No_1_January_1969 — Page 35

Fig. 3. Experimental building and masts under construction

Uncoupled, the sections will be tested for sidesway under seasonal winds of moderate velocity.

Special dynamometers, to be in- stalled across the vertical joints at selected floor levels will record the drag effects of wind between the un- coupled sections. Specially designed flexible joints have been incorporated in the cladding to accommodate any differential sway movements of the component building sections and, in addition, any part of the cladding may be temporarily disconnected from the structural frame.

Wind velocities will be recorded by means of special quick-response gust anemometers, developed by the Elec- trical Research Association of Great Britain and attached at fixed height intervals to a set of free-standing, lat- ticed-steel masts, 175 ft. tall, located about 200 ft. away from the nearest building face. Collectively, these anemometers, approximately 60 in number, will be used to determine the wind spectra and spatial correlation.

Wind pressure distribution over the four faces of the building will be mea- sured by 72 pressure gauges developed by the Building Research Station at Watford, UK. These will be flush- mounted in the glass wall panels at pre-selected points.

In order to measure the sway move- ments and vibration characteristics of

the building special electro-optical tracking instruments are being used. These track the paths of constrasting colour targets attached to various parts of the building and are capable of remotely recording vibration character- istics at distances ranging from about 2 ft. up to 1,200 ft.

To handle the vast amount of data which must be recorded, a special data logger is being installed at ground-

Fig. 4. Plan of building section

floor level, in an air-conditioned and electrically shielded instrument room. This accepts up to 240 channels of low-level analogue inputs in the range +10mv to ±500mv full-scale deflection, time multiplexes the data, makes an analogue-to-digital conversion, and re- cords the binary or binary coded decimal equivalent on in. wide, 9- track magnetic tape. Operation of the system, which is now being installed is illustrated in Fig. 5.

The analogue signals are sampled sequentially at the rate of 10 samples/ second/channel by a reed relay multi-

plexer followed by a solid state sub- multiplexer which also performs the function of amplifying the low-level signals to ± 10 volts f.s.d. for maxi- mum A-D converter resolution. No arithmetic is performed within the sys- tem; therefore all data indicated or recorded will be a function of the analogue signal level and amplifier gain.

Identification data is included in the information recorded on the magnetic tape to provide the means of knowing which groups of channels have been selected for the scan se- quence. This identification data also serves to identify the gain setting of the amplifier, as an individual amplifier gain is permanently associated with a particular channel group.

Analogue data is converted into a 9-bit binary two's complement digital format by the conventional method of successive approximation, each bit conversion occupying a time of 1.5 microseconds. The system accepts a continuous stream of data during the entire data acquisition and recording process, i.e. the scanning of the input channels is a continuous operation.

While interlock gaps are being gen- erated on the magnetic tape, in ac- cordance with IBM System/360 format requirements, digital data is stored in a buffer core store. The store is an Ampex RF-2 of size 4,096 words x

Far East BUILDER, January 1969.

39

Page 35Page 36

Filter-40db.50c/s

1

F

Differential Amp. & Active Filter

Solid State