Design of Experiment (DOE) BLOG 📈 👩🏽‍💻🎖

HEY EVERYONE🤩!!! I am back with another blog 😆🥳. For this blog, I will be showing the case study of an example that i have done🧐 and in my learning reflection i will be talking about learning DOE👩🏼‍🎓 and then applying it to my practical lesson👩🏼‍🔧. I hope yall will gain some knowledge👩🏼‍🏫 while reading it and hopefully try this DOE approach for any of your current or upcoming project🪜!

SO LET'S BEGIN ⬇️⬇️⬇️

FULL FACTORIAL DATA 📉 based on the case study

Effect of single factors📌 and their ranking🥇

These were the values that were keyed into the excel sheet 📉.

From the graph above, it is shown that factor C has the largest gradient which shows it has the largest difference of 1.85g. This means that it has the largest change when the power setting of the microwave is changed from 75% to 100%. It shows that it is the most significant because it has the biggest influence on the reaction, so the higher the power setting, the lesser the mass of bullets. Hence, the higher the popcorn yield. The next factor is B, it has the second largest difference of 1g. As microwaving time increases, the mass of bullets decreases. The least significant factor is A, this is because it has the smallest difference of 0.09g.

Ranking 🎖 of Effect of Single Factors:

1: Factor C

2: Factor B

3: Factor A

Include all tables and graphs📈 both as pictures and as excel file🗂

One Drive link : Eshvin DOE BLOG.xlsx

Data Analysis for Full Factorial Design

Graph for the Full Factorial Design

I find the average when A+, A-, B+, B-, C+ and C- and compare them in a line graph.

Interaction Effects👀

Interaction effect is when two factors interact with each other, the effect of one factor on the response variable is different at different levels of the other factor. Now this is the 3 graph for Interaction Effect for FULL Factorial Design (AxB, AxC, BxC):

Interaction (A X B)

AT LOW B,

Avg of Low A= 1.93

Avg of High A=1.99

At HIGH B,

Avg of Low A=1.03

Avg of High A=0.675

Interaction (A X C)

At LOW C,

Avg of Low A=2.575

Avg of High A=2.03

At HIGH C,

Avg of Low A=0.385

Avg of High A=0.635

Interaction (B X C)

At LOW C,

Avg of low b=3.075

Avg of high b=1.53

At HIGH C,

Avg of low b=0.845

Avg of high b=0.175

Conclusion📝 of the data analysis📉 for full factorial data analysis

Based on the graph, the most significant factor that affect the loss of popcorn yield is C (Power).

FRACTIONAL FACTORIAL DATA 📉 based on the case study

Effect of single factors📌 and their ranking🥇

The runs that were chosen for the fractional factorial design are 1, 2, 3, 6 from the run order. This is because these runs provide all factors that occur the same number of times both at high and low levels, thus these runs are orthogonal and will provide good statistical properties.

These were the values that were keyed into the excel sheet 📉.

From the graph above, it can be determined that factor C is the most significant factor because it has the largest gradient with 2 g. Therefore, it has the largest changes between high and low. When the power setting of the microwave is increased from 75% to 100%, there will be less bullets in the bowl of popcorn. The next most significant factor is factor B, this is because it has the second largest gradient with 0.7 g. Lastly, factor A is the least significant due to its smallest gradient with 0.3g.

Ranking 🎖 of Effect of Single Factors:

1: Factor C

2: Factor B

3: Factor A

Include all tables and graphs📈 both as pictures and as excel file🗂

One Drive link : Eshvin DOE BLOG.xlsx

Data Analysis for Fractional Factorial Design

For my Fractional factorial analysis, I chose runs 1, 2, 3, 6 because all of the factors occur both low and high the same number of times (orthogonal).

Graph for the Fractional Factorial Design

Conclusion📝 of the data analysis📉 for full factorial data analysis

Based on the graph, the most significant factor that affect the loss of popcorn yield is C (Power).

Learning Reflection 👀✍🏼

Tutorial

I felt 🧐these design of experiment (DOE) activities is a little different from the other practicals 👩🏼‍🔧that I have been doing and it has been a little confusing😵‍💫 for me. The main thing i understood about DOE it is used for discovering a set of factors which are most important to the process (or system)✅ and then determine at what levels these factors must be kept to optimize the process (or system) performance✅.

My lecturer explained to us the main fundamentals of DOE 😯 as shown below.

We were then taught on how to apply it in CPDD. We learned that we had to make a lot of prototypes 📦 as shown on the left but obviously it is infeasible to run all treatments thus it will be more realistic to restrict the number of runs 😪. That was when fractional factorial design was introduced and my brain started collapsing right then 🧠 🤒.

We learn that fractional factorial design is more efficient💪🏻 and resource-effective⏳, but theres a risk ❌ in missing❓ certain information. We learned how to fractionalise by learning how to select a subset of 4 runs from a 23 = 8 run factorial design ✅ as there are many possible fractional designs.

The one shown on the left is a balanced design ⚖️. All factors occur (both low and high levels) the same number of times. It is said to be orthogonal 🧊. Thus having good 👍🏻 statistical properties.

We then did a data analysis as shown below for a full factorial design using an example.

So with that what we learn 👩🏼‍🏫 is in the context of product design, DOE could be performed 👍🏻 with one

of these as the objective:

• “Selecting the few factor that matter from the many possible factors and study that few factor further”✅

• “Often we want to "fine tune" a process to consistently hit a target”✅

• “Maximizing or minimizing the response variable within the constraint of the process”✅

Practical

The practical session was undoubtedly a fun 🥳 one. We made use of catapult ☄️ to apply what we have learned about DOE. My group and i split into 2️⃣ smaller groups. My teammate Insyirah 👩🏼‍🔧 and I measure the length 📏 of the catapult ☄️ , measure the weight 🏋🏻‍♀️ of the projectile and measure the stop angle 📐 using protractor. First My teammates Asraf 👨🏼‍🔧 and Valerie 👩🏼‍🔧 perform their experiment 🧪 first as they were doing full factorial design. Then we did the fractional factorial design experiment together.

As you can see, the catapult ☄️ is being taped to the ground to prevent ❌ it from moving when it is being used as to avoid any discrepancy with the measurement of the length.

On the right ➡️ , you can see that we are adjusting the board of sand placement so that the ball ☄️ can land on the sand accurately.

RESULTS

These are the results 📊 we collected from the experiment 🧪.

When arm length increase ⬆️ from 28.1cm to 33.6 cm the flying ✈️ distance of projectile decrease ⬇️ from 305.6 cm to 233.3 cm.

Based on the graph 📉, the most significant factor that affect the flying distance of the projectile is C (Stop Angle) 📝.

When arm length increase ⬆️ from 28.1cm to 33.6 cm the flying ✈️ distance of projectile decrease ⬇️ from 305.6 cm to 204.5cm.

Based on the graph 📉, the most significant factor that affect the flying distance of the projectile is C (Stop Angle) 📝.

After we finished collecting the data ✔️, we actually had a group challenge🏅. The challenge was to hit the targets 🎯 on the table and we were given 3️⃣ tries for 4️⃣ rounds. I think the funniest 🤣 thing was the targets 🎯 were actually a picture 🖼 of the chemical engineering lecturers🧑🏼‍🔧👩🏼‍🔧 .

As you can see my teammate ,the all mighty💪🏻🏋🏼 asraf, is aiming for the target 🎯 and guess what HE SCORE A HATTRICK FOR THE 3️⃣ ROUND‼️‼️‼️

On the right ➡️ you can see that my teammate, Insyirah, was measuring 📏 the length of the catapult ☄️ to the 🎯 to get a precise angle to shoot 🔫 the target 🎯.

This was the final score📝 of the group challenge⛳️.

Although my group did not win😩, we did came 🥈 so thats pretty good👍🏻👍🏻.

Overall, I enjoyed 😁 doing this practical despite it being a little boring 🥱, but it was different 😯 from my previous experiences. But completing this exercise was enjoyable 🥳 and educational 👩🏼‍🏫 because I got to learn something new 😃 and hone my Excel skills 📉 at the same time. It has been a pleasure learning 🤠 about DOE since it employs statistics 📊 to design experiments to obtain knowledge 🧠 with a minimum number of trials. My understanding of it was initially difficult 🤨 during the tutorial 📝, but after the practical 👩🏼‍🔧, I knew I could comprehend the information ✅. A key element 📌 of this practice was determining when to use Full Factorial Design and when to use Fractional Factorial Design💡. Conducting an unlimited number of experiments infinite ∞ will provide the most accurate ✅ results, but this is not something we are physically capable of doing 😩 due to the time factor ⏱ when conducting experiments 🧪. By using a fractional factorial, we can reduce the number of runs 🏃🏼‍♀️ we have to run as much as possible while still getting a general idea of which factor is significant👌🏻. This is very useful 👍🏻and I'm pretty sure this is the foundation of any project ✍🏼, so I will certainly be using this for my FYP 💼.

THANKS FOR READING MY BLOG‼️

HOPE YALL GAIN SOME KNOWLEDGE ALONG THE WAY 👀 AND ARE ABLE TO APPLY IT IN YOUR DAILY LIFE 🌆. CHECK OUT MY PREVIOUS ARDUINO BLOG IF YOU HAVENT! 🌚🌚

I'LL BE BACK SOON FOR MY NEXT BLOG ⭐️ AND I HOPE YALL HAVE A GREAT DAY 🌈