Design Of Experiment (DOE)

Design Of Experiment

Design of experiment (DOE) is a method that allows us to determine which factors used in a experiment is the most influential and how we can optimize the results. Different factors have different effects and levels which will produce different results.

There are 2 methods to help tabulate the data and find the most influential. Fractional and Full factorial. Full factorial uses all the data to come up with its answer while fractional factorial uses fewer number of data/experiments but it is more efficient and resource-effective. However a disadvantage of fractional factorial is that it would be slightly more inaccurate.

PRACTICAL

In this practical session, we were given 2 launcher per group, one for fractional and one for full factorial. We were tasked to find out which factor would produce the furthest or shortest distance for a projectile being launched. 

There are 3 factors, 

1. Arm Length
2. Projectile Weight
3. Stop Angle

For my group, when we tabulated the data for the full factorial, we found out that the Arm Length was the most significant factor that affected the flying of the projectile however in fractional factorial, the Stop Angle was the most significant factor that affected the flying of the projectile. 

Unfortunely, my group did not produce the expected results like other group. Usually, the fractional factorial would produce similar results as the full factorial however our group was different. This could be due to mistakes made during the testing or inaccurate result taking that resulted in the anomaly.

After completion of the given tasks, there was a mini competition where each group were meant to hit down 4 targets using the previously tabulated data to assist us. The targets were miniture of our very own lecturers!😃 It was a great opportunity for students with pent up anger to released at their respective lecturers in secret. As i did not have any anger against Dr Noel and other lecturers, I was not eager to win however my group still finished top along side another group. Hariz was very enthusiastic to knock down all targets👀. 

Here is a satisfying video of a target getting knocked down tho,

With better knowledge on using full and fractional factorial to find out which factors can help us optimise the wanted results we can proceed to solve the case study individually!

CASE STUDY

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

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

Microwaving time, 4 minutes and 6 minutes

Power setting of microwave, 75% and 100%

8 runs were performed with 100 grams of corn used in every experiments and the measured variable is the amount of “bullets” formed in grams and data collected are shown below:

Factor A= diameter

Factor B= microwaving time

Factor C= power

 

Run order	A	B	C	Bullets (grams)
1	+	–	–	3.44
2	-	+	–	2.44
3	–	-	+	0.74
4	+	+	-	1.44
5	+	–	+	0.95
6	+	+	+	0.32
7	–	+	+	0.44
8	–	-	-	3.12

We can tabulate the results and find the amount of “bullets” formed in grams depending on the 3 factors, diameter, mircowaving time and power. We can use 2 types of factorial method, fractional factorial and full factorial. By using the excel template provided on brightspace, we can key in our data and tabulate its results by plotting as shown below.

FULL FACTORIAL

Based off the graphs and results,

Factor A ---> When diameter of bowl increases from 10cm to 15 cm, mass of bullets decreases by 0.1525 grams from 1.69 to 1.5375 grams.

Factor B ---> When microwaving time incrases from 4 min to 6 min, mass of bullets decreases by 0.9 grams from 2.06 to 1.16 grams.

factor C ----> When power setting of microwave increases from 75% to 100%, mass of bullets decreases by 1.9975 grams from 2.61 to 0.6125 grams.

Based off the differences in the amount of bullets, we can conclude that the ranking for most influential factor would be:

1. Power (Steepest gradient and largest difference)
2. Microwaving Time 
3. Diameter of bowl (Gentlest gradient with smallest difference)

Interactions Effects 

A X B

Based on the plots, Low B has a positive gradient while High B has a negative gradient and since there are no intersection between both lines, therefore there is no interaction.

A X C

Based on the plots, Low C has a negative gradient while High C has a positive gradient and since there are no intersection between both lines, therefore there is no interaction.

B X C

Based on the plots, Low C  and High C both have a positive gradient however since there are no intersection between both lines, therefore there is no interaction.

This the excel sheet for both full factorial and interaction factors

FRACTIONAL FACTORIAL

Based off the graphs and results,

Factor A ---> When diameter of bowl increases from 10cm to 15 cm, mass of bullets increases by 0.29 grams from 1.59 to 1.88 grams.

Factor B ---> When microwaving time incrases from 4 min to 6 min, mass of bullets decreases by 0.71 grams from 2.09 to 1.38 grams.

Factor C ----> When power setting of microwave increases from 75% to 100%, mass of bullets decreases by 2.41 grams from 2.94 to 0.53 grams.

Based off the differences in the amount of bullets, we can conclude that the ranking for most influential factor would be:

1. Power (Steepest gradient and largest difference)
2. Microwaving Time 
3. Diameter of bowl (Gentlest gradient with negative difference)

This is the excel sheet for fractional factorial.

CONCLUSION

Personally, I felt that in this topic, the tutorial lesson was pretty confusing😬 especially the excel sheet. When we were first learning about the different factors and how it would affect the results and how we can use it to optimise the desired results, it was understandable at first however the more we were taught, the more confusing it got for me. I did not grasp the concept of the excel initially and I was only putting in values. It took awhile for me to understand what I was plotting and what the graph was actually showing.😃

The practical was slightly less tedious. It was quite fun at first as I was deprived of a very fun childhood😢 playing with cultapults and launchers (JOKING🤡) however having to do 64 tries, I got tired and it got boring however the lil competition at the end of the session did liften up my spirit again. Below is a video showing one way to make the 64 tries less mundane and also a practice for the competition at the end.💪

Personally, I would not do it again unless I could keep changing and adding different factors to have my projectile fly the furthest as possible but that would defeat the point of this practical but I do understand the need to do 64 tries to understand the objective of DOE.

If you are really bored and have nothing else to do, here is a 1 min video of a POV of a soldier operating a catapult and firing at giants (mostly us).😁

-END OF BLOG-