The ceiling tiles were made of wet felted starch bonded mineral fibre material with thickness of 15mm, length 1186mm and width 400mm. The nominal density of the tile was stated to be 360 kg /m³. Assessment of a single tile sample indicated a density of 390 kg /m³.

INSTALLATION OF SPECIMEN :

The ceiling tiles were fitted tightly into a one way exposed grid system using acess (shiplap) tiles, cut back and kerfed on both long sides, with square cut ends.

Where the ceiling passed over the dividing wall, fibreglass strips were inserted over the dividing wall, which then were sealed into place using a 5mm to 8mm thick layer of plaster applied to both sides of the wall between the ceiling and the top of the wall. The plaster used had one (1) hour curing time.

TEST PROCEDURE :

A wide band random noise generator with pink noise filter was connected to a power amplifier and four (4) "flat" loudspeakers. The loudspeakers were positioned on the two walls and the floor in a corner of the source room remote from the dividing wall to excite the maximum number of room modes as required by clause 3.1 of the Standard.

A single microphone was used in each room to sample the sound field. It was attached to a straight line microphone traversing system which was moved slowly across the major room diagonal. The microphone output was monitored by means of a one third octave band real time spectrum analyser which was configured to provide a linear average over a time period of 128 seconds, during which time the microphone completed one traverse of the room diagonal.

The background noise was measured in the receiving room and the source room's noise spectrum was adjusted by means of a graphic equaliser to ensure that the received sound levels were at least 5dB above the background level in all relevant test frequency bands, as required by clause 2.8 of the Standard. The average receiving and source room noise spectra were then measured without altering the input level or equaliser settings.

OVERALL PRECISION OF RESULTS :

The precision of the measurements has been determined by measurements made at a number of fixed positions equally spaced along the length of the traversing system. The 90% confidence limits are as follows :

Third Octave Band

Centre Frequency

(Hz)

Average Sound Pressure Level

Difference

(dB)

CAC

43

(dB)