In March 1964, Barnes Engineering Company forwarded a Model IT-2 to the writer 

 for evaluation. On the first flight the pen vibrations were sometimes half the width of the 

 chart paper, and obviously bore no relationship to water temperature. After successful lab 

 tests a second flight produced the same results, and field tests were run to determine the 

 cause. The IRT, recorder, and power supply were set up on a concrete apron twenty feet 

 from the aircraft and out of the propeller slipstream. The sensing head was aimed at the 

 pavement. When the aircraft engine was started the pen began to swing wildly. Range of 

 these vibrations could be controlled by varying engine power. With the engine off electrical 

 circuits had no effect, but radio transmissions caused the pen to go off scale. Interference 

 from a tioick engine was also noted. A turbojet 400 yards distant caused large scale needle 

 fluctuations while taxiing toward or broadside to the instrument, but hardly affected the pen 

 when moving directly away. Improper or incomplete shielding of critical components was 

 reported to have been the source of trouble. 



Tests on the York River Bridge pointed up several problems, some of which have 

 not been explained. During testing the sensing head was secured to an outside railing on 

 the Bridge Tender's house about 100 feet above the river. Angle of view was nearly verti- 

 cal. Check instruments were mounted on a platform 10 feet above the water in the support- 

 ing caisson. During the week the following "types" of days were experienced: 



1. Sunny, sky clear, relative humidity moderate, wind 5-12 mph 



2. Scattered clouds, relative humidity moderate, wind 25-40 mph 



3. Overcast, relative humidity high, wind 20-30 mph 



4. Overcast sky, dense fog in morning, drizzle in afternoon, relative humidity 

 high, wind 3-15 mph 



During the calm sunny day, readings between IRT and check instruments were ex- 

 tremely close. High wind on the second day, however, affected the readings and made the 

 needle "nervous". A large cardboard box fitted over the sensir^ head eliminated much of 

 the trouble and when the sensing head was moved to the lee side of the building, the needle 

 was stable. It is suspected that even gusts of relatively low velocity affect readings 

 noticeably if the sensing head is exposed. Accuracy was high during the overcast period 

 but fog and drizzle on the fourth day caused errors of large magnitude. IRT readings were 

 generally high, an expected result as the air temperature (and thus the temperature of the 

 fog) was higher than the river water. As the area was relatively protected, no high sea 

 states were observed, but large pressure waves passing through the field of view had no 

 apparent effect. A flat calm sea surface caused anomalous high readings attributed to 

 solar warming of the poorly mixed upper layer. 



Further tests were made by tilting the sensing head at different angles (90°, 80°, 

 70°, 60°, 50°). Normally no changes were noted until 60 or even 50 was reached, but twice 

 at 70 the needle jumped more than 10°F. No clue as to cause could be ascertained from the 

 physical conditions. As no fog was visible, it is concluded that sea state may be a factor. 

 The same unexplained variations have been noted during aircraft banking. 



-65- 



