P18485: Biochar Concrete Roofing Tile Manufacturing and Complete Roof System
/public/

Test Plans

Table of Contents

Test Plans

Design of Experiments

Purpose

The Design of Experiments, or DOE, approach is a simple way to understand the effect of many regressor variables on a wanted response variable. This could drive down the cost of testing, since many variables are being tested together, rather than each variable being independently tested. The formula used for simple DOEs is as follows:

m^k = number of samples

where:

m = number of levels,

k = number of variables/factors

Along with variables, each variable has a number of levels, that can be tested. An example of this would be when looking at the variable speed, you may have three levels, low, medium, high.

Methods

The team used a simple Design of Experiments approach by first brainstorming the many factors that would be considered variables in our experiment. In this DOE our response variable is Strength. After variables are brainstormed, each was researched to see if there was already some data, or some conclusions out there to avoid over testing. Below is a list of the original variables brainstormed:

After some consideration, here are the variables that were used in the final DOE plan:

Results

The team ended up with 3 variables, each with 3 variables. This means, according to the formula, m^k tests need to be run, therefore (3)^(3) = 27 samples need to be created. Here are our variables and levels color coded:
 Variables and Levels

Variables and Levels

For the ratio of aggregate, there mix levels were chosen, in relation to Ricardo’s mix ratio, shown below:

 Ratios Explained

Ratios Explained

The samples will be created as such, with a mold that can hold 9 samples each, over three days, each day having a aggregate mix, either high, equal to, or low:

 Day 1 Samples

Day 1 Samples

 Day 2 Samples

Day 2 Samples

 Day 3 Samples

Day 3 Samples

Plan

Our plan is to choose three days that members can pour concrete for, and begin as soon as all materials have arrived. In the next section, the process from creating the mold will be explained.

Feasibility: Prototyping, Analysis, Simulation

Concrete Test Sample Mold

Based on the Design of Experiments analysis, a 3X3 mold would be most efficient to provide the samples needed. The wood mold was created using information for the previous team, P17485. The Acrylic Backboard was added based on their feedback from using just the wood mold. The Acrylic holds the wood mold so it can be placed on the shaker table. The board also forms a smooth surface on one side of the samples (similar to what we hope to achieve with full tiles) and it helps prevent leakage.
 Specimen Wood Mold

Specimen Wood Mold

The drawing for the wood mold can be found here.

 Specimen with Acrylic Backboard

Specimen with Acrylic Backboard

How wood mold fits into Acrylic Backboard

How wood mold fits into Acrylic Backboard

The drawing for the Acrylic Backboard can be found here.

Sample Creation and Test Plans

During this phase, the team was able to create the samples discussed in the DOE plans from last semester.

Purpose

The purpose of these samples is to be able to test the properties outlined in the DOE. AS a recap, these are the factors and levels being tested.
 Variables and Levels

Variables and Levels

Procedure

Prior to creating the samples, the team created a testing procedure for creating samples. This procedure was followed to create all 27 samples. This procedure can be found HERE, and the following notes discuss the basic procedure followed: Sample Creation
  1. Portion the correct amount of: Pumice, Sand, and Cement
  2. Mix the proportioned amounts together, adding water to the mixture
  3. Clean surface (acrylic) of mold and Coat the insides of each sample mold cell with mineral oil
  4. Fill 2 rows of the sample cells with the concrete mix, following the requirements of the DOE
  5. Portion the correct amount of plastic
  6. Add plastic shreds to the remaining third of the mixture and fill the remaining row of sample cells with mix
  7. Move Mold to the Shaker table
  8. Shake Mold on Shaker table for 1 minute OR until the pumice has visibly risen to the top and cement has sunk to the bottom
  9. Move Mold to an open space in lab to dry

Sample Removal

  1. Let sit for 24 hours
  2. After 24 hours, remove Acrylic bottom and flip mold.
  3. Remove all samples by pushing evenly on all parts of the sample.
  4. Take the specified samples to be cured in the appropriate ways: In the Water tub in the lab ,Wrapped in plastic, or Air Drying in the lab
  5. Remove samples after specified amount of time from their curing environment and prep for testing.
A document for our testing can also be found here.

Samples and Pictures

 Sifting Pumice

Sifting Pumice

 Measuring the Pumice Density

Measuring the Pumice Density

 Weighing Plastic String for Samples

Weighing Plastic String for Samples

 Mixing Pumice, Sand, and Cement

Mixing Pumice, Sand, and Cement

 First Set of Samples Poured on 12/4/2017

First Set of Samples Poured on 12/4/2017

Lessons Learned

During the process, the team was able to make a "lessons learned" log, to improve each consecutive sample creation, or for future testing. Below is an example of those lessons:
 Lessons Learned During Sample Creation

Lessons Learned During Sample Creation

Testing

To test the highest tensile stress in the specimens, we will be utilizing ASTM Standard C293 Standard Test Method for Flexural Strength of Concrete (Using Simple Beam With Center-Point Loading). This is the same testing standard P17485 used. The logic for going with a tensile test as opposed to a compression test is because of the nature of the tile assembly. With the way the tiles are assembled and supported on the roof, the tiles will be in tension on the bottom surface. This is the anticipated failure mode.

Predictions from Literature or Observations

This subsection contains any predictions created by the team, in terms of how the results from our future tests will show:

Test Roof Plans

Some of the engineering requirements are mapped to features of the assembled roof. To test that our roof meets those requirements, we need to construct a test roof made of tiles and a test roof made of zinc sheets. Since the zinc sheets are what are currently being used, we want to show that our roof has more benefits than a zinc roof.

Frame

 Test Roof Frame

Test Roof Frame

Full Test Roof

 Full Test Roof

Full Test Roof

Budget

 Test Roof Budget

Test Roof Budget

Tests

Below are the tests that are going to be performed on both the tile roof and zinc roof.
  1. Loudness Test
  2. Solar Gain Test
  3. Waterproof Test
  4. Wind Resistance Test

Recap of Testing

To satisfy all Engineering Requirements, we need to be able to test against those requirements. Below are the respective tests for each Engineering Requirement. It is worth noting that the "Bottle String Width" and "Bottle String Length" requirements are contingent about the specimen testing that will be conducted at the beginning of MSD II.
 Test Plans Mapped to Engineering Requirements

Test Plans Mapped to Engineering Requirements

With the Samples created and cured, the next step of the DOE is to test each sample. This includes reaching out to Lab Coordinators to find time for training and time to test. Matt was able to train with the Lab Coordinator. Below is a picture of our fully cured samples:

Finished Cured Samples

Finished Cured Samples

Test Plan Summary

Concrete Strength Test

Arrangements have been made to begin sample testing on February 6th, 2018 with completion expected by February 9th, 2018. These results will test us the strongest mixture. From there, we will determine the most feasible combination of materials to use to make out tiles.

The samples will be test based on ASTM C293 Concrete Center Point Loading.

Mold Materials Test

In order to prepare for using the vacuum former, it was decided that we would conduct a test on two different types of styrofoam with two different protective coverings to determine which performs best. Our initial test plan involved more material types, but the cost of such a test was too high and we reduced it to only the necessities. We will have two samples each of styrofoam and smoothfoam, and cover these samples with bondo and plaster of paris. We will vacuum form over these samples with a piece of 1/8 inch thick ABS, and make the positive for the mold out of whichever material performs best.

Tile Tests

To recap, below is are plans we have to test each engineering requirement.
Test Plans Mapped to Engineering Requirements

Test Plans Mapped to Engineering Requirements

Below are the test instructions and documents: