Thruster Performance

Static thruster testing was performed in order to determine the overall thrust profile, current draw, and saturation characteristics of Ranger NBV's thrusters. A single thruster was mounted on a floating support on the surface of the tank at the NBRF as shown in Figure 2-4. This support was then connected to a precision digital scale on the opposite end of the tank with high strength fishing line.

Figure 2-4 Thruster Test Configuration

Using a closed loop control system, the test thruster was commanded to rotate at a specific angular velocity. Once the thrust level had become constant at this propeller speed, the force reading on the scale was recorded, along with the electrical current passing through the thruster. The current was measured on the motor side of the PWM amplifier. The amplifier was driven by an 18V source.
After the data was recorded at a specific angular velocity a new speed was chosen, and the test repeated. This sequence was repeated at intervals of 65 RPM until the maximum angular velocity attainable by the thrusters at 18V was reached. This data is shown in Table 2-5.

Table 2-5 Thruster Data (Forward)

After this sequence was completed, the thruster was rotated 180 degrees, and the entire test was repeated in the reverse direction. Since the propellers are optimized for forward thrust, the profile in reverse was expected to be substantially different. The data from this sequence of tests is shown in Table 2-6.

Table 2-6 Thruster Data (Reverse)

The small forces shown at 0 RPM were caused by the weight of the line hanging into the tank, along with small water currents perturbing the floats and thruster. The results of the first test are graphed in Figure 2-5.

Figure 2-5 Thruster Performance Data (Forward)

The thrust profile is parabolic from zero to about 700 RPM. At this point the profile becomes almost linear to about 910 RPM where the thrust level has increased to about 9.6 lb. At this point the propeller begins cavitating. Further increases in propeller angular velocity produce little to no increase in thrust. In interesting qualitative observation was that at cavitation velocities the thruster would initially jump to a thrust level beyond those reported in the data. Within a couple of seconds, the force output drop to the levels reported in the data. The initial level was difficult to measure accurately.
The current profile also initially increases parabolically, and eventually transitions into a linear region at about the same angular velocity as observed in the thrust profile. The parabola describing the current draw does not intersect the y-axis at 0 Amps. This is due to the fact that the motor must not only fight water drag, but also the friction of the bearings and seals. Another interesting observation is that after cavitation starts and the thrust output begins to fall off, the current (i.e. torque) required to maintain the desired angular velocity drops with almost the same profile.
Figure 2-6 shows the thrust and current profiles for the same thruster in the reverse direction. The propeller is not optimized to operate in this direction, so reduced performance was expected.

Figure 2-6 Thruster Performance Data (Reverse)

The thrust profile in reverse appears to be parabolic from zero to the point where cavitation begins. Additionally, cavitation begins at about 780 RPM, or about 86% of the speed at which it begins in the forward direction. The thrust level at which cavitation begins is about 5.0 lb. or about 52% of the force in the forward direction. After cavitation begins, the force output is reduced, where in the forward direction it stayed almost constant.
The electrical current profile increases parabolically all the way to the point of cavitation. This parabola describing the current draw also does not intersect the y-axis at 0 Amps due to the friction of the bearings and seals. Another interesting observation is that current draw for a specific angular velocity was similar, and at higher angular velocities even larger for operation in the reverse direction while producing much lower thrust. (Figure 2-7 & Figure 2-8)

Figure 2-7 Comparative Thrust Data (Fwd. & Rev.)

Figure 2-8 Comparative Current Data (Fwd. & Rev.)

This thrust mismatch produces some unexpected problems for closed loop control. If Ranger NBV is attempting to perform a pure yaw maneuver, with the two port x-thrusters rotating forward and the two starboard x-thrusters at the same speed in reverse, the result will be a rotation plus an undesired forward translation since the two forward rotating port thrusters produce more force for the same angular velocity as the reverse rotating starboard thrusters. In closed loop control as the vehicle is constantly applying a little yaw left, and then a little yaw right, and so on, the mismatch will cause the vehicle to slowly translate forward. To help alleviate this problem, the speed of each thruster is scaled back in the forward direction so that the maximum output is the same in both directions. The scale factor was determined by trial and error. Eventually, .69 was determined to produce no translation during a basic attitude hold. This number turns out to be fairly sensitive. For example, a value of .65 demonstrates a noticeable reverse translation.

Figure 2-9 Comparative Thrust with Forward Values Scaled

Although this factor was determined empirically under dynamic conditions, when the static thrust data is compared with the forward thrust velocity scaled back by the factor .69 (Figure 2-9), the two curves match almost exactly. This fact is very interesting because the scale factor was determined under dynamic conditions, while the thruster data above was the static thrust output. This seems to indicate some degree of correlation between static and dynamic thrust profiles. The down side of this scaling of the thruster output is that half of the power in the forward direction is unused. Thus, the maximum force in either direction is (using the British system of pounds as a unit of force) 5 lb. or (SI units) about 22.2 N.