necessary; however, change in output speed, or wave period, was 

 accomplished by other means (see components f and g) . 



(d) A three-shaft input differential gearbox coupled to the 

 transmission by chain and sprocket drive. 



(e) A constant-speed, alternating current (a.c.) main drive 

 motor. This 7. 5-horsepower motor was connected to one input 

 shaft on the differential gearbox by belts and pulleys and sup- 

 plied the output power for the differential. 



(f) Two smaller variable-speed direct current (d.c.) motors, 

 with one connected to each of the two remaining input shafts of 

 the differential by flexible couplers. The rotation speed of the 

 single output shaft of the differential was determined by the 

 difference in speed of the two d.c. motors. The single output of 

 the differential drove the remaining power train to the bulkhead. 

 Output speed, or period, of the wave generator was therefore set 

 by regulating the speed ratio of the d.c. motors. 



Cg) Electrical and mechanical apparatus to control the speed 

 of the d.c. motors. Functional parts consisted of a d.c. tachom- 

 eter and slave resolver (a.c. phase-shifting mechanism) located 

 on the wave generator and a master resolver, an a.c. to d.c. 

 rectifier-amplifier, and a variable-speed drive connected to the 

 master resolver located in the control room. By adjusting the 

 varidrive speed, the master resolver speed was controlled, and, 

 through the slave resolver and amplifier, the speed ratio of the 

 d.c. motors was set. 



The interaction of the generator and control system components, and 

 operating procedures are given in Appendix B. 



3. Wave Period Control . 



In addition to setting an initially correct wave period for a* given 

 experiment, it was necessary to monitor the period throughout each indi- 

 vidual run. The varidrive stability was affected by temperature rise of 

 the unit during a run, but this effect was minimized by running the vari- 

 drive for at least 20 minutes before starting the run. If not allowed to 

 warm up, the drive was adjusted to the desired period within the first 5 

 minutes of the run and was monitored more closely than the normal schedule. 

 For the normal schedule, the period was checked and recorded a minimum of 

 five times during each run, with three checks during the first hour of 5- 

 hour runs and four checks during the first hour of 2-hour runs. After 1 

 hour the drive stability was found to be within the human error of obser- 

 vation. By repeatability tests and the comparison of checks by experienced 

 and inexperienced personnel, conscientious observations rarely varied more 

 than 0.2 second per 20 revolutions of the generator shaft. The procedure 

 to check the period was to count 20 revolutions of the generator crank 

 arms (or 60 revolutions of the speed reference shaft; see App. B) while 



25 



