EGR 150-06 Team Consensus

Here is the Team Consensus for my EGR 150-06 team. This is a basic concept of I think the final project should be.

These are the questions that Dr. Harris posed:

1. What questions do you ask? (Think of IDEO)
2. Where do you get the answers you seek?
• internet
• books
• interview
3. What are the engineering skills you have to be successful?
4. What are the engineering skills you lack to be successful?
5. What is the solution trying to accomplish?
6. What are the criteria that will be judged? (think safety in the IDEO video)
7. What other questions are there?

Dr. Harris references a company called IDEO, they are a design company that tries to innovate while creating unique designs and solutions. There are some intriguing videos about IDEO that are worth watching.

Here is my consensus:

EGR 150-06 Final Team Consensus
Project: “Tweetable Coffee Pot”

• What is the main goal?
• To create a Twitter activated coffee pot

• What questions will your team ask? 
• How will we interface the arduino with the internet?
• What electrical components will we need?
• What will the layout be?
• How will the 3D printer, laser cutter, and vinyl cutter be incorporated into this project?

• Where will your team get the answers it seeks?
• The main source of information will be the internet. We will use websites like Instructables and use citations when needed.
• This article will provide much of the needed information 
• Magazines may also be used.
• We will also talk to the TAs in the FabLab for advice.

• What engineering skills does your team have to be successful?
• Basic circuit layout and building
• Basic 3D design and layout
• Creativity
• Time mangement

• What engineering skills does your team lack to be successful?
• Programing skills with Python and C++
• Fabrication

• What are the criteria that will be judged?
• How was the 3D printed part, vinyl cut part, and laser cut part incorporated?
• What was the moving part?
• Team work
• Presentation of the project
• Performance of the final product
• Incorporation of the Arduino board

• What other questions are there?
• How much money will be spent?
• Where will the resources come from?

Arduino Output Project

This article will cover some basic arduino design, layout, and coding. The parameters for this project was to do a sample project that comes with the arduino kit. The sample project was a basic blinking LED circuit and then modify the code and/or the design. It is composed of five components, not including the bread board and arduino itself. Here is the project:

Several weeks ago Dr. Harris told the class to order an arduino kit from SparkFun, an online electronics supplier. The choices for kits were these: the SparkFun Inventor's Kit for $100 or the SparkFun Mini Inventor's kit for $50. As the name suggests, the mini kit was simply a smaller version for half of the price. However, the regular kit include 2.5 to 3 times as much stuff as the mini.

First I had to download the arduino software. It is simple to install and straight forward to use. Once the software was installed the correct type of board had to be selected. In this case it was an Arduino Uno, the SparkFun RedBoard is an equivalent board. The basic programming for the project was already done, it was an example called 'Blinking an LED'. This is in the example program files and can be easily found.

There was one hiccup with the software, the program could not be uploaded to the board. This was due to the drivers for the arduino board not being installed properly. The proper drivers had to be downloaded and installed. The drivers can be found (here).

Now that all of the software was working correctly and the program could be uploaded, the hardware was next. The example that the class was told to use includes a diagram of the board set up. Here is the diagram and the board.

Source: SparkFun PDF, page 22

The arduino kit included a small tray to hold both the bread board and the arduino itself. The arduino has numerous output and input pins. The bread board is interesting, it is simply a group of contacts that can hold onto a pin or a lead. This allows simple and quick prototyping of a circuit as it does not need to be soldered together. The board's rows are number top to bottom and the lettered from left to right. This gives a concise way to notate where each component should go. These are essentially coordinates, like a10 or f5  In this case pin #13 was used as the output. The output wire, a green wire was used, was plugged into hole e2. The LED was plugged into c2 - c3. The negative leg of an LED is the shorter one and that side of the LED has a flat ridged. A small 300 Ω resistor was also used. The board set up can be seen the videos below.

The power for the board can be an external power supply or supplied through the USB port. In this case, the power was supplied through the USB port. The positive and negative wires were plugged into the proper holes and the circuit was complete.

Once the set up was done the test code was then uploaded. Here is the test code working:



The code makes the arduino send a signal through pin 13. This signal is then relaid to the bread board through the green wire. The code emits a signal through pin 13 that makes the LED blink for one second intervals.

Once the basic code was working it could then be modified. I tried to determine what would be a good project or challenge to code. I decided to stick with the single LED, but make it blink a pattern. I wanted the pattern to actually mean something, so SOS was chosen. The Morse code for SOS is this: dot-dot-dot-dash-dash-dash-dot-dot-dot. This is fairly simple and code could actually be useful. There are numerous brands of flashlights that can flash SOS.

Here is the SOS flash:



The board set up is the same as before, but the code was changed. Here is a link to the code used. The code has been saved as a .txt file type which can be copied and pasted to the arduino software.

This project was not too hard. The programming for the unit converters and beam calculator prepared the class for this project. As is typical with this type of project, the hardest part was getting the interface software to work, arduino to computer. But once that was squared away, everything else went smoothly. Another interesting, but tedious, project. I am looking forward to programming further for an arduino.

3D Print Project

This post is about designing and printing an object with a 3D printer. It will cover what software was used to design and print, along with the parameters of the project. This post will also cover the types of materials, mainly plastics, that were available to use.

The main constraints for this project was that it needed to be designed by the student and fit within 1.5 x 1.5 x .25 inch box. Dr. Harris suggested some 3D modeling software such as SoildWorks or Tinkercad. I choose to use another software called Creo. I am currently using Creo in my Engineering Graphics class, so I am already familiar with it. Here is the process:

The first step was to sketch the design. The design is the final dimensions, except for height. This sketch is not included, as the one below is a better example. Once the paper sketch was completed the design could then be put into Creo. The design needed to be finalized before starting in Creo. If you are going to use Creo, you need to know exactly what the design is. As mentioned in previous posts, Creo is a not meant for creativity. Here is the first Creo sketch:

Sketch 1. This does show the dimensions.

The design was fairly simple. First a 1.5 x 1.5 inch square was sketched out in Creo. The design consists of a backwards 'E', a 'P' and the a small 'S'. This stands for my initials, EPS, and is effectively my monogram. The main pieces are 1/4" wide which makes the two gaps 3/8 x 5/8 inch. The 'P' was also simple. A circle was made with a diameter of 3/4" and then two lines were made. These two lines were made tangent with the circle and then this shape was connected with the backwards 'E'. A concentric circle was made with the center point of the larger with a diameter 1/4" and the same process was repeated.

Once those steps were done, the 'S' was next. This was difficult because I did not know how to make a letter in Creo. As it turns out, there is a text option. Choosing the font for the 'S' took the most time, about 10 minutes.

The next step was to extrude the sketch. This was done with the extrude tool. The overall thickness needed to be 1/4", so the logo was extruded to 1/8". Here is the extruded sketch:

Extrude 1, this was 1/8"

There needed to be a base in order for the 'S' to stay in place. This was done by making another sketch on the bottom face. This was also a 1.5 x 1.5 inch square.

Sketch 2, Creo can be hard to use, that is why it is sideways.

The sketch was then extruded 1/8" making the final piece 1/4" thick.

Extrude 2

There are a few limitations with a 3D printer, such as sharp corners. 3D printers handle smooth or rounded corners better than sharp ones. The four outside corners were smoothed with the round tool, the radius was set to 1/8". It was a simple matter of clicking on the corners to rounded. The model took an hour to complete, a small part of a free evening. Here is the final model:

Final model

Once the model was completed it needed to be saved to the correct file type. The default file type for Creo is a '.PRT', but the 3D printer software needs an different type such as a '.STL' or '.OBJ'. This was done by selecting 'save as' and then choosing the file type. The file type used was '.STL'. This was tricky, the part does need to be selected and then saved. A .STL file basically saves the model as polygons and simple shapes.

The printers in the FabLab are on a server. One of the laptops in the FabLab had to be used in order to access the server. Once this was accomplished, the printer could then be selected.

The printer that was used is a Printrbot Simple. This printer seemed to give good results and was simple to use. The Printrbot takes a code called 'G-code'. G-code gives the printer points or concordats to follow. It is accomplishes this by slicing the model into .8 millimeter slices or layers. In order to turn a .STL or .OBJ file into G-code, a third-party program had to be used. One of the TA's recommended a software called Cura. Cura is convenient because the model of printer is selected and then the file is loaded into Cura. The model is then positioned in Cura and then printed.

There were two different options for the printing material, known as filament. This filament is 1.5mm thick, comes in spools and a variety of colors. The two types of plastic available were ABS and PLA. ABS is the classic plastic, Lego bricks are ABS. PLA is a corn based product which will eventually degrade. For this process, Dr. Harris wanted the students to use PLA as these are first prints and in case of any mistakes. PLA can go into a landfill, but if ABS is used, your mistakes with literally last forever.

Here is the print in process with purple PLA filament:

Beginning of print

Approximately 1/3 done

Printing progress bar

Approximately 3/4 done

The overall print time was 45 minutes. The printer took 10 minutes to heat up and then 35 minutes to print. This printer gave fairly clean results. There were three other printers running at the same time and those were not printing nearly as cleanly. The results are decent, especially for a first print. Here is the finished piece:

Top-down view

Slanted view

There are still lines on the print, which could be sanded smooth. There is a trick with ABS to remove the lines on the piece. The piece is suspended over acetone bath in a sealed container. The acetone vapors react with the ABS plastic, acetone will dissolve ABS. The process does not take too long, maybe 45 minutes to an hour. However, this process can be sped up if the acetone is heated. This would create more vapor. It is recommended that you read this article as it goes into further detail. This article does talk about using an acetone bath which would accomplish the same effect.

Since this piece is simply a rough piece, more a proof of concept, it does not need to be finished. There was some basic clean up, a few thin strands of material from the print head. You will also notice that the curves are not smooth. This is due to the file type. The .STL saves the model as simple shapes, so the curves are made of triangles.

Overall, this project was quick and fun. This project took 2:15 - 2:30 to complete. The model took an hour or so to design it, another half hour to convert it and prep for printing. The overall print time was 45 minutes. The most tedious step was the print it self. It did feel like watching paint dry, unexciting.

There is really nothing that I would change, I might use ABS in a future print. I am happy with the print, and the color is not half bad either. This piece came out exactly 1.5 x 1.5 x .25 inches, which is one of the parameters. I will eventually mirror the design so that it can be used to make a sand mold and then casted in either copper or brass. This will  be used as a brand for branding various objects.

I do not wish to share these files, as they are of a private/personal nature and I do not want the them or the print to be recreated. If you do wish to obtain the files, drop me an Email.

Beam Testing

This write up is about testing the beam that was designed and built in the last three projects. This write up will show the expected results, actual results, and the reasons why the two did or did not match. The calculations will be both on paper calculations and from the beam calculator. Here are the results:

Firstly, the beam could not deflect more than 0.1 inch at a concentrated load of 250 lbs. The beam also had to be no more than 2 x 2 inches with a length of 24 inches.

The first step of testing the beam was to do some rough calculations. These are the ones from before I built the beam:

Rough calculations

This shows the on paper calculations and results. It also shows the results from the online calculator that was made previously. There is a fair bit of difference in between the two. The paper calculations gave that the inertia would be 0.821 in^4 and a deflection of .0253 inches. The calculator stated that the inertia would be .685 in^4 and a deflection of .0304 inches. The projected weight, according to the calculator, should approximately be 160 grams. The paper calculations are less accurate due to rounding in between steps. This should be the main source of error.

The beam was already built and needed to be tested. The beam had to be tested in the FabLab and one of the TAs was required to help or at least supervise. Firstly, the beam had to be weighed. There was an electronic scale at the FabLab. I brought the beam in on two separate days, a rainy day and a dry, sunny day. The first day the beam weighed 221.5 grams and the second day it weighed 175.5 grams. The change was astonishing, 40 grams of difference. This is due to how porous basswood is. It can readily absorb water. Now that the beam was weighed, testing the deflection was next. Here are two pictures of the beam in the testing apparatus:

Front view

Side view

The apparatus is simply an A-frame with roller supports and a downwards facing hydraulic jack, which and was operated with a socket wrench.. The jack provides a concentrated load, the same type as the calculations. There is also a dial gauge underneath the beam, it is visible in the first picture. The beam is placed in and the span was set. The rollers were already 18 inches apart, which is the correct span. The whole apparatus was clamped down to the table so that it could not move.

There is a small weight sensor between the jack and the beam. The sensor uses a USB connection with a computer to display how much force is applied to the beam. The sensor was ratted to 500 lb and if the beam was to be destroyed, the senor needed to be removed.

Onto the testing. The results were recorded by the TA who was helping to run the apparatus. Here are the results of the 2 x 2 by 3/16 thick basswood I-beam:

Weight of beam: 175.5 grams

Beam width: 1.97 inches

Beam height: 2.07 inches

Weight applied at 0.1 inch of deflection: 219 lbs

Deflection at 250 lbs: 0.12 inches

Pass or fail: failed

The beam did not pass all of the tests even though the calculations showed that it would. It was 0.07 inches too tall and deflected 0.02 inches too much. The calculations were incorrect. I believe that main source of error was the density of this bass wood. This basswood is rather thin, so any deviance in grain structure or pattern at all would cause major problems. I did try to pick the 'cleanest' wood, without blemish or knots, that could be found. However, the densities of materials are typically measured with large blocks of the material, not thin sheets. I should have realized this sooner and tested these specific pieces, but I did not.

The beam did pass the weight test, even though this was not an actual requirement. The beam was 24.5 grams under the implied weight restriction, which is pretty good. But it was still 15.5 grams over the projected weight. The 0.07 inches of extra height could have been corrected with light sanding on the tops of both flanges. However, the 0.07 inches was due to a slight gap while the glue cured and would not effect the deflection significantly. The beam was also 0.03 inches under the 2 inch restriction.

I am generally pleased with the results, though slightly disappointed that the beam failed the test. But it is still impressive that 3/16 inch basswood can support well over 250 lbs. If I could redo this project I would definitely test the density of the wood by finding the volume and weight. I would also test the deflection of the wood and see if it would pass. If I did go up to 1/4 inch wood I believe that the beam would pass the deflection test, but fail the weight restrictions. The aforementioned deflection calculations were significantly off, but the weight was also incorrect. This happened because the glue was not included with the calculations. An extra 10 - 15 grams should make up the difference of weight, so roughly 5 - 7.5 grams per glue joint.

Another fun project with some cool gadgets and calculations.