Stain Removal Experiment

IN2023 Design of Experiments

Constantin Kraus, Juan Fernndez, Csar Gardida, Jorge Martnez, Jazmn Vzquez
April 28th, 2017

Abstract

Lipstick stains in cloth garments are of common occurrence and difficult to remove. The purpose of
this experiment is to maximize the stain removal by identifying the factors and levels that yield the
best results. For this experiment, two blocks (Machine A and machine B) and four factors (half two
levels and half four levels) have been considered for the design of a full factorial experiment. Three
variables were shown to be significant and optimal at: soak time (10 mins), detergent (Blancanieves)
and water temperature (warm). The other two variables were set at levels machine A (Maytag) and
pre-wash detergent (chloride) since these levels reduced the cost.
 1. Introduction

Stain removal is the process by which a piece of cloth that has been stained or soiled is cleaned. The objective of
stain removal is to prolong the use of a garment or a household decoration while not damaging it. While multiple
solvents and detergents are commercially available for the stain removal process not all of them work for the same
types of stains. To establish the ideal combination of factors that would maximize the stain removal of a lipstick stain,
a design of experiment has been executed. A design of experiments is used to design and execute an experiment in
order to process and obtain information on the effect of a factor over a response variable. For this experiment, four
factors were considered: soak time, pre-wash detergent, detergent and water temperature.

The purpose of this paper is to design, run and analyze a full factorial experiment to maximize the stain removal
process of a cloth that has been stained by lipstick. The specific objectives of this project are: design a full factorial
experiment, run the experiment to identify the variables that are significant to the stain removal process, and once
these has been done, the levels that yield a better response will be selected. After these objectives have been achieved,
a confirmatory experiment was conducted. In order to minimize variability in the experimentation process it has been
defined and standardized to ensure that the process can be replicated accurately.

2. Methodology

As a first step, the materials used for the experiment were selected. The cloth was selected to be a white cloth in order
to have a contrast of the fabric with the color of the lipstick stain. The cloth was cut into small squares in order to
execute the stain removal process that will be selected. The stain was selected to be lipstick due to its resistance to be
removed from fabrics and how commonly it is used. Each small square was marked by a line of lipstick of 7 cm and
then it was soaked, either for 1 min or 10 min, in a pre-wash detergent, either glycerin or chloride. After that, they
were machine washed in a mini wash setting, in a regular wash cycle, with three different detergents (Ariel, Axin
and Blancanieves) in three different types of water temperatures (Warm, Lukewarm, and Cold). The design of this
experiment was executed in two blocks, in a Maytag machine and a Whirlpool machine. After the experiment was
executed, each fabric square was graded by cleanliness in a scale of 1 to 4 via an established pattern designed by the
team (1 counting as the least clean and 4 being the cleanest)

3. Results and Analysis

The experiment was designed in two blocks and four factors in a full factorial design. After the 72 experiments were
conducted, the results (found in Table 2) were analyzed with the statistical software called Minitab. The software was
used to identify the significance and contribution of each factor. A statistical DOE Factorial analysis was executed
and it showed that the variables that had the biggest contribution (Figure 1) to the stain removal were soak time
(45.57%), detergent (25.53%) and water temperature (25.53%). A main effects plot (Figure 2) shows that the best
combination of factors and levels for maximizing the results are: Soak time 10 mins (2), detergent Blancanieves (3)
and warm water temperature (3). As there is no significance between the blocks and the pre-wash detergent the
recommended levels for these insignificant variables are: Maytag block (1) and chloride as prewash detergent (2)
since these are the cheapest options. In fact, this combination yielded the best result during the experimentation
process (ID 3/19).
Figure 3. Lipstick stain (before washing) Figure 4. Worst result (ID 10) Figure 5. Best result (ID 3)

4. Conclusions
To sum up, a full factorial design was created, executed, and analyzed. The results demonstrated that to maximize
the cleanliness of the cloth, it is necessary to soak clothes for 10 mins in chloride, after that, Blancanieves detergent
should be used, since the type of machine (blocks) has no influence on the result, and pre-wash detergent is marginally
less significant than the rest, it is recommended to use a Maytag washing machine and chloride as pre-washing
detergent. At the end, soak time, detergent, and water temperature proved to be the deciding factors.

For these reasons, we advise to focus on these factors and invest in them, while opting for the cheapest option in the
machine type and pre-wash detergent. As well the cleanliness is proportional to how high the setting is, so it is
recommended to use as much as ones budget allows.

5. References

Montgomery, D.C. (2000) Design and Analysis of Experiments. 5th Edition. Wiley Text Books, New York.

Cornell University. Removing stains at home. April 28th, 2016, from web site:
http://www.human.cornell.edu/fsad/outreach/upload/removingstains.pdf