P18463: Water Powered USB Charger

Customer Handoff & Final Project Documentation

Table of Contents

Lightning Talk

The lightning talk is roughly a two minute breakdown of our entire Senior Design project. It focuses on the need for such a device, how the team went about designing and building a product that would meet those needs, and how well the product performed against our expectations.

The lightning talk slide and script can be found below:

Lightning Talk Slide

Lightning Talk Slide

Final Design

River Tested System

The final iteration of the Trick Charger is shown in the color-coded cross-section below. The device is oriented so that the water would be moving right to left, with the cable coming out the right side of the image through the bearing shaft extension. The final housing was 3D printed out of ABS and bathed in acetone for waterproofing and durability. The blades ideally would have also been printed out of ABS, but availability and size constraints restricted us to using PLA. The device housing and modular blades were designed and 3D printed at the Construct and FabLab here at RIT.

Legend Assembly

Legend Assembly

Final Device Picture

Final Device Picture


Assembly Materials:

- Exterior housing

- 2 lip seals

- Interior housing

- Steel interior housing

- Motor

- Cable with sliced end for knotting

- Wire cutters?

- Metal wire shaft

- Metal drive shaft motor extension with set screw

- Silicone Adhesive

- Coffee stirs from MSD office for application!

- Caulk gun

- Paper towels

- Epoxy adhesive

- Drive shaft fin anchor

- Wire shaft fin anchor w/ bearing

- 2 pieces of 5 cm cotton shoelaces

- Graphite lubricant

- Aluminum stuffing body

- Plastic turbine blade holders

- 3 Turbine blades

- 5/64 Allen key

- Aluminum turbine blade plate securer.

Additional Testing Materials:

- Access to pool

- Scale from Apps Lab

- iPhone for slow mo

- Paper towels

Assembly Procedure:

Part I:

1. Thread the cable through the wire shaft fin anchor (w/ bearing) and then the metal wire shaft so that both are close to the sliced end of the cable. The two should be able to be press fit together by hand on the cable.

2. Thread the sliced end of the cable through the top of the exterior housing.

3. Apply epoxy adhesive to the inner diameter on the top of the exterior housing. Slide the metal wire shaft into the exterior housing as far as it can go. Apply additional epoxy and silicone adhesive at the base of the metal wire shaft as well as the metal wire shaft-wire interface.

4. Tie 2 knots at the end of the sliced cable. Apply silicone adhesive on inside surface and give wire a rug from the outside to securely sit the knot at the inside of the top exterior housing.

Part II:

1. Cover both shoelaces in graphite lubricant.

2. Insert shoelaces in lip seals for a snug fit.

3. Lube inside tunnel of stuffing body with graphite lube.

4. Place both lip seals inside stuffing body with the insides facing the aluminum casing from both ends. Apply epoxy of outer edges of lip seals to permanently connect them to the stuffing body.

5. Insert drive shaft motor extension into stuffing body until both lip seals are sitting on the round surface of the shaft.

6. Apply silicone to the inner edges of the three columns of the plastic interior housing and place stuffing body within that housing. Do not push housing past the point where the mating surface of the stuffing body matches top surface of the three plastic columns. Let the silicone set for 20 minutes.

7. Insert metal housing through other end of plastic interior housing. Push motor through steel housing and make sure motor shaft slips into aluminum drive shaft.

8. Align flat head of motor shaft with set screw and tighten using allen key.

9. Apply epoxy to mating surface of stuffing body and stick entire inside assembly into lower exterior housing and let it rest for 10 minutes for epoxy to set.

Part III:

1. Place both o-rings inside their designated locations on exterior plastic housing.

2. Begin screwing both pieces of exterior housing. Put large amount of silicone in between the two pieces. The complete screwing and wipe off extra silicone squeezed out with coffee stirrer.

3. Apply silicone between any place where there is a visible gap in the seal between the two housing pieces.

4. Place the two plastic turbine holders on the ends of both aluminum shafts. Face the openings outwards.

5. Place blades inside the holder openings.

6. Place aluminum plate on driveshaft and push against blade tips. Secure plate with set screws inside the plate with allen key.

7. Epoxy end of cable to the metal cable tube and cover with teflon tape. After waiting for 10 minutes for the epoxy to set, cover with one final layer of epoxy on top of teflon tape. Let it set for 20 minutes.

System Test Results

RPM & Power Production: Without Boost Converter

The system test was run in the Genesee river on a fairly calm day with moderate amount of wind. The water speed seemed to be slightly faster than that in the pool so the turbine blades seemed to spin much faster than they did in the pool. Unfortunately, the device was too far for us to use our sensor or our slow-motion camera method to attain a velocity. But based visually, the turbine was spinning much faster than it was in the pool. It was still not fast enough for the turbine to generate enough voltage to power the power bank. The turbine did stay at the desired depth based on how much cable slack was given. While there was not enough voltage being output by the turbine at the given water velocity, it would have generated more voltage at a higher water velocity. This was proven when the device was returned to RIT to and the drill was attached to the system, providing enough rpm to generate the required voltage.

Running this system test showed us viable modes of anchoring the system. We initially tried our original plan, by extending the 8 foot aluminum pole and placing the turbine in the water. The resultant moment was too great for the system to handle, and we realized that had we placed the arm on the tripod (as originally planned), we would have broken the tripod legs. Note the bending in the aluminum rod; this force would not have been secured by the tripod. The water speed was also not nearly high enough on the shore (due to the boundary layer thickness being longer than our pole).

Rod being held at shore

Rod being held at shore

Therefore, we took the system to the center of the bridge, where the velocity of the water was the greatest. Even from such a high point on the bridge, we had a lot of extra unwanted cable. Hence we are cutting off 10 feet of cable. In a more ideal test, the water velocity would have been faster, and the bridge would have been lower. We would want a lower height for the bridge so that more of the cable is underwater to keep the depth of the device at the ideal spot.

View from shore

View from shore


One of the growing concerns throughout the second semester of MSD was waterproofing. Initially, water was leaking through the lip seal that held the motor shaft extension. This issue was addressed through the use of a stuffing body. The stuffing body was built from aluminum, had a lip seal on either end, had a cotton shoelace coated in graphic lubricant, and more graphic lubricant on the tunnel inside. Graphite lubricant was used as it lubricates effectively and breaks down with use, enhancing the lubricating effects even more. Both ends of the contact surface between the lip seals and aluminum stuffing body housing were then coated in epoxy to permanently close the enclosure. Below is a picture of the entire inner housing (without the motor):

Stuffing body with inner housing assembly

Stuffing body with inner housing assembly

Even after the lip seal issue was addressed, there were growing concerns for the cable end leaking. We realized that the bigger turbine blades caused the system to twist even more than expected. Because of the twist of the cable, water was coming in through the end with the cable shaft. We initially tried to fix the problem by using silicone which did not solve it. We even tried epoxy, but that too was of no benefit. We then realized that the best and most effective way to waterproof this end of the system was by using the epoxy to coat the mating surface between the cable and shaft, and then cover it in teflon tape. We then applied more epoxy outside for good measure. The final tests showed that the system was waterproof as the motor was still working and generating the same power as before at the same rpms even after extensive testing in the river. Below is a picture of the epoxy and teflon tape connection point:

[Picture of teflon tape + epoxy for cable end/shaft]

Boost Converter

A boost converter was implemented to attain a constant higher voltage output, at the cost of current. Since the purpose of the device is to charge the battery at a trickle charging rate, the lower current output was not an issue. The boost converter is set such the voltage output would be approximately 4.5V at any given time. This was set slightly lower than the 5V requirement so that in case there is any unwanted fluctuation in the voltage due to electrical components, the voltage would still be in the safe region. The data we attained showed us that we were getting a charging current of ~50 mAh at 270RPM which is similar to the RPM we got at the river in the final system test. That would mean that a normal Iphone 8 battery would be charged within 36 hours. While this may seem long, it is not an issue as it is expected that this device will be kept in water for weeks on end, and hence can charge an external power bank day and night at that trickle charging rate.

Bill of Materials

The total cost for the device, accessories, and rod system $181.65. However, if the cost of the rod is removed due to its ineffectiveness in the field, the cost drops down to $143.51. While it is uncertain of the manufacturing costs of the Water Lily and Estream, they are both retail at least $199.

The team could further drive down manufacturing costs by manufacturing large amounts of blades using molds rather than individual prints. The device also currently has 50 feet of cable attached to it, but after testing off of a bridge it was assessed that 30 to 35 feet of cable would be more than adequate in nearly every application. The cable was one of the most expensive things on our BOM, and reducing the amount needed for each device by roughly 40% would be an effective way to drive down manufacturing costs. All other components that are required for assembly would also be purchased in large quantities in order to reduce unit cost.

The current Bill of Materials can be found in the following link: Final Bill of Materials

Risk and Problem Tracking

Nearly all of our risks and problems have been addressed by the final review. In this final phase, Risks #1 and #40, waterproofing the housing and producing steady power at a voltage that will charge a battery, were both solved by the team. These were the primary risks hindering system functionality.

The Risk Assessment Document can be found here: Risk Assessment

The Problem Tracking Document can be found here: Problem Tracking

Potential Improvements


All engineers can reflect on a project and see where mistakes were made, how they may have been corrected at the time, and evaluate how the best way to fix them if completely redesigned. After designing, building, and testing the Trick Charger, the team easily identified potential areas of improvement that could be used to develop Trick Charger V2:

1. Aluminum Housing. The housing would have been significantly more durable and far less susceptible to deformation, especially on the threads. By the end of the project, our final ABS housing literally did not have threads; we had been sealing it together with silicone adhesives. This was effective for the duration of the project, but threads would provide a much more secure fit that would be easier to open and reliably waterproof with O-rings. Due to the deterioration of the threads over time, the O-rings were able to stay in place less and less, and eventually became completely obsolete.

2. Blade Connection Arms. While our fin structure held strong in the river, even using the blades with the thinner connection arms, the a fin did fail in this location in a test performed in the pool in which the device was aggressively pulled through the pool. In higher velocity waters, there is an increased potential risk of a fin breaking. The blades need a much thicker and solid interface with the rest of the body.

3. Shape of Blades. The blades have the potential to produce adequate power but the pressure drag for them due to the current pitch angle is very high. This is one of the reasons a blade connection arm snapped during a previous test. We have found recently that blades with varying pitch angles along the full length of the fin helps minimize this drag while still using the same radius.

4. Entire Shape of Housing & Blades. The design utilizes external flow around it in order to rotate the blades. We believe that while this may be an effective way to generate power and avoid river debris, a design with an internal blade surrounded by an external casing would have the potential to produce significantly more power. Early on we had drawn ideas of devices that funnel flow through a circular bladed cross section; they are shown below:



Funnel Impeller

Funnel Impeller

5. Boost Converter Housing. The inclusion of a boost converter would let the whole setup create the required 5V at any given rpm. The only thing that would be fluctuating would be the current. Since a big part of the trick charger is to charge the battery pack at a "trickle" charging rate, low current, such as 0.02-0.03 amps is not a big issue at all. Since the inclusion of this device was so recent, there was no housing built for this. Ideally, the design would be altered to include this device inside the charger itself. However, if that were to not be possible, a small enclosure would be built from normal 3D printed plastic to house the chip/board. The male USB from the power bank would run into the female USB on one end of this housing containing the boost converter. Then out of the other end would be a male USB plug which would plug into the female USB port of the charger cable itself.


One of the crucial lessons preached in MSD I was the importance of using systematic processes in order to further progress. While designing the Trick Charger during MSD I, it was easier to manage and stick to these processes. The team would meet during every class period, all action items would go on Trello, and weekly emails would be sent out in order to update the team on weekly action items.

MSD II brought about challenges during our build and test periods that hindered (and in some cases altered) the processes in place. For example, weekly emails started becoming less frequent in the midst of the chaos; this often led to us focusing on the short term goals at hand but sometimes losing track of important dates that stretched further ahead.

Looking back, there are a few things team members could have used in order to stay more organized and on target with their objectives:

1. More group meetings. Scheduling our MSD II work time during other team member’s class time has undoubtedly reduced the amount of time we collaborated as a full team. This overlap increased the chances of communication gaps and minimized the diversity of ideas within the group.

2. Group meetings vs. group work sessions. It is difficult to brainstorm and plan ahead as a team and actually complete the working action items simultaneously. It would have been useful to organize “meetings” for collaboration and planning with “work sessions” to actively work on our individual or group deliverables as a team. This would have ensured everyone on the team had opportunities to express their design input and complete action items (or express concerns) in a timely manner.

3. Agreeable software. The team experimented with three different project management tools: Microsoft Project, Teamweek, and Trello. Each of these had their advantages and disadvantages, but none of them individually fit all of the team’s needs for this project. For example, while Trello was useful for tracking action items and progress, it was not useful as an overall scheduling tool like Teamweek. In the end, many of these tools ended up being superfluous to the team and went generally unused other than times close to Design Reviews. The team needed one piece of encompassing software that we could agree on so that the team remained diligent about checking and updating it with goals and progress.

4. Printouts. Any document that was particularly important to everyone on the team should be printed and handed out to everyone on the team. Having a physical copy undoubtedly forces the owner to see its content often, which will increase the chance of it actually being used. Printing out schedules, for example, would have forced the team to refer to and abide by them on a stricter, more frequent basis.

Key Successes

1. Guaranteed an output of 4.5V at all RPM values for river velocities greater than 1 knot. As river velocity fluctuates, so will the current output. While faster currents will lead to higher RPMs, currents, and charging speeds, the device will be able to “trickle charge” in a wide range of water velocities. This stepped-up voltage is due to the added boost converter which has proved to be the essential remaining component used to obtain system functionality.

2. Successfully waterproofed the interior housing using a stuffing body - an apparatus used to seal boat propeller shafts - containing lip seals and graphite-infused water resistant grease. Silicone and epoxy adhesives were also used to waterproof the cable inlet to the device.

3. Optimized 3D printed blades to increase RPMs and easily be attached, removed, or replaced by the user.

4. Successfully used DC motors as generators to produce power from the rotating turbine blades.

5. Properly secured all internal device components to handle the rigors of operation.

Final Project Documentation

Below are links to all the of the Final Project documentation.

The Technical Paper document can be found here: Technical Paper

Imagine RIT Poster can be found here: Imagine RIT Poster

The Customer Handoff Outline can be found here: Customer Handoff Outline

Imagine RIT Demo

The demo at Imagine will model the same setup as our previous dry system test. The Trick Charger will be fully assembled and connected to a drill that can supply rotational energy to the device's driveshaft, causing the blades to rotate and simulating its reaction to flowing water.

A video of our previous demo setup can be found here: [DEMO LINK]

Home | Planning & Execution | Imagine RIT

Problem Definition | Systems Design | Preliminary Detailed Design | Detailed Design

Build & Test Prep | Subsystem Build & Test | Integrated System Build & Test | Customer Handoff & Final Project Documentation