The strengths of this design greatly outweighed its weaknesses.  All established tasks and milestones were completed.  The main strengths are listed below:

• Effectively collects high speed flow rate data
• Able to collect preliminary data
• Overall Mechanical System Integration
• Flexibility for Array Setup and Turbine Type
• Streamlined Design

The weaknesses of the mechanical system inhibited certain performance characteristics.  The weaknesses summarized indicate the major impacting factors on data collection and system functionality.  The main weaknesses are listed below:

• Ineffectively collects flow velocity data
• High friction losses within system
• Misalignment
• Torque Data Skew from Motor Assembly
• Motor to Upper Shaft Coupling
• Fabrication
• Welds on Struts to Cylinder
• Cylinder Distortion
• Inner Threading of Pipe

The strengths of this design greatly outweighed its weaknesses.  All established tasks and milestones were completed.  The main strengths are listed below:

• Effectively collects Torque, Rotational Speed, and Frontal Force data
• Overall Electronic System Integration
• User Friendly Interface with LabView Program
• System Organization

The weaknesses of the electrical and data acquisition systems inhibited certain performance characteristics.  The weaknesses summarized indicate the major impacting factors on data collection and system functionality.  The main weaknesses are listed below:

• Lack of Power Supply and Rheostat Reproducibility
• Wiring Short and Instrumentation Overload
• Communication Failure
• Equipment Mislabeling
• Multi-voltage Power Supplies to Instrumentation
• Exposure to High Humidity Environment
• Wiring Containment Box not Watertight
• Power Supply Not Connected to Carriage
• Over Sizing of Load Cells

FUTURE IMPROVEMENTS

Despite the successful operation of the mechanical system, there were numerous opportunities for improvement.  The upper and lower drivelines contained significant friction.  This was largely due to misalignment in the motor, linear bearings, and motor coupling.  To fix the problem, the upper driveline bearings could be replaced with deep groove ball bearings or be better aligned.  The coupling on the motor could be redesigned to reduce motor run out.

The gimble contained a significant amount of play between the motor and the load cell joint.  This could be reduced with a plastic insert to reduce vibration and yield more defined torque results.

The carriage movement on the tracks is not smooth or uniform.  This results in high variation in all data types.  To improve this, open linear roller bearings could be installed as a replacement track.  The drive system could also be redesigned to balance driving forces on the carriage.

The drive system on the tow tank, consisting of a 7.5 hp electric motor, variable frequency drive (VFD) does not function.  After the first data run of the second experiment, the drive system failed.  This situation should be rectified by fixing/replacing the motor and/or VFD.

The mechanical motion produced by the turbine was translated to the upper driveline via a chain.  Though efficient, this extra step in the data collection process, induces uncertainty in the form of vibrational and frictional losses. Also, with the lower housing open to the intrusion of water, the chain would carry water up the struts, which resulted energy losses from lifting the water and cause instrumentation be exposed to water. To overcome this issue, the housing of the lower driveline could be sealed.  This sealed housing would contain inline torques sensor and slip ring. 

Despite the successful operation of the data acquisition system, improvements could be made in several areas. Using a stopwatch to measure the velocity of carriage produced consistent results during testing, however the team originally planned on using a second encoder mounted on the carriage drive system to measure the velocity of the carriage.  Knowing the diameter and angular velocity of the drum would make it easy to calculate how quickly the cable connected to the carriage was moving, producing the carriage velocity.  This hasn’t yet been implemented due to the signal degradation that would likely occur attempting to connect the moving parts on the carriage and the stationary drive train encoder to the same DAQ.  A system already exists to run cables to the moving carriage, and a plan is in progress for keeping the laptop in the control room (rather than on the carriage) and running an Ethernet cable from the DAQ to the laptop.  With the laptop now in the control room, connecting the drive train encoder would be a simple matter.

Another area of concern was the electronics box.  The team has improved accessibility to the electronics in the box by building a tilted stand, but due to the wires coming out of the box, it cannot be fully sealed and protected.  This problem could be solved by drilling a hole or two in the wall of the box for the wires to exit, then using quick connects to sealing up the extra space.

Both of the power sources simply sat on top of the carriage, which could potentially be dangerous.  There was little chance of them sliding into the water, but it was possible.  Ideally, it would be best to have power sources that fit into the box with the other electronics.  Steps should be taken to ensure they are securely attached to the carriage.

The rheostat which controlled the motor was far from desirable.  An investment in some type of electronic controller for the motor would likely improve that part of the system considerably.

While the instruments and signal conditioning cards selected successfully measured the desired data, accuracy might be improved by scaling down a little.  The load cells especially seemed to register readings well below the full scale values.  On the positive side, with the current instrumentation, turbines of more than ten inches in diameter (the size of the turbine tested) could be evaluated without having to change the electronic and data acquisitions systems.