Resonant Chamber is an interior envelope system that deploys the principles of rigid origami to transform the acoustic environment through dynamic spatial, material and electro-acoustic technologies. Our aim is to develop a soundsphere able to adjust its properties in response to changing sonic conditions, altering the sound of a space during performance and creating an instrument at the scale of architecture, flexible enough that it might be capable of being played. The project is developed through three streams of iterative research and development in both computational testing and full-scale prototype installation: Dynamic Surface Geometries; Performative Material Systems; and Variable Actuation and Response. Resonant Chamber is funded through the 2011 Research through Making Grant, U-M Office of the Vice President for Research, 2011 Small Projects Grant, U-M Center for Wireless Integrated Microsystems, Social Science and Humanities Research Council of Canada Research Creation Grant.

Significant work in the areas of kinetic tessellated architectural systems and variable acoustic surfaces using specific geometry includes David Serero's 2005 "Variable Geometry Acoustical Domes", Mani Mani's 2009 "Tunable Sound Cloud", the current research of Brady Peters, particularly his 2011 "Distortion II" and Eddy Sykes' 2008 "Yakuza Lou" kinetic rigid origami structure. Resonant Chamber advances this work by exploring the application of multi-functional material treatments within a volumetrically variable acoustic space paired with kinetic operation digitally controlled via environmental sensing. 

The faceted acoustic surface is comprised of the composite assembly of reflective, absorbtive and electroacoutsic panels, clustered around an electronics panel that contains circuit controls for linear actuation, electro-acoustic amplification of the distributed mode loudspeaker (DML) embedded speakers and a set of sensing inputs. A single electronics panel may contain enough processing to control four DML speakers, local sensing of acoustic pressure and three sets of linear actuators which in turn controls three flat-folding cells.

Actuation and geometric configurations was predictively modeled in Rhinoceros 4.0 and its plug in Grasshopper and add on Kangaroo to script and simulate the physical relationship between vertices and applied forces. Simulations allowed the models to learn how to predictively arrange panels into optimum acoustic variations based on input constraints such as reverberation time, absorption coefficient, directional amplification and early/late acoustic response. When the physical envelope is deployed within a bounding space, this same software can translate predictive modeling into optimized measured lengths of displacement for actuation within the digital model to PWM signals actuating the system’s kinetic components and positioning its tessellated surface.

The flat-folding attribute of rigid origami allows the change of exposed surface properties of the operating envelope with a predictive number of DOF. These flat-folding cells are geometrically translated by linear actuators mounted parallel to the top of the panel. The linear actuators are controlled through the PWM signal to control the fold angle resulting in a proportion of exposed surface. This approach offers the capacity to modify aesthetic form while simultaneously manipulating the acoustics of a space. The DOF in the system allow for a limited (versus serial) number of actuation points to be coordinated across the continuous surface resulting in customized transformations based on input criteria of acoustic optimization.

Three types of composite panels have been developed: reflective, absorptive and sound generating (via electro-acoustics) in collaboration with consultants at Arup Acoustics who undertook rigorous material based digital acoustic simulations as well as scaled physical panel prototype testing to determine optimal geometry and material characteristics relative to their acoustic performance when combined within the geometric configurations proposed. Solid panel inserts to the rigid frame fastened to the membrane offer a 120 Pa (N/m2) density, optimum for acoustic reflections. This performance factor of is described in terms of density allowing thickness of the panel to be specified relative to the material properties. Porous Expanded Polypropylene (PEPP) acoustical panels provide the acoustic absorption coefficient desired for sound dampening effects and the panel rigidity for milling to the desired geometric configurations for flat-folding origami. The PEPP panels are faced with a perforated surface allowing a minimum 25% exposure for optimal noise reduction with noise reduction coefficient of .85. The rigid panel system includes distributed model loudspeakers (DML) mounted to 9.5mm face panels. This piezoelectric technology allows sound to be produced through the face panels themselves by introducing vibrations through an electro-acoustic exciter. A set of four DML modules may be controlled through the electronics panel for a multichannel control system.

The simultaneous dynamic control of material exposure and geometric configurations through actuated surficial deformation offers a new model for adjusting acoustic environments. Though there are precedents in control of late room response through the use of systems to control height or orientation of a ceiling reflector, the use of a kinetic system to control wavefront curvature, level, and time of arrival of early reflections is as yet unprecedented in acoustic research and in the investigation of robotic architectures and responsive envelopes.

Project Team: Geoffrey Thün, Wes McGee, Kathy Velikov [leads] Lisa Sauvé, Mary O’Malley, Adam Smith, Colin Ripley, David Lieberman [design research collaborators] Katie Wirtz, Ian Ting, Lief Millar [prototyping]; Collaborators: Raj Patel, Terence Caulkins and Dave Rife [ARUP Consultants] Jerome Lynch, Devki Desai, Mike Kane [Embedded Sensing and Wireless Integrated Microsystem Integration]