10 November 2001



Prof. Richard Aynsley


This is a critical time in Australia’s efforts to achieve energy efficient buildings and reduce its greenhouse gas emissions.  Australia needs verifiable scientific methods for determining realistic R-values for external envelopes of buildings.  Current test methods fail to adequately account for radiant heat transfer or the high temperatures commonly experienced in roofs in Australia. One reason for the inadequate accounting for radiant heat transfer in roof spaces is the difficulty posed by the geometry (view factors) of roof spaces.  The Australian home owner deserves reliable R-values for insulation products derived using a common method for all products.  Now is the time to deal with these issues with some good science.


United States encountered similar problems some years ago, including R-value feuding among insulation manufacturers. To resolve the issue of radiant heat transfer in roofs, USA built the large-scale climate simulator at the Oak Ridge National Laboratories in Tennessee. A full-scale residential test module can be accommodated inside the climate simulator chamber.  I had the opportunity to inspect this ORNL facility on March 14-15, 2001, and to discuss many issues with Dr. Desjarlais and Dr. Wilkes.  ORNL is about two and a half hours drive north of where I work in Marietta, GA.


Probably the most valuable feature of this large-scale climate simulator facility is its ability to simulate dynamic climatic conditions, including radiation, which in the US follow the Typical Meteorological Year (TMY).  This allows direct comparison of heat fluxes from dynamic physical simulation with output from the computer simulation software ATICSIM written by Dr. Ken Wilkes.  It uses TMY climate data (Petrie, Wilkes, Childs, and Christian, 1998).  This software was validated against the large-scale climate simulator data and is now the ASTM C 1340-96 method for determining R-values in roofs.


It is extremely difficult to validate dynamic computer simulation software such as NatHERS etc. from field study data. The beauty of the large-scale climate simulator is that the facility can determine the dynamic performance of all types of construction with any type of insulation under a very wide ranges of climatic conditions. A large-scale climate simulator offers the Australian insulation industry a level playing field; a single method for determining R-values for any form of insulation or envelope construction under Australian climatic conditions.


Impediments to progress:

·      Feuding within the thermal insulation industry

·      Inadequate standards for establishing R-values and testing thermal insulation products

·      Difficulties in accommodating radiant heat transfer through roof spaces in estimating R-values

·      Difficulties in accommodating effects of air change rate through roof spaces in estimating R-values

·      Lack of a common method for testing thermal resistance of bulk and reflective insulation products, or combinations of these materials.

·      Australia does not have a large-scale climate simulator such as that at ORNL in the USA which was established to address the issues listed above.


Benefits of full scale thermal simulation approach:

·      Full scale simulation avoids scale effect issues associated with radiant heat transfer

·      Physical full scale simulation reproduces roof space geometry influences on radiant                   heat transfer

·      The same Australian month-by-month, hourly climate data used in computer models such as NatHERS can be simulated physically in the large-scale climate simulator, including radiation and roof space ventilation rates.  This avoids difficulties associated with trying to compare output from computer models with data from field studies.

·      Provides opportunity to validate computer modelling against physical simulation (not just other computer simulations).

·      Provides a common independent means for establishing overall R values for both                       reflective and bulk insulation products (or combinations) and envelope construction.

·      Validated ORNL computer heat transfer software which runs under DOS is available now in Btu/°F units.  ORNL is interested in having the software rewritten by a third party to run under Windows using dual British/SI units.



Petrie, T.W., Wilkes, K.E., Childs, P.W., and Christian, J. E. (1998) Effects of radiant barriers and Attic ventilation on residential attics and attic duct systems: New tools for measuring and modeling. ASHRAE Transactions., Vol.104, Part 2, Paper number TO-98-20-1, pp. 1175-1192.


(Note: this paper includes a detailed description, and diagrams, of the ORNL large-scale climate simulator.  Under severe summer conditions simulated in the tests, the average cooling benefit from the radiant barrier was 34% with ducts in the attic and 29% with no ducts in the attic.  The computer model ATICSIM uses attic geometry and thermal properties of the attic together with hourly climatic data [Typical Meteorological Year] to calculate the heat flux through the ceiling).


Richard Aynsley B.Arch(Hons I), MS (Arch.Eng), Ph.D.

Dean, School of Engineering Technology & Management

Southern Polytechnic State University,

1100 South Marietta Parkway, Marietta, GA. USA

Tel. 1 (770) 528 7205, Fax 1 (770) 528 7134

email <raynsley@spsu.edu>


Adjunct Professor of Tropical Architecture

James Cook University, Townsville, QLD.