Heneghan Peng's Grand Egyptian Museum in Giza presented an acoustic conundrum for engineer Buro Happold. But clever detailing served to diminish the 'cocktail-party effect'.
At first sight, the gallery design for Heneghan Peng Architects' Grand Egyptian Museum in Giza seems like an acoustician's nightmare. When it opens in 2011, the museum is projected to receive up to 3,800 visitors per hour. Most will be in groups of 20 to 40 accompanied by personal guides, similar to the Cairo museum's existing system. Cathedral-like dimensions and materials - stone walls and oors - and a large proportion of stone artefacts make for a highly reverberant environment. But the artefacts themselves, in particular the taller statues, columns and display cases, form part of a complex acoustic strategy in the main gallery.
Acoustical engineer Lawrence Hughes of Buro Happold, which has been involved in the project since competition stage, describes the museum's acoustic environment as a sequential experience which parallels the museum-goer's path through the building. After passing through airport-like security at the site entrance, visitors will approach the building's translucent stone facade on foot to experience a grandeur similar to Egypt's temple complexes. They will then pass into an internal courtyard where a travelator will lead up to level three, and one of the museum's panoptic views of the Giza pyramids.
These views were the point of departure for Heneghan Peng's 2004 competition-winning scheme. From this point, visitors will pass through a set of wide double doors - which act as an acoustic and environmental barrier - into the 24,000m 2 main gallery space, where it is estimated that roughly 80 per cent of the visitors will opt for a 90-minute guided tour of the museum's highlights, which include a timeline of artefacts through the Early, Middle and Late Kingdoms of Egypt; Tutankhamen's tomb; and a solar boat from Cheops.
Audio headsets might seem like an obvious solution to this acoustic conundrum, but they were initially rejected by the museum in favour of guides, because of the desire to retain an important source of local employment and also because of a maintenance concern about how to keep more than 3,000 headsets an hour in constant working order. Yet the acousticians were concerned about controlling the 'cocktail-party effect', where each guide would have a tendency to speak 10dB louder than the background noise level, replicating the noisy environment of the existing Cairo museum.
One early solution drew on the cathedral precedent of the pulpit. The architects developed a design for raised niches within the gallery walls where guides could stand to speak to a small group. This was eventually rejected due to lack of exibility.
Another solution looked at the use of tent-like canopies which could recreate the imagery of an archaeological site and provide acoustic absorption, but this approach was ruled out due to issues of dust and maintenance. A variation on the conventional audio headset has now been adopted: guides will speak with microphones which will communicate directly to visitor headsets - supplied and maintained by hotels rather than the museum.
Headsets aside, another critical early decision for the acoustic design was determining an acceptable Reverberation Time (RT) for the main gallery, which would maintain enough reverberation to create a sense of grandeur without compromising the intelligibility of the public-address system. Initial modelling indicated an RT of over 13 seconds - way above the accepted range of 1 to 2. An RT of 2.5 seconds was agreed, but the next challenge was how to achieve this.
Due to the gallery's vast scale and spatial complexity, the design approach was similar to that of a Tube station or airport. The gallery was divided into subsections to define a strategy which could then be applied to the whole, and the design was developed in 'bays' through a process of iterative 3D computer modelling, using CATT (Computer Aided Theatre Tool). Acoustic modelling with the artefacts in place showed that locating key objects to serve as sound barriers at critical points could reduce RT to about 8 seconds through specular diffusion - similar to the refracted reection created when you crush aluminium foil.
The next task was to introduce as much absorbent material as possible to reduce the RT from 8 seconds to 2.5.
Absorbent panels were put at high level in the 'Toblerone' space above the open metal mesh ceiling of the galleries, but this is not where it was most needed and was an inadequate solution. A modelling exercise which looked at the entire surface of the gallery walls showed that all the stone would have to be covered to achieve the desired acoustic control, clearly unacceptable because of the architectural imperative of exposed stone walls.
The solution involves using the void between the stone facing and the concrete wall behind to introduce slabs of blackfaced Isowall. The acoustic design hinges on the depth of that void and the width and spacing of the joints between the stone slabs. Joints had to be kept to a minimum at low level to avoid maintenance problems, but at high level a 'banding' of joints was developed to meet the acoustic requirements.
CATT modelling showed that, in fact, the joints provide little overall absorption, but there was an unanticipated serendipity which Hughes says proved to be 'magical' for establishing the acoustic performance for the gallery. The gaps in the stone, together with the Isowall, were able to absorb low-frequency sound in the 100-125Hz range - the most troubling for acoustic comfort in the gallery. Iterative CATT modelling was used to determine the optimal depth of the cavity and spacing of the joints until the 2.5 seconds RT could be met.
But will it work? There is a possible plan to limit the number of guides and people in any one bay at a given time.
Stephen Greenberg of Metaphor, responsible for the exhibition design, observes that the Egyptian guides are 'extremely clever and will quickly adopt shortcuts and find the best spots'. Much will depend on how the gallery is used, and human behaviour is difficult, if not impossible, to model.