seat can be roughly determined through an analysis of the load distribution in height direction”; “…it is a good seat when the body pressure pattern shows a variance of pressure along the body’s shape around both ischiatic nodes”. Consequently, Kamaijo et al. (1982) opposed the guidelines of Tichauer (1987 and 1978). How to compare several studies? A similarity between the studies mentioned above is that both comfort and seat pressure were measured. However, they differed in: the experimental methods used for the experiments; the measuring equipment; the duration of exposure; the type and number of seats; the number of subjects used and parameters used for data processing. Most experiments were done statically (in a mock-up); some were done dynamically (driving in an instrumented vehicle). Several methods were used to process the seat pressure data. Some authors divided the matrices resulting from the seat pressure measurements, into various areas to interpret the data quantitatively. These areas were mostly determined visually when inspecting the matrices based on anatomic aspects of the human body and mechanical properties of the seats. The resulting areas focused

Downloaded from SAE International by Chongqing University, Sunday, April 10, 2016

mainly on the area around the buttocks, the upper legs and several areas on the backrest. It must be mentioned that the areas differed in size, position and shape amongst different studies. The parameter most commonly used was mean pressure. In some cases a pressure ratio was used (Lee and Ferraiuolo (1993), Park et al. (1998)). This ratio was defined by the total pressure for a certain area of the seat, divided by the total pressure for the whole seat. Evidently, when analyzing the differences between several papers, a standard method to process seat pressure data in relation to seat comfort has not been determined yet. This makes detailed comparisons of results virtually impossible.

To summarize, seat pressure relates to seat comfort in general. It is, however, not clear what part (the backrest or the seatpan) of the seat contributes more strongly to seat comfort. Finally, most authors seem to agree on the need for an even pressure distribution. However, it is not clearly defined what an even pressure distribution is.

METHODS

The methods used will be described in this section. The subjects, the experimental procedure, the independent and dependent variables, the measurement techniques and the techniques for data analysis will be described in the sections below.

SUBJECTS

A homogeneous group of subjects (in stature and weight) was selected for the experiments in order to be able to compare pressure data with comfort ratings for certain areas of the human body. These areas would differ significantly when using an inhomogeneous group of subjects. The main focus of the experiment is biomechanics instead of anthropometry. Hence, anthropometric differences are no contributing parameters. Therefore, the subjects’ stature and weight was close to the Dutch average (stature 1810 mm, weight 81 kg). The subjects were selected based on their driving experience (more then 5 years driving experience and at least 20.000 km each year). Another selection criterion was absence of problems or injury to the lower back and lower and upper extremities. The subjects’ age ranged from 22 to 55.

EXPERIMENTAL TASK AND PROCEDURE

A within subject design was used. The various conditions were rotated between subjects in order to prevent any kind of order-effects.

Prior to each condition the subjects were instructed about their task. They were asked to drive in a normal way. Extreme driving was prohibited. They were instructed to assess comfort.

Each condition consisted of a task driving for one hour. The task contained 8% city streets, 39% country roads and 53% highway (Bubb, 1995). The moving base of the driving was active, providing realistic driving movements. After each condition the subjects completed a questionnaire using a computer. Completion of the questionnaire took approximately 10 minutes.

The TNO Human Factors driving simulator was used for the experiment (see Fig. 1). This simulator enables subjects to drive in a controlled environment. The driving simulator was validated as a tool to replicate on-road driving by Hoekstra, van der Horst (2000), Kaptein et al. (1996), Kaptein, Claessens (1998) and Kaptein, Korteling (1998). Two mock-ups were used for the experiment: one mock-up of a MPV and one of a sedan. Furthermore, the simulator consisted of a computer generated image system, a moving base (the moving base gives the subjects a physical feedback of driving motions (braking, cornering, accelerating and road conditions) and a 3 dimensional sound system.

Fig. 1 The TNO Human Factors driving simulator equipped with the sedan mock-up

INDEPENDENT VARIABLES SEATS

Ideally a seat would be used where several parameters such as geometry, shape, mechanical properties (stress-strain, damping, friction) can be manipulated independently. However, such a seat was not available for the comfort experiments. Therefore, two seats were selected for the experiments. The seats were selected based on their mechanical characteristics: a medium soft seat (Volvo V40 seat) and a firmer seat (a standard BMW 318 seat) were used (Verver and de Lange, 2002). The medium soft seat will be referred to as seat 1, the firm one as seat 2.

TRUNK INCLINATION

Subjects were seated in two different seated postures: with an upright trunk position and a more reclined one. A

Downloaded from SAE International by Chongqing University, Sunday, April 10, 2016

mock-up based on an MPV design (Renault Espace) was used for the upright trunk posture, a sedan design (BMW 318) for the more reclined trunk posture. Consequently, the driving simulator was equipped with two different mock-ups: one for the sedan and one for the MPV. The seats in the mock-ups were positioned with the same Seating Reference Point (SgRP) and seat reference lines (SAE J1100, 1997). The angle between the reference lines for the seatpan and the backrest was 105° for both mock-ups. This angle was fixed in order to obtain comparable seat pressure data. For instance, the pressure on the seatpan and backrest would change when adjusting the angle between the seatpan and the backrest, which had to be prevented. Additionally, the aim of the experiment was not to find the optimal backrest angle, but which seat provides the optimal support given a certain backrest angle.

The angle between the horizontal plane and the seatpan reference line differed between the two mock-ups. The seat in the sedan mock-up was installed in a more reclined position (4°) compared with the design for the BMW 318. The seat, pedals and steering wheel were adjustable in order to assure the constant relationships between the subjects and the mock-up. The pedals were horizontally adjustable over a range of 100 mm. The steering wheel was adjustable in height (50 mm) and depth (80 mm).

DEPENDENT VARIABLES AND MEASUREMENT TECHNIQUES COMFORT

Comfort was measured through questionnaires. The questionnaire was based on the work of Krist (1993), Park et al. (1997, 1998) and Reed et al. (1991). Also, Localized Postural Discomfort (LPD) ratings were included (van der Grinten (1991) and van der Grinten, Smitt (1992)). Furthermore, the subjects rated the firmness, and the support for various areas of their back and buttocks.

SEAT PRESSURE

Seat pressure was measured continuously using pressure mats, with a frequency of 0.5 Hz. The FSA measuring system was used. The dimensions of the mats were 21” x 21”. The sensing area dimensions were 17”x 17”. The mats had a 16 x 16 array sensor arrangement. One pressure mat was attached to the seatpan, another to the backrest. Two edges of the mats were positioned touching the intersection between the seatpan and the backrest. The pressure mats were properly secured to the seat in order to prevent folding and sliding.

In order find quantitative differences between both seats, the seats were divided into different areas (see Fig. 2). The seatpan and the backrest were divided in 20 and 12 areas respectively based on anatomical aspects of the human body, and certain parts of the seats.

Fig. 2 The mean seat pressure on the seatpan and the

backrest for both seats and the areas analysed.

These areas distinguish separate parts of the seats i.e. the cushions for sideways support and the cushions for general support of the buttocks and the back or separations in the cushions. Also, the areas distinguish separate parts of the subject’s body (the area around the ischial tuberosities, the legs and the lower, middle and upper back). This method could be used because of the anthropometric similarity between the subjects: all subjects were average in stature and weight. An ANOVA was used to analyze the pressure data for each area. This was done for the seats, the two mock-ups and the left and right side of the seats as independent variables. In addition, the mean pressure and the corresponding standard deviation of the areas were compared using a posthoc Tukey-test. The results were considered being significantly different with a p<= 0.05.

Fig. 3 The areas selected for the backrest and the seatpan (marked red)

In order to visualize differences between the two seats, certain areas of the backrest and the seatpan were selected. These were the areas with the highest-pressure values and the most contact between the human body and the seats. Evidently, the areas (cells) with a pressure equal to 0 mmHg were taken out. This resulted in two areas: one for the buttocks and one for

Downloaded from SAE International by Chongqing University, Sunday, April 10, 2016

the back (without the sideways support). The mean pressure for each row (1 to 16 and 1 to 10 for the backrest and the seatpan respectively, see Fig. 3) was determined for each horizontal array of pressure cells.

DATA ANALYSIS COMFORT

The subjective measurements consisted of a questionnaire and Local Perceived Discomfort (LPD) measurements. The subjects were asked to rate the comfort for specific areas on the back and the buttocks in a specific part of the questionnaire (see Fig. 4). The mean ratings and the corresponding standard deviations were calculated for each question. The results were analyzed using MANOVA. The MANOVA was carried out on the subjects’ ratings using the subjects, the mock-up (MPV or sedan), and the seats as independent variables. Differences were considered being significantly with a p<= 0.05.

SEAT PRESSURE

The mean ratings for all subjects and the corresponding standard deviations were calculated for each area. A Student test (two-tailed) was carried out to check the differences between the resulting data. The results were considered being statistically and significantly different with a p<=0.05.

Fig. 4 The areas (sideways, lower, middle and upper back) on the back including the buttocks.

RESULTS

Firstly, the comfort results will be given, followed by the seat pressure results. Finally, the relationship between seat pressure and comfort will be described. COMFORT

Both seats were comfortable. However, the overall comfort of seat 1 differed significantly from the seat 2, i.e., an overall lower comfort rating for seat 2. There was no significant effect found for comfort and trunk inclination. Both seats were rated similar in both mock-ups.

The subjects were asked to rate the support provided by several areas of the backrest. The results for all areas of the back showed significant differences between the two seats. The ratings for all areas analyzed were lower (= worse) for seat 2 (see Table 3 and Fig. 4). The largest differences were found for the lower back. Again, no significant effect was found for trunk inclination. As a result, both seats provided the same amount of comfort in both mock-ups for all areas on the back (including the buttocks).

Table 3 The mean comfort ratings for several areas of the back (1= dissatisfied, 5= satisfied). Seat 1 Seat 2

Mean Mean Upper back MPV 3,72 3,00

Sedan 3,76 3,03 Middle back MPV 3,78 3,30

Sedan 3,79 3,09 Lower back MPV 3,06 1,28 Sedan 3,21 1,69

The rating for the sideways support was significantly lower for seat 2 than for seat 1. The rating for the sideways support did not differ between both mock-ups.

Table 4 The mean comfort ratings for the sideways support (1= dissatisfied, 5= satisfied).

Seat 1 Seat 2

Mean Mean Sideways support MPV 3,09 2,61 Sedan 3,18 2,82

The subjects also rated their LPD. Significant differences were found only for the LPD for the lower back. The results showed a higher discomfort level for the lower back in seat 2 (see Table 5). There was no difference for the LPD between the seats in the two mock-ups.

Downloaded from SAE International by Chongqing University, Sunday, April 10, 2016

Table 5 The mean LPD level for the lower back in both seats for both mock-ups (0= maximum discomfort, 5= no discomfort) Seat 1 Seat 2 Lower back MPV 4.4 3.8 Sedan 4.8 3.8

The subjects were asked to rate the seatpan’s comfort for its support of the buttocks, the upper legs, the sideways support, the seat’s width, depth and height. No significant differences were found between the two seatpans for these aspects. The two seatpans were rated identically in both mock-ups. The subjects were also asked to rate the firmness of the two seats. There was no preference for seat 1 (medium soft) or seat 2 (firm).

SEAT PRESSURE

The mean seat pressure distribution differs between the two seats as can be seen in Fig. 5 (see appendix). The differences between the two seats are detected easily. The seat pressure of seat 1 is evenly distributed over a larger area; the seat pressure of seat 2 is higher in partial areas. There was almost no pressure measured for the sideways support of seat 2.

Also, differences were found between the seats in the two mock-ups (MPV and sedan). The mean seat pressure on the seatpan was higher for the MPV. The mean seat pressure on the backrest was higher for the sedan.

Fig. 6 (see appendix) for the seatpan shows clear differences in mean seat pressure:

• The mean pressure for seat 2 in the MPV mock-up

is higher;

• The pressure is more evenly distributed for the seat

1 in both mock-ups;

• The pressure distribution in the MPV mock-up for

seat 2 close to the backrest is lower than for seat 1 seat;

• There is a small difference in the mean pressure in

the sedan mock-up for both seats, the mean pressure is higher for seat 2;

• The pressure is evenly distributed for both seats in

the sedan mock-up;

• The pressure close to the backrest in the sedan

mock-up for the seatpan is lower for seat 2.

Fig. 7 (see appendix) shows clear differences in mean seat pressure for the backrest. The results show: • A similar pressure distribution for the seats in the two mock-ups;

•

A higher mean pressure for both seats in the sedan mock-up;

• An evenly distributed seat pressure for seat 1, seat 2 shows an uneven distribution;

• • A higher mean seat pressure for seat 2;

A clear peak in mean seat pressure for seat 2 and not for seat 1;

• A higher mean pressure for the lower back for seat 1 and lower mean pressure for seat 2;

• A lacking support for the lower back (0 cm to 15 cm from the intersection with the seatpan) for seat 2; •

An overcompensated support for the middle back (15 to 30 cm from the seatpan) for seat 2.

RELATIONSHIP BETWEEN SEAT PRESSURE AND COMFORT

The mean seat pressure results showed significant differences between seat 1 and seat 2. The mean seat pressure for seat 1 was evenly distributed which was not the case for seat 2. A lower comfort was found for seat 2, especially for the backrest. The seatpan was rated to be comfortable for both seats. No significant differences were found between the seatpans of both seats, despite significant differences in mean seat pressure.

The mean seat pressure from the bottom (near the seatpan) to the top (near the shoulders) of the backrest and the comfort ratings for the lower, middle and upper back is shown in Fig. 8 for the sedan (see appendix). The lower and middle back was estimated to be 15 cm high each. The figure also shows that seat 2, providing less support for the lower back, and a rather high pressure concentration for the middle back, was experienced as being less comfortable compared to seat 1. Evidently, the lacking lumbar support and the uneven pressure distribution in seat 2 seat was less comfortable. Seat 1, instead, with more lumbar support and an even pressure distribution was more comfortable. Therefore, one may conclude that an even pressure distribution results in a higher comfort for the backrest.

Sideways support:

Table 6 shows the mean seat pressure and the comfort ratings for the sideways support. The pressure results are mirrored by the comfort ratings. The comfort rating shows that seat 1 is more comfortable compared to seat 2 concerning sideways support.

Table 6 An overview of the mean seat pressure (SP) for the sideways support) and the corresponding comfort ratings (CR) for the two seats in the two mock-ups (comfort rating: 1 = dissatisfied, 5 = satisfied).

SP CR MPV Seat 1 3,8 3,09 Seat 2 0,6 2,61 Sedan Seat 1 9 3,18 Seat 2 3 2,82 Downloaded from SAE International by Chongqing University, Sunday, April 10, 2016

DISCUSSION

The aim of this study was to find a relationship between comfort and seat pressure. A relationship was found between comfort and seat pressure for the backrest. The backrest with a more even pressure distribution gave a higher seat comfort. This relationship was not found for the seatpan. The comfort of both seatpans was similar despite significant differences in seat pressure. It could be that the subjects are more sensitive to discomfort caused by the backrest compared to the seatpan. As a result, differences in properties of the seats’ backrest could be noticed earlier. This finding can probably be explained by the difference of the anatomy of the human back and buttocks. The area around the buttocks contains, for most people, more deformable tissue, contrary to the back. Consequently, the buttocks have a higher potential to compensate for differences between the seats and the buttocks’ shape, contrary to the human back. In other words, comfort awareness for the buttocks is lower due to a higher potential to compensate for shape differences.

Literature both supports and contradicts the above-mentioned results. Most studies have found a relationship between seat pressure and comfort. Most studies also support the need for an even seat pressure distribution. However, not all studies mentioned support our findings for the backrest, some are unclear and two are even contrary to our results. Most authors only looked for the relationship between seat pressure distribution and experienced comfort for the complete seat. They did not look for a specific contribution of the backrest or the seatpan to seat comfort. Milivojevich et al. (2000), Park et al. (1998), Kamaijo et al. (1982), Demontis, Giacoletto (2002) and Inagaki et al. (2000) revealed clear differences between the seatpan and the backrest concerning the relationship with comfort. Table 7 shows several of these seat comfort experiments. It also compares these experiments with our experiments on three parameters:

A. Is seat pressure distribution related to comfort?

B. Are the results similar with our results, namely, that

comfort is related to an even seat pressure distribution?

C. Does the backrest contribute more strongly to

comfort compared with the seatpan?

Most studies have found a relationship between seat pressure and comfort. Most studies also support the need for an even seat pressure distribution. However, not all studies mentioned support our findings for the backrest, some are unclear and two are even contrary to our results. Most authors only looked for the relationship between seat pressure distribution and experienced comfort for the complete seat. They did not look for a specific contribution of the backrest or the seatpan to

seat comfort. Milivojevich et al. (2000), Park et al. (1998), Kamaijo et al. (1982), Demontis, Giacoletto (2002) and Inagaki et al. (2000) revealed clear differences between the seatpan and the backrest concerning the relationship with comfort.

Table 7 A comparison between several seat comfort experiments and the experiment described in this paper (A= a relationship was found between seat pressure and comfort, B= an even pressure distribution is needed, C= a relationship was found for the backrest only but not for the seatpan, + = similar result, ? = unknown, +/- = similar and different result, - = different result)

ABCLee (1993)???Park (1998)+++Kamaijo (1982)+++/-Milivojevich (2000)++-Uenishi (2000)++/-?Reed (1991)?-+Zhao (1994)++/-+/-Michida (2001)+++/-Demontis (2002)++-Inagaki (2000)+++ Is an even pressure distribution needed? There is a common conclusion that an even pressure distribution is needed for seat comfort. However, it is not clear what an even distribution is. In fact, the pressure distribution was uneven for both seats in the experiment. Only, seat 2 showed a clear peak at the middle back and seat 1 showed a higher mean pressure for the lower back compared to the upper back. The pressure was more evenly distributed for seat 1 than for seat 2. The highest comfort ratings were found for a seat pressure distribution with a lumbar support, which is higher or equal to the support for the middle back and with a gradually diminishing pressure starting from the middle back to no pressure for the upper back.

The pressure distribution was not evenly distributed for the seatpan either. The results showed substantially higher pressure under the buttocks. Still, the seatpans were rated to be comfortable. Evidently, an even pressure distribution is not desirable for seat comfort. Seat pressure Finally, the following issue remains: is a low mean pressure preferred over a high pressure or not? The seat pressure data showed similar pressure distributions for the backrests of both seats in the two mock-ups. The only difference was a higher mean pressure for the sedan mock-up. This could be expected because the center of gravity moves aft when rotating the body backwards. More weight rests on the backrest as a result. However, both backrests were rated similarly in

Downloaded from SAE International by Chongqing University, Sunday, April 10, 2016

comfort. Additionally, the seat pressure on the seatpans of both seats differed significantly. The mean seat pressure was higher for seat 2; still the comfort ratings were similar. Additionally, the subjects had no preference for the firmness of seat 1 (the medium soft seat) or seat 2 (the firm seat). Evidently, the mean pressure itself, within the boundaries as tested, does not relate directly to seat comfort. However, there is always a tradeoff to be made. It is possible that seats with firm upholstery will cause discomfort due to a poor fit between the human body and the seat causing pressure peaks on certain areas. These fit problems could be overcome with seat adjustments. On the other hand softer seats, can deflect more easily, are less sensitive to these fit problems. Consequently, an optimal fit is harder to achieve with a firmer seat than with a softer seat, which allows for more deformation.

CONCLUSIONS

This paper focused on the relationship between seat pressure and comfort. It also focused on the fit between the seated driver and the seat. Dynamic aspects, involving real car vibrations, were not taken into account. Dynamic aspects were taken into account regarding the driving task itself. The relationship was investigated using the TNO Human Factors driving simulator as an instrument. This simulator offers the possibility to drive in a controlled environment.

The following can be concluded:

• In line with other experiments, a relationship was

found between comfort and seat pressure for the backrest. Consequently, seat pressure measurements can be used to predict seat comfort. This offers the oportunity to reduce the amount of test time needed for seat comfort tests. Additionally, this relationship can be used to predict seat comfort using applicable human modeling systems;

• The pressure values, being high or low, do not relate

directly to comfort. However, the pressure distrubution relates directly to comfort. The comfort differences were only found for different pressure distributions but not for similar pressure distributions with different pressure values;

• There is no direct preference for a firm versus a

medium soft seat. Both types of seats can be comfortable. However, the human body’s fit is more critical for a firm seat. Hence, firm seats could be assessed being less comfortable due to a poor fit; • The backrest has a stronger contribution to seat

comfort than the seatpan;

• An uneven pressure distribution for the backrest will

be less comfortable. However, an even pressure distribution is not desirable. The pressure distribution for the backrest at the lower and middle back should be similar/constant and gradually

deminishing starting from the middle back to no pressure for the upper back for an optimal seat comfort.

REFERENCES

Bubb, H. (1995). Anwendung von RAMSIS - Cockpitgestaltung in Kraftfahrzeug. DGLT Bericht 95-02

Demontis, S, Giacoletto, M. (2002). Prediction of car seat comfort from human-seat interface pressure distribution (SAE Conference 2002, SAE no 2002-01-0781). Warrendale, PA: Society of Automotive Engineers.

Hoekstra, W. & van der Horst, A.R.A. (2000). New road designs tested in a driving simulator. Proceedings DSC 2000, Paris, 506-509.

Inagaki, H, Taguchi, T, Yasuda, E. (2000). Evaluation of riding comfort: from the viewpoint of interaction of human body and seat for static, dynamic and long time driving. (SAE Conference 2002, SAE no 2002-01-0781). Warrendale, PA: Society of Automotive Engineers.

Kamaijo K, Tsujimura, H, Obara, H, Katsumata, M. (1982). Evaluation of Seat comfort (SAE Conference 1982, SAE no 820761). Warrendale, PA: Society of Automotive Engineers.

Kaptein, N.A. & Claessens, F.M.M. (1998). Effects of cognitive road classification on driving behaviour: a driving simulator study. Soesterberg, TNO Human Factors, Report TM-98-C048.

Kaptein, N.A. & Korteling, J.E. (1998). The effect of image resolution on driving simulator validity. Soesterberg, TNO Human Factors, Report TM-98-A048.

Kaptein, N.A., Theeuwes, J. & van der Horst, A.R.A. (1996). Driving simulator validity: Some considerations. Transportation Research Record 1550, 30-36.

Lee, J, Ferraiuolo, P. (1993). Seat comfort (SAE Conference 1993, SAE no 930105). Warrendale, PA: Society of Automotive Engineers.

Michida, N, Okiyama, H, Nishikawa, K, Nouzawa, T. (2001). A study of drivers; fatigue mechanisms during long hour driving (SAE Conference 2001, SAE no 2001-01-0381). Warrendale, PA: Society of Automotive Engineers.

Milivojevich, A, Stanciu, R, Russ, A, Blair, G.R, Heumen, J.D. (2000). Investigating Psychometric and body pressure distribution responses to automotive seat comfort (SAE Conference 2000, SAE no 2000-01-0626). Warrendale, PA: Society of Automotive Engineers.

Downloaded from SAE International by Chongqing University, Sunday, April 10, 2016

Oudenhuijzen, A.J.K, Verver, M.M, Happee, R, Delleman N.J. (2000). ACHIEVE: An overview of Driver comfort. Report TNO-TM-00-D011, TNO Human Factors, Soesterberg, The Netherlands.

Park, S.J., Lee, Y.S., Nahm, Y.E., Lee, J.W. & Kim, J.S. (1998). Seating physical characteristics and subjective comfort: design considerations (SAE Conference 1998, SAE no 980653). Warrendale, PA: Society of Automotive Engineers.

Reed, M.P., Saito, M., Kakishima, Y., Lee, N.S. & Schneider, L.W. (1991). An investigation of driver discomfort and related seat design factors in extended-duration driving (SAE Conference 1991, SAE no 910117). Warrendale, PA: Society of Automotive Engineers.

SAE 1997, SAE J 1100, Motor Vehicle Dimensions (april 1997).

Tichauer, E.R. (1978). The bimechanic basis f ergonomics. New York: Wiley.

Treaster D, Marras, W.S. (1987). Measurement of seat pressure distributions. Batelle Memorial Institue, Columbus, Ohio

Uenishi, K., Fujihashi, K. & Imai, H. (2000). A seat ride evaluation method for transient vibrations (SAE Conference 2000, SAE no 2000-01-0641). Warrendale, PA: Society of Automotive Engineers.

Van der Grinten, M.P (1991). Test-retest reliability of a practical method for measuring body part discomfort. In Qeuinnic, Y. and Daniellou, F. (eds.), Designing for everyone (London Tayloar and Francis), 54-56.

Van der Grinten, M.P., Smitt P (1992). Development of a pratical method for measuring body part discomfort. In: Kumar, S. (ed.), Advances in industrial ergonomics and safety IV, Proceedings of the anual Ergnomics and Safety Conference, 1992. Taylor and Francis, 311-318.

Verver, M.M, de Lange, R (2002). ACHIEVE: Seat modelling in MADYMO using multi-body techniques. Report TNO-02.OR.BV.037.1/MVV, TNO Automotive, Delft, The Netherlands.

Zhao, L.Q, Xia, Q.S, Wu, X.T. (1994). Study of sitting comfort of automotive seats (SAE Conference 1994, SAE no 945243). Warrendale, PA: Society of Automotive Engineers.

CONTACT

Aernout Oudenhuijzen, TNO Human Factors. P.O. Box 23, 3769 ZG Soesterberg, the Netherlands. E-mail: oudenhuijzen@tm.tno.nl

Downloaded from SAE International by Chongqing University, Sunday, April 10, 2016

APPENDIX

Fig. 5 A typical example for the mean seat pressure for a subject on the seat and the backrest for seat 1 (left) and seat 2 (right). The vertical and horizontal axes at the left and bottom of the pictures represent the cells of the measuring device. The vertical axis on the right of the pictures shows the scale for the seat pressure (in mm Hg).

Downloaded from SAE International by Chongqing University, Sunday, April 10, 2016

120100Pressure mmHG80seat 1 MPV60seat 1 sedanseat 2 MPVseat 2 sedan4020035810131518202325Distance to backrest (cm)Fig. 6 The mean seat pressure for the area of the buttocks of the seatpan for both seats in both mock-ups

353025201510seat 1 MPVseat 1 sedanseat 2 MPVseat 2 sedanPressure (mmHG)41 5035810131518202325283033363848Distance to seatpan (cm.)0

Fig. 7 The mean seat pressure for the area of the back (without the sideways supports) of the backrest for both seats in both mock-ups

Downloaded from SAE International by Chongqing University, Sunday, April 10, 2016

Fig. 8 The mean seat pressure and the comfort ratings for the lower, middle and upper back for the sedan (QE (comfort rating): 1 = dissatisfied, 5 = satisfied, LPD (discomfort rating): 0 = maximum discomfort, 5 = no discomfort), NS = not significant (no significant differences were found).