Customer Needs Assessment:



Staff | |

|Team Member |Discipline |Role |

|Wayne Walter |ME |Faculty Guide |

|Erin Gillespie |ME |Consultant |

|Anthony Squaire |ISE |Project Manager |

|Alan Mattice |ME |Lead Engineer |

|Brian Bullen |ME |ME Support |

|Cody Ture |ME |ME Support |

|Charles Trumble |ME |ME Support |

|Jeff Cowan |EE |EE Support |

|Aron Khan |EE |EE Support |

|Andre McRucker |CE |CE Support |

Open Architecture Thruster Design for Underwater ROV

Concept Development Review

Project 08454

Introduction

The primary objective of the Underwater Equipment Technology project is to take the underwater ROV that was constructed by the Underwater ROV Project 06606 during the 2005-2006 Senior Design sequence and expand its capabilities. Currently, the ROV is submersible to a depth of 400 feet of water and is equipped with a digital video system which allows the user to search for or observe underwater objects. The current projects will focus on improving lighting, thrusters, and equipment housing.

The mission of this team will focus on improving propeller shape and selecting other components to create an optimized thruster. The team will use hardware and software that is very similar to that used with the scaleable land-based robot and the lighting system. The team will also use the same housing to enclose the thruster system that is used to enclose the lighting system.

The figure below is a sub function flow diagram depicting how the team plans to engage the thruster project.

[pic]

Customer Needs

The establishment and clarification of customer needs is a very important part of the concept development process. These customer needs were developed by the Thruster Team after meeting with their customers Dr. Hensel, Dresser Rand, and Hydroacoustics. The customer needs were then translated from the voice of the customer to engineering metrics and specifications stating what the team will attempt to achieve in order to satisfy all customer needs.

Customer Needs Assessment:

1. Thruster must have powerful motor with high torque

2. Thruster must have long operation life per battery charge

3. Thruster must be easy to mount on the ROV frame

4. Thruster must use standard (off-the-shelf) fittings and connections

5. Seals able to handle underwater depth pressures of at least 400 ft. (180 psi)

6. Thruster is able to travel in the reverse direction

7. Thruster gives off very little vibration when being operated

8. Thruster must be operational in at least 400 ft. (180 psi) of water

9. Thruster must be less than or of the same weight as the Seabotix thruster currently used

10. Thruster must have a volume comparable to the Seabotix thruster currently used

11. Thruster must be operational in temperatures that range from 32-80º Fahrenheit

12. Thruster must be modular

13. Thruster must be open source and open architecture

14. Thruster must comply with federal, state, and local laws

15. Thruster must comply with RIT policies and procedures

16. Thruster must be durable

17. Thruster must have a lifetime of at least 168 hours

Specifications

After defining the customer needs, the needs were then translated into detailed statements and specific values that the Thruster Team collectively agreed on. The specifications developed for the thrusters are listed below. All specifications will attempt to be satisfied, however some tradeoffs might occur due to time, space, and money constraints.

[pic]

Propeller Concepts

Choosing a type of water propulsion device falls to what the main purpose of the pump will be. An axial flow device takes in water and accelerates it along a perpendicular axis to the rotation of the impeller. A centripetal pump intakes a high velocity flow through a small inlet and uses rapidly expanding flow corridors to decrease velocity but increase the pressure of the flow to pump is large distances (vertically or horizontally). A magnethydrodynamic engine would use electrodes to create magnetic fields in the fluid thereby repulsing or attracting each individual water molecule. If the fields are pulsed in a certain direction, then the molecules would then be pushed in that same direction creating thrust.

[pic]

Based on the Pugh’s Matrix and the concept descriptions above, the only feasible type of propulsion device for the small ROV application is the axial flow pump. This will use a small motor to turn a set of blades in the fluid, creating thrust. A centripetal pump is used for applications such as sub-pumps and supply lines of municipal systems where water needs to be raised a certain distance. The magnethydrodynamic engine is just not feasible for the small-scale ROV applications because this method is far beyond the scope of a 22-week project. The amount of energy that is surely needed to created and pulse the magnetic fields would most likely hamper the energy consumption of the ROV.

Given the scope of the project and the time allotted for in-depth research, it was decided that designing a propeller from the ground up would be impractical. Consultation with the customer and members of the original ROV team revealed that thruster manufacturers will often choose propeller geometries empirically. In looking at the propeller designs of the Seabotix and Tecnadyne thrusters, both resembled geometries typically used to move air rather than water. This seemed a natural choice given that after being geared down, the motors only spin in the neighborhood of 800 RPMs. With such a low speed, an aggressive propeller design was best. Propellers designed to move air are easier to find in a variety of geometries, materials, and size and are cheaper and easier to modify.

[pic]

Muffin Fan Impeller

• Comes in an assembly in which most parts are not needed

• Curved ends allow for best curve fit along shroud inner diameter to reduce tip vortices

• Hard plastic resistant to corrosion and deformation

• The muffin fan pushes a very high volume and will most definitely overload any motor for our application

• Cost $5.99 per unit with a lead time of one week

[pic]

Fan Impeller

• Very cheap and easy to make: cut and bent out of a sheet of stainless steel

• The thin stainless steel is more likely to corrode or deform over the life of the thruster

• Flat edges allow large tip vortices

• Cost $2.12 per unit with a one week lead time

Motor Concepts

The currently used thruster on the ROV contains a brushed DC motor. The team’s new design uses a Permanent Magnet Synchronous (Brushless DC) Motor. The Brushless motor has an efficiency advantage over the brushed motor with identical torque speed characteristics. The elimination of the commutator decreases the friction caused by the brushes rubbing against the rotor assembly. Brushless motors have an average efficiency of 85% to 95%, where as the brushed models peak at 75% to 80%.

Anaheim Automation BLWRPG170S-24V-4200 Brushless DC Motor

Also, brushless motors have no chance of sparking, an attractive feature when the drive electronics for the motor needs to be in close proximity. Hall Effect sensors are used to provide information about the position of the motor to the drive electronics which then activate the phase winding to achieve the desired direction and speed of rotation. The Hall sensors lends naturally to the implementation of a closed loop control system. While the brushless motor has a tacit mechanical advantage, the electronics needed to drive the motor are more complicated than its brushed counterpart.

[pic]

|  |(Current) |(Proposed) |(Proposed) |(Tecnadyne) |

|Motor Type: |IG42GM |BLWRPG170S-24V-4200 |BLWRPG111S-24V-10000 |RBE(H)-00712 |

|Torque (oz-in) |7.92 |24.07 |5.66 |63.30 |

|Speed (rpm) |5900 |4200 |10000 |12000 |

|Gear Ratio |17 |4.90 |14 |6 |

|Power (W) |34.70 |25 |41.88 |124 |

|Voltage (V) |24 |24 |36 |150 |

|Run Current (A) |1.45 |1.04 |1.16 |0.83 |

|Geared Torque (oz-in) |112.49 |117.94 |79.30 |379.80 |

|Geared Speed (rpm) |325 |857.14 |714.29 |2000 |

|Efficiency |1053.56 |4043.76 |1352.44 |6125.81 |

The only two feasible motors of this list are the first two: IG42GM and BLWRPG170S. Both of those motors run at 24V, while the second BLWRPG motor runs at 36V and the Tecnadyne motor runs at 150V. From the assertion that the motor needs adequate power with a decent amount of revolutions to spin the impeller, the BLWRPG170S-24V-4200 motor is the only choice as its meets the torque of the current motor used by the land vehicles but has roughly 2.5 times the revolutions per minute. The BLWRPG170S-24V-4200 motor will cost $86.60 per unit and the Brushless DC Motor Driver will cost $7.50 per unit. Both items have a lead time of three weeks.

Sealing Concepts

To seal the cap on the enclosure:

One end of the thruster enclosure will be separable from the enclosure to allow servicing of the motor inside. The cap of the enclosure will have a step machined around the edge that will line up with the machined flat surface of the enclosure. There will be a rubber o-ring lined around the step of the cap to ensure a water tight seal. This method was chosen because Dan Scoville at Hydroacoustics currently uses this method and recommends this as a cost efficient and reliable seal. The material of the o-ring will likely be nitrile because the material has excellent resistance to oil and is considered inexpensive.

[pic]

Picture of step around edge for the cap

The seal between the motor shaft and enclosure:

A rubber o-ring will not be a suitable seal because the shaft of the motor will wear the o-ring quickly when it is rotating and lead to a leak. We considered using a rubber and/or plastic shaft seal, but the reliability of this seal greatly depends on the dimensions and assembly of the motor shaft, enclosure and the seal itself. We also considered a hydraulic piston seal since the thruster enclosure is expected to handle at least 200psi. And since the enclosure will be filled with oil, a failed seal will lead to an environmental hazard. To decide on a seal, we checked what our competitors use. Tecnadyne uses a magnetic coupling on their units to transfer the rotation of the motor to the propeller with high reliability. This coupling also allows the motor to continue to rotate inside the enclosure when the propeller is caught with debris. This saves the motor from failure from heat build-up if it were physically connected as one piece.

[pic]

Picture of disk type magnetic coupling

There are several different types of magnetic couplings. The disk type is the cheapest with two hubs facing each other. A stainless steel or plastic barrier is required between these hubs to act as the enclosure seal. Using a disk type magnetic coupling is complicated because the magnets on the end of the hubs will corrode if they are exposed to severe conditions. Also, Magnetic Technologies recommended two additional bearings for the shafts on either side to maintain the exact locations of the hubs. We decided to use co-axial magnetic couplings because the inner hub will be sealed in the enclosure with a stainless steel barrier, and the outer hub is already sealed. Magnetic Technologies manufactures co-axial couplings that will be reliable up to 450psi.

[pic]

Picture of co-axial magnetic coupling

The price of a MTC-0.3 coupling and stainless barrier will cost $265 per unit and have a lead time of four weeks. This unit also has a synchronous design which will not allow any slip at any speed unless the propeller is caught.

[pic]

Control Concepts

The motor control is required in order to operate and optimize the performance of the thruster. The control unit needs to be able to do the following:

• Fit in the thruster housing unit

• Have enough PWM channels to operate a BLDC motor

• Use power efficiently

• Operate the motor in both the forward and reverse direction

The selected control unit will be a slave to the current control unit on the Hydroacoustic ROV (ATmega128). The selected control unit will communicate with the ATmega128 through a RS-485 cable through which it will listen for messages that pertain to it. All other messages will be disregarded. In order to operate the motor in both the forward and reverse mode, an H-Bridge chip will be controlled by the control unit. As an improvement from the previously designed project, the motor of the thruster will give feedback to the control unit. The BLDC motor contains a Hall Effect sensor which will send a signal to the control unit containing the speed of the motor. This feedback can be used to increase or decrease the voltage to a motor if the motor is not running at the same speed as the other motors.

Control Unit Selection

Last year’s design team decided to control the motor with a FPGA board (Altera). An alternative control unit to FPGA is a microcontroller. The following table compares FPGA and a microcontroller:

|Motor Control Concepts |Advantages |Disadvantages |

|FPGA |Cheaper |Source code needs to be loaded with every |

| |Better performance |power up |

| |Able to use code from previous design to |Difficult to change design |

| |control motor | |

|Microcontroller |Easy to program |Motor control code has to be written from |

| |Design can be changed easily |scratch |

| |Source code remains stored in memory |Some of the functionality may go unused |

Figure 1.1

Based on the comparison in Figure 1.1, a microcontroller for the control unit is the ideal selection. By using a microcontroller, the source code does not need to be loaded before each use of a thruster. If a FPGA was used to control the motor, this aspect would become a hassle because before every run of the ROV, the source code for each motor would need to be loaded. The greatest benefit of using a microcontroller is that its design can be changed easily. This allows for our design to be modular and scalable for future use by design teams.

Microcontroller Selection

[pic]Figure 1.2

[pic]

Figure 1.3

Figure 1.2 is a Pugh Matrix comparing two microcontrollers (the ATmega168 and the dsPIC) to the reference Altera (FPGA). The ATmega168 received a higher score than the dsPIC mainly due to its lower power consumption and smaller size. The ATmega168 was designed for low power consumption, drawing 250 μA at 1 MHz when it is in active mode. Although the dsPIC has a higher processor speed than the ATmega168, the built-in Hall Effect encoder in the ATmega168 is suited perfectly to control the motor. From the results of the Pugh Matrix and the comparison of both microcontrollers, the ATmega168 is our first choice to use as the control. The ATmega168 will cost $4.00 per unit and the development board will cost $83.79 per unit with a lead time of eight weeks.

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download