Data Acquisition System
The data system consists of a dual-processor 150-MHz Pentium Pro PC with a 16-bit analog-to- digital converter. It is networked with all other computers in the computer room using a Microsoft Windows NT network. To avoid network interference, a serial cable is used to transfer data between the on-line plotting PC and the data acquisition PC immediately after it has been recorded.
The data system nominally has 6 extra data channels for customer use. Each channel of the data acquisition system has an associated signal conditioner and amplifier. Data are taken and averaged at a sampling rate of 100 samples per second. Data are only taken if the dynamic pressure is on-condition, i.e. if the dynamic pressure is within 1% of the target value. Each test point consists of five seconds of on-condition data. The on-condition criteria can be changed to meet individual test requirements, for example, on-condition only if the velocity is within 1 ft/sec.
Daily Calibration Process
In order to maintain the integrity of the data, a balance calibration is conducted twice a day by hanging check weights at half-scale for each component. The calibration (BALCAL) includes a system calibration (SYSCAL). The system calibration automatically checks and corrects the amplifier outputs and analog-to-digital converter (ADC) using a programmable voltage
standard (PVS). The operator is notified if any portions of the data acquisition system provide outputs that are out of pre-established tolerances.
The data acquisition system moves the model through a preprogrammed set of angles. The list of desired angles can be a series of subsets. An example for pitching the model: -10° to - 2° by 2°, -2° to +2° by 1°,+2° to 32° by 4°. Furthermore, if the actual angle of the model is not within ±0.02° of the target angle, the system can be setup to retry until the actual angle is
within tolerance. The tolerance can be re-defined by the operator.
The model positioning motors move the model at a yaw rate of 1 per second and a pitch rate of 2 per second (for the standard five-inch distance between trunnion and pitch arm attachments)
A model position calibration is conducted for each model. Angles of yaw are measured after determining the yaw = 0° point. Angles of attack are measured by an internal model accelerometer. Each model should be built with a cavity for housing the KWT model accelerometer. See Section Miscellaneous for more details.
If the model cannot accommodate the internal model accelerometer, the angles of attack are determined by an encoder that measures the position of the pitch arm. The encoder reading is correlated to the model angle of attack using a precision inclinometer (±0.01° resolution). The calibration provides angles of attack to within 0.02°. A detailed drawing of the KWT inclinometer is in the Miscellaneous Section, to be used in creating a mounting block for the customer’s model. The customer is responsible for providing a reference surface for placement of the inclinometer.
Inputs to the system may be either full or half bridge resistive transducers or dc voltages. Transducer excitation is normally 5 or 12 volts. The amplifiers are gain switchable in ten steps to give a full scale input from 8 to 245 millivolts. In addition, the amplifiers have a ten step low pass filter with cutoffs from 1/8 to 256 Hz, normally set at 1 Hz.