90 



NATURE 



[September 24, 1914 



such a channel proved to be very difficult, and 

 the more important of these difficulties are illus- 

 trated by records of fluctuations obtained. A 

 satisfactory desij^n was ultimately obtained and 

 put to a full-scale test in the new four-foot channel 

 shown in Fi^. 2. The results of this final test 

 were considered to be quite satisfactory, and 

 since that time a three-foot channel has been 

 erected and put into use, whilst a seven-foot 

 channel to the same general design is nearly 

 completed. An almost exact copy of the four- 

 foot channel has been erected at Boston, U.S.A. 



From the figure it will be seen that the channel 

 consists of a square trunk, the part from the left 

 to the centre being the working section in which 

 uniform flow is required. The fan is four-bladed 



large bodies of air without the production of 

 , vortices is the great problem in the production of 

 ' a steady wind-channel. 



Not only has the air-flow been rendered satis- 

 factory, but the aerodynamic balance has been re- 

 designed and rendered suitable for work of the 

 most general description. It is possible, and has, 

 in fact, become part of the regular work of the 

 National Physical Laboratory, to measure all the 

 forces and couples acting on any body, however 

 unsymmetrically placed on the wind. The pre- 

 cautions taken in the setting of the balance are 

 described in some detail in the report, and show 

 that considerable care is essential to success ; the 

 order of accuracy required can be realised from 

 the fact that to obtain the resistance of an aero- 



FiG._2. — General view of the new four-foot wide channel, showinj trumpet intake, working section of the channel, and 

 distributor. From " Technical Report of the Advisory Commiitee for Aeronautics for the year 1912-13." By permission of the 



Controller of H. M. Staii mery Office. 



and similar to an aeroplane propeller, and is 

 situated in the short circular section to be seen 

 just past the steps. This propeller draws air at 

 speeds up to 50 ft. per sec. through the bell- 

 mouth and a honeycomb, past the model, and 

 through a second honeycomb, finally delivering 

 the air into the perforated chamber or distributor 

 seen to the right of the picture. The air leaves 

 the distributor at speeds less than 3 ft. per sec, 

 and the disturbance of the air cannot easily be 

 detected except within a foot or so of the perfora- 

 tions. The first honeycomb has a use not readily 

 suspected; miniature whirlwinds form on the floor 

 of the building, and unless broken up by a honey- 

 comb, the result on the model is a succession of 

 blows at uncertain intervals. The movement of 

 NO. 2343, VOL. 94] 



plane wing correct to i per cent, the wind-direc- 

 tion must be known to within 0*05° relative to the 

 arms of the balance. 



There is one measurement other than those 

 provided by the balance which is necessary before 

 the results can be expressed in absolute units, 

 viz. the velocity of the air-current. Experiments 

 have been carried out on the whirling arm of the 

 National Physical Laboratory, from which it has 

 been possible to fix the constant of a standard 

 anemometer to o'l per cent., an accuracy which 

 makes the uncertainty in absolute value far less 

 than the errors of observation in most experiments 

 for aeronautical purposes. The accuracy of aero- 

 nautical experimental work is often limited by the 

 effects of apparatus for holding the models, and 



