Customer Handoff & Final Project Documentation
Team Vision for Final Demo and HandoffPhase Objectives:
In this phase the team would like to accomplish these tasks:
- Complete normal tile mold
- Complete top tile mold
- Pour and cure all tiles
- Complete construction on Zinc Roof
- Complete construction on Test Roof
- Complete all tests for tile verification
- Meet with Business Team, finalize costs
- Finalize documentation for customer hand off
In this phase, the team has divided these goals into these themes and assignments:
- Matt - will be leading the tests and creation of test roofs
- Derick - will be leading project management tasks, working with Saunders team integration, problem tracking, creation of documentation, and guiding team discussions.
- Dan - will be in in charge of the creating the tiles, and leading tile creation and curing
- Connor - will be in charge of researching risks and finalizing manufacturing requirements.
Phase UpdatesPulse of the Project Being the last project phase, these objectives have been completed thus far and will be talked about.
- Molds have been created
- All Tiles created and tested
- Costs have been finalized for this project, new costs may appear for Saunders team
- Frame has been expanded, now being used for both tiles
- Test Roof has been built, and used to test
- All Tests have been completed with results
Test Results Summary
DescriptionThe strength test was conducted in the Student Life Center by stacking weights onto the tile in 5 lb increments until failure. Two planks of wood were used to simulate the roof supports. The weights were stacked on top of a plexiglass sheet supported by a pair of shoes, in order to simulate the surface area and weight distribution of a person standing on the roof.
The tile broke at 340 lbs, almost 100 over the ideal engineering requirement of 250 lbs.
Solar Heat Gain
DescriptionBoth test roofs were placed outside in direct sunlight for 2 hours. The test took place on April 27th, 2018. The temperature on the inside surface of each roof was recorded every minute using an Onset HOBO External Temp/RH Logger. The sensors were switched from one roof to the other after one hour to ensure that they were providing accurate readings.
The point-by-point data can be found here.
SummaryThe inside surface of the tile roof was on average 10 degrees Fahrenheit cooler than the metal roof. The metal roof also changed temperature more rapidly than the tile roof. The switching of the sensors from one roof to the other confirms that each sensor was showing an accurate reading. It should also be noted that at around the 82-85 minute mark and 100-101 minute mark the sensor became detached from the tile roof. This is the reason for the missing data points in the “Solar Gain Temperature Difference” graph.
DescriptionThe test was completed using a typical garden hose and a variable nozzle. The first setting used was the Shower setting. This setting is similar to light but steady rain. The second setting used was the Stream setting. This setting consisted of a single stream of water at a higher pressure than the Shower setting. We used this setting to simulate heavy rainfall.
DescriptionThe waterproof test was designed to test the water resistance of the tiles when assembled together on the roof. This test was conducted by assembling a test roof with 4 normal tiles and 2 top tiles. We sprayed the tiles for one minute with a hose nozzle on the "shower" setting to replicate rainfall. We wrapped the sides and bottom of the roof in plastic sheets, and lined the bottom with dry paper towels in order to observed any water that dripped through the roof.
Water Durability Test
DescriptionThe water durability test was conducted in order to determine the ability of an individual tile to withstand constant water flow. This test was conducted by putting a tile in a sink and letting a stream of water run over it continuously for 4 hours. The water flow rate was about 67 mL/sec. After 4 hours, the tile was inspected for signs of water penetration.
Wind Lift Force Test
DescriptionThe wind lift force test was designed to determine whether hurricane winds could detach a tile from the roof. Based on ASTM standards, a class 1 hurricane has wind speeds of 74 mph, while a class 2 hurricane has wind speeds of up to 110 mph. Using the formula for force of a fluid on a flat stationary plate
Test ResultsThe results of this test showed it took 75 lbf to remove one wire, meaning it would take 150 lbf to remove the tile entirely. This is well above the ideal criteria of 81 lbf.
Optimal MixFor our concrete mix, we continued to use the ratio of 35% Pumice, 35% Sand, and 30% Cement by volume. We ordered 1/8th Fines pumice from Hess Pumice. When we poured our first tiles, we used a mix buffer of 1.35, as this ratio had worked for us when we poured our concrete samples. This rounded up to a recipe of
- 7 cups of pumice,
- 7 cups of sand, and
- 6 cups of cement per tile.
However, during the process of pouring tiles, we discovered that we needed a larger buffer. We settled on a ratio of
- 8.5 cups of pumice,
- 8.5 cups of cement, and
- 7.25 cups of cement per tile.
We have not accounted for why the buffer amount changed for the tiles vs. the concrete samples.
Final Tile Designs
Normal Tile MoldTo create the normal tile mold, we took the following steps:
- Used the CNC router in The Construct to machine a large foam mold for plaster
- Lined the foam mold with tape in order to waterproof it
- Mixed and poured plaster into the foam mold to create a cast
- Allowed the plaster cast to cure and removed it from the mold
- Drilled holes in the plaster cast to allow for airflow
- Vacuum formed ABS sheet over plaster cast
We attached wooden frames to the finished molds to allow them to stack.
Top Tile MoldTop Tile Mold 3D Model and Sheet Metal Drawing
Finished TilesPICTURES OF FINISHED TILES
Problems With Finished Tiles
- The tile nubs were too small to reliably hold their shape during the drying process. The size of the nubs should be increased, and their shape should be changed to be wider at the base than at the top for added stability.
- The inverted design of the molds was designed to keep the concrete from sliding off while drying. However, it made it necessary to cut the plastic into a precise shape, and resulted in the plastic curling up on the ends, deforming the corners and edges of the tile. The tile mold should be made non-inverted, more like the Ecosur molds.
- There were many bubbles on the top surface of the tile that had not been shaken off. This caused imperfections and divots in the surface that could potentially let in water. This problem could be solved with more shaking, or a less viscous mix ratio.
- The tiles were slightly too long to fit well onto the roof when the supports were 18 inches apart. The tiles should either be shortened, or the supports should be installed further apart.
Final Process MapBelow is the final process map to be used a reference for any future business owners.
Cleaning of molds: After tiles are taken out of molds, use a cloth to wipe down remaining concrete from molds
Cleaning of metal frame: After wax paper with tile is moved onto the mold to form, wipe down the metal frame to remove any remaining concrete with a cloth
Keeping track of inventory: Using colored tape outline the areas in the inventory for which a batch (21 tiles) can be stacked within. Use a different color for each day the tiles must be cured in the inventory. Whenever a batch is placed in inventory the date is recorded, the inventory name, the batch number, and the color code the batch was placed within. Using a different color diagonal to create different color patterns (I.E. Red tape with blue slash through) so that only a limited amount of colored tape needs to be purchased do make each batch area unique. This data will be added to a master set that keeps track of all inventory. Batch numbering allows for tracking of defected parts.
Move tiles to next inventory/cure area:' Check the board/data tracker to determine which batches of tiles need to be moved forward in inventory. Gather a rolling cart. Go to the inventory and place tiles onto the cart. Move rolling cart with tiles to next inventory. Place the batch in the new inventory and mark the date which moved and the color of the batch. Each day this will be done first thing in the morning to have available molds for production.
- Pumice Crushing
- Plastic Shredding
- Metal Snipping
- Mixing of concrete (based on workers experience)
- Top tile
- Metal wire addition
- Roof installation
- Operator Manual
- Service Manual
- Recommendations for future work
Costs DiscussionThis section reviews the final costs of the roof. Using data collected by the team, market research, and stakeholder information, the Saunder’s team (Commercializing Pumice Roof Tile in Nicaragua) was able to estimate the production cost of a tile to be $1.69 USD. This cost was driven by material costs, material percentages per tile, and a number of logistical, overhead and labor costs.
Risk and Problem TrackingBelow is the risk assessment for this phase.
Below is our problem tracking table. This tool is useful in conquering some of the obstacles in an organized manner. Note that not all problems have been solved yet.
Plans for Wrap-up
Overall Project ConclusionsThe team concludes the following items after the conclusion of the project:
- Zinc has lower capital cost $512
- Pumice has a capital cost $844
- Pumice has a better yearly investment $28 per year vs Zinc = $34-68
- 2-3 workers including marketing, order generation, and admin
- Equipment cost about $3500
- Production cost for Tiles $1.69
- 3000 square foot facility to meet 21 roof demand, 21 tiles a day
- Operating at a 3 sigma level, about 5700 tiles produced
- 224 Normal Tiles needed
- 13 Top Tiles needed
- 237 total tiles needed
- 35% Pumice is better for costs but 40% was stronger
- Tiles was about 15-17 lbs
- Similar to proven processes (ECOSUR)
Hand OffThis is a list of items that will be handed off to the customer.
- Digital Documentation
- Tile Molds
- Example Tiles
- Foam to be used for future Molding
Recommendations for future teams
- Optimize concrete/pumice mixture for cost, as we are well over the strength goal
- Redesign mold to be non-inverted
- Improve nub shape to be stronger/more consistent
- Work on ways to get all bubbles off of the top surface
- Work on pumice grinding method
- Look into using current roofing material as molds/components of molds