Proceedings of the Institute of Acoustics

Measurements of auditorium seats “in the time of cholera” – Challenges in designing concert hall auditorium seats during the pandemic

AK Klosak, Cracow University of Technology, Poland; archAKUSTIK, Cracow, Poland (email: andrzej.klosak@pk.edu.pl)
AC Gade, Gade & Mortensen Akustik, Denmark (email: acg@gade-mortensen.dk)
B Ziarko, Cracow University of Technology, Poland (email: bartlomiej.ziarko@pk.edu.pl)

1 INTRODUCTION

The design process for auditorium chairs often emphasizes aesthetics (upholstery style, shade, and feel), longevity (resistance to fabric wear), comfort (cushion firmness and bounce), and price. Sound absorption attributes aren't typically a priority, so details on sound absorption, for both whole chairs and their parts (textile, cushion), are primarily available for premium products designed for concert venues. In a modest 350-seat musical academy auditorium inaugurated in 2021 in the industrial town of Jastrzębie Zdrój, Poland, the budget for each chair was capped at around 200 Euros. This excluded most chairs with known sound absorption properties provided by manufacturers. The first two authors, in charge of the venue's acoustic design, subsequently collaborated on material choices with a regional chair producer. This producer embraced the task of crafting audience chairs at that rate, ensuring they met strict aesthetic and sound standards. Herein, we offer a concise overview of our insights from this partnership.

2 SEAT SHAPE, FOAM AND FABRIC SELECTION

The development phase started by selecting a fairly narrow chair (55 cm wide) from the manufacturer's collection of existing designs, which featured a wooden (non-absorbent) back and base, a reclining seat bottom, and solid armrests (Fig.1). Given the restricted size of the Jastrzębie concert venue, the objective was to craft a lightly-sound absorbing seat, with potentially high absorption primarily at low frequencies. This adjustment was necessary due to the hall's minimal low frequency absorption, given that vast portions of the walls were lined with smooth concrete [see separate paper at IOA 2023]. In order to investigate the possible influence of the position of the seat bottom when not in use (vertical or inclined), and the size of the 'gap' between the back and base on absorption at low frequencies, two versions of the seat bottom were developed (Fig.2). Architecturally, the seat was intended to be finished in black painted wood, with black 'woven' fabric (Fig.2, right).

Figure 1: Preliminary seat design

Figure 2: Left: two types of seat bottom inclination angle prepared for final measurements of sound absorption of a block of 20 seats in reverberation chamber; Right: fabric “Type 1” used in all seats in those measurements (Rs=115 Pa×s/m).

Figure 3: Influence of the lamination process on the sound absorption coefficient, as measured in an impedance tube, for two types of foams (typical open-cell foam, and a 'mixed' open-cell foam from 'rebond' processing of different types of recycled foams), and two types of fabrics: velour 135 g/m² with 200 Pa×s/m ('Type 2') and woven 354 g/m2 with 2560 Pa×s/m ('Type 3').

The next step was to select the appropriate foam [1]. Typically, low-cost chair manufacturers utilise closed-cell PE foam (injected into moulds), as it is cheaper, easier to shape, and more durable than open-cell foam. Acoustically, closed-cell PE foams have significantly higher airflow resistance (1400 and 6300 Pa×s/m in our case) than open-cell PE foams, so closed-cell foams were dismissed for this project. After further measurements, a softer open-cell foam (~29 kg/m³ and ~300 Pa×s/m – see Fig.3, upper image) was planned to be used for the seat backs, whilst a slightly firmer open-cell "mixed" foam created from “rebond processing” (~63 kg/m³ and ~390 Pa×s/m – see Fig.3 lower image) was designated for the seat bases. Firmer foams are favoured for seat bases due to durability requirements, whilst softer, more elastic foams are necessary for comfortable backrest support. The subsequent step – arguably the most aesthetically important – was the selection of fabric. The architects desired a "wavy" appearance for the chair fabric, but around 10 different fabrics selected by the architects, based purely on visual preference, had excessively high values of airflow resistance when measured (range of 1000 ~ 2560 Pa×s/m). It is worth noting that none of these fabrics provided any information about their airflow resistance in the datasheets, so access to equipment allowing frequent measurements of airflow based on ISO 9053 (part 2 in our case) was vital. Eventually, a slightly more expensive, but highly permeable polyester fabric “Type 1” was found with an airflow of only ~115 Pa×s/m. This fabric, with its "wavy" pattern, was approved by the architects and it was decided that the chairs would be upholstered in this material (see Fig.2).

3 INFLUENCE OF LAMINATION PROCESS

Based on a simple summation of the airflow resistance of fabric and foam, we expected the total airflow of the upholstery to be in the range of 500-600 Pa×s/m, which should give a good absorption of seats upholstery (α_W=0,95 when measured for 50 mm sample thickness). This airflow value was confirmed in measurements with just fabric placed on top of foam. However, the manufacturer decided to use so-called “laminating foam” to connect fabric to foam using a heat lamination process. Connecting fabric to foam is vital for durability. The negative influence of the lamination process was first observed while measuring the absorption coefficient in an impedance tube for two early upholstery variants, shown in Fig.3. It is clearly visible from Fig.3 that the absorption coefficient is greatly reduced from approximately 500 Hz upwards, for the much heavier woven fabric (Type “3”, which likely requires more heat to laminate), than for the much lighter velour fabric (Type “2”). These measurements also show, that the shape of the absorption curve for velour fabric (Type “2”) laminated to "mixed" foam aligns more with its theoretical expected absorption based on airflow resistance, than when the same fabric is laminated to much lighter open-cell foam. Samples prepared for these early tests were small (25×25 cm), so we hoped the final process might display different behaviour. However, as shown in Table 1, the results of airflow measurements taken on 10 final samples (with final fabric “type 1), each approximately 50 mm thick and 120 mm in diameter, cut from several full-scale seat cushion elements, are even more influenced by the lamination process. As can be seen, the lamination process generated a significant increase in airflow resistance, and it’s uneven quality (different heat and pressure?) resulted in a large spread of measured airflow depending on where the sample was cut. This created a fear that, after spending so much time and effort on material selection, we might end up with uncontrolled sound absorption of the final seats, where the absorption coefficient would be linked solely with the amount of heat and duration of lamination, which was certainly beyond the manufacturer's control. Due to the aforementioned concerns, we requested measurements of 20 seats in a reverberation chamber with two variants of backrests (from both types of foam), to see if sample variation and the lamination process were still visible in the final seats and a larger selection of chairs. It was also observed that denser open-cell "mixed" foam was much less prone to lamination influence than the more elastic open-cell one. So from purely acoustical point of view, open cell “mixed” foam is a safer choice when fabric is to be laminated to foam. Thus, replacing the less dense foam preferred for backrests due to audience comfort, with “mixed” foam, was still a possible choice.

Table 1: Airflow resistance measured for 10 samples (50mm thick) cut out from several different seat samples.

4 LABORATORY AND IN-HALL MEASUREMENTS

Before 350 seats were manufactured, a group of 20 seats was tested in reverberation chamber of Silesia Technical University in Gliwice (Fig.4). Five seats 55cm each, were set up in four rows (sample size 288x364 cm). Row to row distance was 90cm. Following the K&K method [2,3] chairs were placed in the room corner and surrounded on two sides with 90cm high plywood screens. Due to time constrains we measured only one unoccupied configuration with and without screens (similarly as in [4]). Two versions of seat bottom were tested (vertical and inclined, as shown on Fig.2). Two types of foam on backrest in three thicknesses (50 mm, 30 mm and 0 mm /just wood/) were also tested. In total 14 configurations were measured, including 4 with audience. Based on those measurements, a variant with a vertical seat bottom, and 50mm open-cell foam on backrest and 50mm open-cell “mixed” foam on seat bottom was used in the hall. The results for this configuration alone are discussed further below.

Figure 4: Seats in reverberation chamber, before installation of screens on two sides.

During final installation of seats in the building, we also measured their’s in-hall sound absorption, and compared it (Fig.5-6) with our earlier lab measurements and Beranek curves [5]. We also used Bradley [6,7,8] P/A correction taken from [8, Table 4] to compare lab measurements to Beranek [5,9] absorption coefficients. What was observed from in-hall measurements (Fig.5), that sound absorption of choir seats (which were installed at choir balcony with slope ~33º) is higher in low frequencies, than sound absorption of other seats located at much flatter parterre (~19º) and flat side galleries. Therefore, sound absorption data from ISO354 measurements of flat seat samples should be cautiously applied to steeper audience areas when calculating reverberation time (RT). As sound field in a concert halls is not perfectly diffuse, placing absorbing audience area on a steep slope, especially high-up in the hall, might reduce reverberation time more than expected, as this introduce absorption into the upper reverberation volume, which otherwise can be used as a “reservoir” for longer RT [10].

Results from laboratory measurements of occupied seats, with added Bradley corrections (shown on Fig.6, blue curve) are consistent with Beranek “lightly absorbing seats” up to 500 Hz. The reason for the lower sound absorption of Jastrzębie seats for higher bands is the lamination process.

Figure 5: Sound absorption of empty chairs measured in reverberation chamber and in hall.

Figure 6: Sound absorption of occupied chairs measured in reverberation chamber.

As we found ourselves amidst the pandemic in April 2021, we were unable to permit individuals to occupy the reverberation chamber for obvious reasons. In order to simulate the sound absorption of a seated audience, we utilised twenty polypropylene mannequins [similarly as in 11], each weighing 5kg and dressed in realistic clothing – dress 20 mannequins was really a refreshing experience!

Figure 7: Our “pandemic-proof” audience!

5 SUMMARY

1. When designing new concert hall seats, full acoustic oversight, including airflow measurements of seat elements (fabric, foam, both), is key to anticipate final product sound absorption.
2. Smaller chair manufacturers often lack understanding of the link between manufacturing and acoustic properties.
3. Process of lamination of fabric to foam reduces airflow and chair sound absorption (above 500 Hz in our case). Could be acceptable in specific venues (like Jastrzębie Hall), but can cause issues elsewhere, especially if fabric is laminated to lower-density foams.
4. Sound absorption coefficient of steep audience areas in final hall, can be underestimated when measured in reverberation chamber for a group of chairs placed flat on a floor.

6 REFERENCES

1. NJ Mills: Polymer Foams Handbook, Butterworth-Heinemann, 2007, ISBN 9780750680691, https://doi.org/10.1016/B978-075068069-1/50009-X.
2. U.Kath, W.Kuhl, ‘‘Messungen zur Schallabsorption von Personen auf Ungepolsterten Stuhlen,’’ Acustica 14, 50–55, 1964
3. U.Kath, W.Kuhl, ‘‘Messungen zur Schallabsorption von Polstersthlen mit und ohen Personen,’’ Acustica 15, 127–131, 1965
4. WJ. Davies, RJ. Orlowski, YW. Lam, “Measuring auditorium seat absorption” J Acoust Soc Am 96, pp.879–888 (1994) https://doi.org/10.1121/1.410263
5. LL.Beranek, T.Hidaka, “Sound absorption in concert halls by seats, occupied and unoccupied, and by the hall’s interior surfaces”. J Acoust Soc Am 104, 3169–3177 (1998) https://doi.org/10.1121/1.423957
6. YJ.Choi, JS. Bradley, DU.Jeong, “Effects of edge screens on the absorption of blocks of theatre chairs”, Applied Acoustics, vol.73, no.5, 2012, pp.470-477, https://doi.org/10.1016/j.apacoust.2011.11.007.
7. JS.Bradley “Predicting theater chair absorption from reverberation chamber measurements”, J Acoust Soc Am 91, 1514–1524 (1992) https://doi.org/10.1121/1.402483
8. JS.Bradley, YJ.Choi, DU.Jeong, D.-U. (2013). Understanding chair absorption characteristics using the perimeter-to-area method. Applied Acoustics, 74(9),1060–1068. https://doi.org/10.1016/j.apacoust.2013.03.009
9. N.Nishihara, T.Hidaka, L.Beranek, “Mechanism of sound absorption by seated audience in halls”. J Acoust Soc Am. 110. (2001) 2398-411. https://doi.org/10.1121/1.1409970 10. AK.Klosak, AC.Gade, “The Penderecki concert hall in Radom, Poland: acoustic design and performance” in Proc. of IOA 8th international conference on Auditorium Acoustics (Dublin, Ireland) vol. 33 (2011)
10. RHC. Wenmaekers, CCJM. Hak, MCJ. Hornikx, “How orchestra members influence stage acoustic parameters on five different concert hall stages and orchestra pits”, J Acoust Soc Am vol.140(6), pp.4437–4448 (2016) https://doi.org/10.1121/1.4971763