P17027: Starfish Gripper

Detailed Design

Table of Contents

Team Vision for Detailed Design Phase

Plans for phase:

Completed during phase:

Left to do

Progress Report

A more detailed list of tasks completed and decisions made thus far (up to 11/22/16), as well as tasks to complete for the rest of the phase are listed in the document below.

Team P17027 Project Update

Prototyping, Engineering Analysis, Simulation

Pump Operation Test

The pump was tested recently to ensure that it would operate using our specified power source. The test was successful, with the battery easily powering the pump. While freely flowing, we measured around 1psi of pressure, and the pump used 1.4 amps of current. When the outlet tube was pinched, we measured about 4 psi of pressure, and about 2.4 amps of current use. More testing will be performed with a more suitable pressure gauge, as well as with the actual gripping material.

Pump Test

Valve/Pump test

Valve Test

BJTs Schematic

BJTs Pump

More detail about Valve/Pump testing in the link below

Pump and Valve test documentation

Soft-Body Limb Fabrication

Further development of soft-body gripper fingers were designed and reviewed, with a smaller variation from previous designs being selected for further use. The new design reduced overall mass, and moderately decreased thickness of the limb, in order to both conserve the material consumed in production, and address the potential root cause of the air-bubble presence,seen in the previous iteration of limb fabrication, as seen below:

Limb v3 Fabrication Bubble issue

Limb v3 Fabrication Bubble issue

The new limb iteration, which was utilized in the design of associated mounting components, both slimmed the width of the limb profile, and added a relief feature, to account for the designed point of flow insertion into the limb.

Rendering of Current Soft Body Limb Design

Rendering of Current Soft Body Limb Design

Both of these fabrications utilized the in-lab equipment for addressing air-extraction from the silicone polymer, during the cure time. Through several iterations of limb fabrication, this was deemed an appropriate method through which to remove the vast majority of air bubbles and irregularities within the limb's body, with the noted observation that initial operation of the vacuum chamber, and subsequent extraction of the majority of bubbles (occurring within the first few moments of chamber activation) caused a notable degree of polymer material to spill/seep from the molding assembly. This amount was considered minimal enough to disregard at the current time.

Limb Curing Position with inclusion of Vacuum Chamber

Limb Curing Position with inclusion of Vacuum Chamber

Some issues were still experienced during curing with trapping of bubbles in the silicon material. The team is trying to get in contact with Smooth-On, the company that makes the silicon material, to come up with a solution to fix the bubble issue.

Limb v4 defects after curing

Limb v4 defects after curing

Finalized Limb Fabrication Process

  1. Mixing Procedures:
    1. Using a clean beaker or container, pour equal parts of the binary polymer compound (1A:1B)
    2. Add a small amount of the silicone thinning agent to the mixture (No greater than 10% of the mixture’s weight)
    3. Thoroughly mix the compounds together, slowly churning the mixture, so as to minimize bubble air bubble introduction into the mix (approximately 2-3 minutes)
  2. Molding Procedures:
    1. Assembly the mold components, ensuring proper part mating, and that all parts align
    2. Gently pour the polymer mixture into the mold, overlaying all features with an even spread of the compound. Attempt to pour such that the laying of the compound doesn’t inherently trap the air beneath it within the mixture pour.
    3. Having evenly filled the mold, and allowing the compound to settle into place, let the mold sit, under observation, for approximately 5 minutes, to allow for trapped major air bubbles to rise from the mixture.
    4. Pop and/or extract risen air bubbles, using a syringe.
    5. After all preliminary major air bubbles have been found and removed, place the mold within the vacuum chamber, on a level surface, and seal the chamber.
    6. Activate all connected vacuum nozzles, and ensure suction within the chamber has been achieved.
    7. Wait approximately 10-15 minutes while the mold is under vacuum suction for remaining bubbles and micro-bubbles to be extracted by the chamber.
    8. Having observed all significant bubbles be extracted, via suction, disconnect vacuum suction line from the chamber, and allow for the chamber pressure to return to standard pressure.
    9. Open the chamber and use remaining compound mixture to replenish the mold for whatever amount was forcible displaced by the preliminary vacuum extraction.
    10. Make sure nice obvious bubbles are present after re-pouring, and reseal vacuum chamber.
    11. Re-Activate all vacuum suction lines to the chamber.
    12. Wait an additional 15-20 minutes, while remaining bubbles in the mixture are extracted.
    13. Visually confirm all apparent major bubbles have been extracted via vacuum suction
    14. Disconnect vacuum suction, and allow for chamber to bleed to standard pressure.
    15. Using a rubber scraper, collect as much spilled compound around the mold as possible, and gently layer back on top of the curing compound.
    16. Remove the entire mold from the vacuum chamber, and place on a level surface at room temperature.
    17. Visually inspect the resting mold, and extract any and all remaining, observable air bubbles, using a syringe.
    18. Let mold sit for the remaining time of cure, at standard room pressure (approximately 4-6 hours for Smooth-On DragonSkin 20 + 8% Silicone Thinning Agent)
    19. After curing has been completed, peel away any excess compound that has set around the mold components.
    20. Gently pry apart the mating components, leaving the cured limb on the mold base.
    21. Gently remove the limb from the base, being mindful of the specific forming features of the mold base.
    22. Trim any remaining excess compound from the limb’s edges, and discard excess.
    23. Inspect all regions of the limb for defects and impurities, such as critical air bubbles, mis-formed features, inadequate mixture pockets, and any potential impurities in the mixture.
  3. Sealing & Finishing Procedures:
    1. Acquire and cut to form the designated non-elastic layer material for the limb (Ex. Duct tape)
    2. Mix a small portion of the binary polymer compound. No addition of the silicone thinning agent is required.
    3. Gently layer the compound mixture into the limb cap mold (the shallow dish formed to the same footprint profile as the limb)
    4. Inspect that not apparent bubbles have been trapped in the mixture. Remove any identified.
    5. Gently center the non-elastic layer on top of the set compound, and gently tamp into the mixture.
    6. Gently pull some of the mixture to layer on top of the non-elastic layer, thinly embedding it in the cap layer, and ensuring a polymer surface is present for the limb body to bond with.
    7. Gently center the cured limb body on top of the cap layer, and place down onto the mixture.
    8. Do NOT apply force to the application of the limb body, as it will force out the curing mixture. Nothing more than a light application should be used.
    9. Check to ensure the limb is centered above the cap layer, apply minor side-to-side adjustments if need be.
    10. Leave mold on a level surface at standard room pressure for remaining cure time. (approximately 4-5 hours for Smooth-On DragonSkin 20)
    11. After curing has been completed, peel away any excess compound that has set around the mold component.
    12. Gently remove the limb from the base.
    13. Trim any remaining excess compound from the limb’s edges, and discard excess.
    14. Inspect bonding of the limb body and cap layer, paying attention to the set position of the non-elastic layer within the compound.
    15. Clean and store all molding materials and tools. Discard any non-reusable/expended resources.

Comparison of processes

By comparison to the former method of fabricating, wherein limb curing was entirely done within the vacuum chamber, the above noted process was implemented for the mold design that was current at the time of process development, below are the results.
Left, limb molded using process outlined above. Right, limb molded constantly in the vacuum chamber.

Left, limb molded using process outlined above. Right, limb molded constantly in the vacuum chamber.

While this limb contained the most minimal presence of air pockets to date, pressure testing for actuation revealed a design flaw with the rigidity of the design. With the absence of any relief slots in the limb body, sufficient actuation was not made possible, and, upon testing, critical rupture in the material was experienced, when the limb was subjected to upwards of 6PSI. Below is evidence of this failure. Note of this design need was made, and incorporated to the next design iteration.

Critical Rupture in Limb Design

Critical Rupture in Limb Design

Current-State Limb Design
Utilizing all above noted details to limb design and the fabrication process, a new mold design was developed, seen below.
Latest Molding Variation Design, including Relief Slots

Latest Molding Variation Design, including Relief Slots

The resulting limb model from this design was fabricated and sealed, now utilizing a non-elastic layer of duct tape. The result may be seen below.

Finalized Limb Design for Senior Design I

Finalized Limb Design for Senior Design I

Finally, this limb, having been cleared of any major physical defects or impurities, was subjected to pressure actuation testing. This testing showed the limb capable of fully actuating, and easily handling pressure applications of upwards of 9PSI without complication.

Limb Actuation Testing, 0 - 8 PSI

Limb Actuation Testing, 0 - 8 PSI

Heat Dissipation Analysis

Heat Dissipation Analysis

Heat Dissipation Analysis

With all electronics running continuously, the internal temperature of the enclosure will equalize to approximately 50°C (122°F). This is not representative of the true operating conditions of the robot, which will be operated very intermittently, so this analysis represents the very worst case possible.

The Seaflo pump is rated for operation up to 43°C. With this in mind, the team does not anticipate any operational issues due to heat dissipation from the electronics enclosure, as all electronics will be operated intermittently and the internal temperature of the enclosure is highly unlikely to reach even 43°C.

Component Selection

Pressure Sensor

Many pressure sensors were examined for the gripper. The pressure sensor needed to be sensitive over the gripper’s operating pressure range of approximately .1-10PSI. The other major requirement is that the sensor be able to be exposed to water. Due to the water requirement, one of the original ones picked was an engine oil pressure sensor for an automobile. However, there was very limited information available on this sensor, so it was not purchased. Unfortunately, the selection of non-heavy-duty water compatible sensors is very limited. There were a few different sensors that fit this criteria, however they needed supply voltages that are not currently in use in the system. Therefore, the best sensor found is the MS5803 sensor from TE Connectivity. The team contacted TE to obtain a demo kit for the part, however the cost was significant. Sparkfun sells a breakout board with this sensor for $60, which the team will purchase two of. The sensor is waterproof, outputs using SPI or I2C, and measures 0-14Bar in sensitivity of 1 / 0.6 / 0.4 / 0.3 / 0.2 mbar.

Pressure Sensor Datasheet

Note that during this phase, the team decided to use two pressure sensors with the robot, in order to account for the pressure differences at various water depths.


The Venom 20C 2200mAh 11.1V lipo battery was selected to power the project. This battery was chosen as it was low cost, well reviewed, and well over the estimated battery size needed for the project. If there is extra money left in the budget as Imagine RIT approaches, we will purchase more batteries to extend the number of demonstrations we can perform during the day.

Venom 20C 2200mAh 11.1V

Drawings, Schematics, Flow Charts, Simulations

Top Level Electrical Schematic

Top Level Electrical Schematic

See the document below for updates on schematic progress: schematic status report.docx

See Schematics Folder for schematics of each individual component.

Design and Flowcharts

Pump Waterproofing

Previously, we had planned to waterproof the pump by covering seams and sensitive areas with a silicone-based coating, in the hopes that this would properly seal the pump off from water contact, as well as experimenting with taking some of the panels off of the pump electronics and placing some marine grease over sensitive areas. The team had some concern with this method being high risk and difficult to properly seal.

Now, we are planning to place the pump in the waterproof enclosure with the rest of the electronics. Waterproof cord grips/strain reliefs will be used with metal tubing to bring fluid to and from the pump. More detail is shown below in the model screenshots.

Model Screenshots

Full Assembly

Full Assembly

To give a sense of scale, the robot's envelope is approximately 10"x14"x16".



Tabs and slots cut into sides for easy slide together assembly. Brackets between the sides and bottom plate secure. Sheets to be water-jet cut from PVC.

Extension Mechanism

Extension Mechanism Extension Mechanism Extended

Extension mechanism mostly machined aluminum components. Nylon washers added at all joints to minimize friction.

Servomotor Waterproof Container

Servo Container

Servomotor and servohorn (arm/crank) to be purchased. Block and lid to be milled from aluminum (shown transparent here for visibility).

Exploded View:

Servo Container Exploded View


Manifold to be machined from Delrin plastic block obtained from machine shop. Support/mounting angles to be machined from standard aluminum angle. Fastened together with screws underneath.

Isometric View:

Manifold Isometric View

Front View:

Manifold Front View

Side View:

Manifold Side View

Waterproof Enclosure

NOT SHOWN: Battery, Arduino proto shield (attaches to top of Arduino), auxiliary circuits

Isometric View:

Enclosure Isometric View

Top View:

Enclosure Top View

Control Panel

Control Panel

To be 3D printed from PLA or ABS plastic, and fastened with 4 screws.

Note: The slot in the top of the panel will contain an SPST power switch. A model was not provided by the distributor, but the mounting cutout size was given.

Engineering Drawing Package

PDF Drawings Folder

As of 12/6/16, not all drawings have been completed. We plan to finish these by the time of the Gate Review.

Bill of Material (BOM)

public/Detailed Design Documents/BOM_DEC3.jpg

Link to PDF File

So far the pump, valves, battery, waterproof container, switches, transistors BJTs, and silicon dragon skin material have been received for initial testing. By the end of this semester, all of the items listed on the table will be ordered in preparation for MSD II. Most of the important component had been purchased for initial testing.

The BOM currently has a total cost of $701.55, which is 93% of the overall budget. This accounts for all parts needed to complete the project, but does not account for shipping and handling costs. While this is close to our budget limit of $750 dollars, there are $91 of screws, fittings, and bolts that can potentially be found in the machine shop. Along with this, the team is contacting Hi-Tec to see if they will be willing to donate the servo motor($49).

Test Plans

Below are test plans for all engineering requirements that are testable. These are to be tested during MSD II.

Deployable Depth Test Plan

Deployable Depth Test Plan.pdf

Life Cycle Test Plan

Life Cycle Test Plan.pdf

Successful Capture Rate Test Plan

Successful Capture Rate Test Plan.pdf

Plans for next phase

Overview schedule for MSD II:

Home | Planning & Execution | Imagine RIT

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

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