DOE (Design of Experiment)


This is the blog for DOE. 
It is going to be actual hard work writing this blog. 
It is so rush as CNY is coming up with many projects due. 
But since this is graded, i am "encouraged" to do it. 
Hope this 3% is worth it. 

So I will simply document the process of creating a series of experiments to obtain the required results for examination via Full Factorial and Fractional Factorial data analysis methods; and my learning reflection of my DOE Practical session.


We learn about full and Factional factorial 
Full means all 
Factorial means faction or part of (e.g 4 out of the 8 runs)
It has to be not bias of any sort 
 
What could be simpler than making microwave popcorn? Unfortunately, as everyone who has ever made popcorn knows, it’s nearly impossible to get every kernel of corn to pop. Often a considerable number of inedible “bullets” (un-popped kernels) remain at the bottom of the bag. What causes this loss of popcorn yield? In this case study, three factors were identified:

  1. Diameter of bowls to contain the corn, 10 cm and 15 cm

  2. Microwaving time, 4 minutes and 6 minutes

  3. Power setting of microwave, 75% and 100%

 

FactorsLOW (-)HIGH (+)
A - Diameter of bowl to contain the corn10 cm15 cm
B - Microwaving Time4 minutes6 minutes
C - Microwave Power75%100%


8 runs were performed with 100 grams of corn used in every experiment and the measured

variable is the amount of “bullets” formed in grams and the data collected are shown below:

Factor A= diameter

Factor B= microwaving time

Factor C= power



The effects of Single Factor are totaled and averaged separately. This is shown on the graph below.


 
This is the graph for full factorial
 
 

 
It might look confusing with all the different values but let me explain what this graph shows 

From the graph above, we can infer that 
  • When the diameter increases from 10cm to 15cm, the mass of "bullets" increases from 1.35g to 1.50g.
  • When the microwaving time increases from 4minutes to 6minutes, the mass of "bullets" increases from 0.88g to 1.97g.
  • When the power setting of the microwave increases from 75% to 100% , the mass of "bullets" increases from 0.52g to 2.33g.

So we can concluded 
The most significant is Factor C followed by Factor B and lastly Factor A




Interaction 

A vs B 


The gradient of both lines are different. Therefore, there is a significant interaction between A (Diameter) and B (Microwaving Time). 

A vs C
 


 
 

The gradient of both lines are different. Therefore, there is a significant interaction between A (Diameter) and C (Microwave Power). However it is very close 0.525 and 0.455 So the interaction might be not that significant.

 
 
 
B vs C




The gradient of both lines are different. Therefore, there is a significant interaction between B (Microwaving Time) and C (Power Setting).




Discovery

The full factorial shows that Mircowaving Power is the most significant followed by Mircowaving Time and lastly, size of bowl. 
It also shows that at "+, +, +" The mass of "bullets" is greatest. This is very confusing to me. As higher diameter gives a higher surface area so that more corn can become popcorn so the mass of the "bullets" should decrease instead. Also higher mircowaving power and time should cause more popcorn to be form. But in my case, with the greatest power, time and bowl size. The amount of bullets is at its greatest.




Now moving on to Fractional Factorial 

A fractional factorial is ‘less than full’
 
- It provides fewer than all possible treatments are chosen to still provide sufficient information to determine the factor effect.
 
It is more efficient and resource-effective, but you risk missing information. So we have to be smart when choosing

We learn in class that we cannot anyhow choose any 4 but we have to be decisive and smart when choosing. 

We have to make sure 
- All factors are considered/varied
-  It is balance



What you see in the picture shows a good and balance fractional factorial
It includes all factors with both its high and low levels. Including the same number of time 



Therefore, these are the 4 values that I choose. Runs 2,3,5 and 8
 These are the values
 
This is the graph I got for the fractional factorial values.

It looks similar yet different to the full factorial graph. 

From the graph above, we can infer that 
  • When the diameter increases from 10cm to 15cm, the mass of "bullets" decreases from 0.85g to 0.70g.
  • When the microwaving time increases from 4minutes to 6minutes, the mass of "bullets" increases from 0.60g to 0.95g.
  • When the power setting of the microwave increases from 75% to 100% , the mass of "bullets" increases from 0.27g to 1.29g.
From this we can see that the the most significant factor is still Factor C, followed by Factor B and lastly the least significant factor is Factor A. 

This is exactly the same as the full factorial. This means that the values I have chosen for the fractional factorial is good and balance. 

There is a difference though. In this case, Factor A decreases, while in the full factorial it increases.

For the Fractional factorial.
It shows that Mircowaving Power is the most significant followed by Mircowaving Time and lastly, size of bowl. 

The bowl size in this case makes sense as, higher diameter gives a higher surface area so that more corn can become popcorn so the mass of the "bullets" should decrease. Which matches the graph. But higher mircowaving power and time should cause more popcorn to be form. But in my case, with the greater power and time the amount of bullets increase too



Learning Reflection

So,for bothe the tutorial and practical, I actually enjoyed and learn a lot. Although the tutorial was kinda boring and long. My lecturer Mr Chua made it as enjoying as he could. He taught us that a good fractional factorial data analysis consists of the equal amount of low and high level. 
 
 After learning all these, I managed to finished my pre-experiment quickly.He also guided us through the excel process which made it fun for me to do. I really want to thank him for guiding us through the process thoroughly.

For the practical lesson, I would like to thank my beloved teammates, Enzo, and Redza for helping me with the excel. To be honest, I just read the distance of how far the ball landed and told them. They were so patient and help me key everything in.  We just have to simply key in the values during our experiment and the graph will appear.
 
I was SHOCKED that when the length of the arm of the catapult increases, the flying distance decreased. I thought its the other way round where when the length of the arm of the catapult increases, the flying distance will increase due to long arm meaning can "reached" more. 
 
Also firing the cataput was so fun and satisfying. I had a lot of fun firing it. 

Athough my team mater Redza, broke the 3D printed arm, we still had a spare "arm" and continued
 
During the group competition, we were shocked that we missed the second nearest target. We only manage to successfully hit down all 2 targets which led us to receive 6 marks. I felt it is because our team was under prepared and we taught we had 3 practice tries for each target. But it turns out we only had 3 practice tries for all targets. 

Furthermore we were the first group and didnt do any practice on the side, thus we used trial and error. Ended up failing a lot and causing confusion in the team.
 
All in all, design of experiment is really interesting and fun. I learnt a lot. This is my favourite practical, although I felt we could have done better in the competition if we had more time . 


Time to study now since my MST is coming.
Will update my blog during the next practical !!!!!!!!!!!!!












Links to sheets




















Comments

Post a Comment