The Gary Hilliard FQC Scholarship, made available through the Foundation of FTA (FFTA), provides a grant for students to conduct a flexography-focused research project.
I was awarded this scholarship in 2018 to pursue a project examining how press setup variables influence soft-touch coating results. Specifically, I wanted to discover if consumers could discern between samples generated through different press setups, and whether or not they showed a preference for particular variables.
So why focus on soft touch? With significant advancements in high-quality image reproduction, flexography is becoming an increasingly attractive option for high-end packaging. In this market, tactile coatings are trending as a way to engage with consumers and provide a luxury feel to a package. In my research, I found the sense of touch extremely difficult to measure. However, its impact on multi-sensory communication is well recognized and I wanted to focus on consumer response instead of attempting to quantify tactile outcomes.
3D-rendered images of the fictional product.
All photos and data courtesy of Emily Anderson
To generate distinct soft-touch samples, I established a three-factor, two-treatment experiment, where the factors included tint sleeves, anilox rolls and substrates. Specifically, I used:
- A 60 degree, 12 bcm anilox roll
- A high-volume, alternate geometry anilox roll
- A 55 shore A tint sleeve
- A 67 shore A tint sleeve
- A polyethylene terephthalate (PET) substrate
- A polypropylene (PP) substrate
These were chosen according to the range suggested by the manufacturer of the coating formula used, and were limited to what was available in our on-campus labs. All samples were printed on our Mark Andy 2200 press. I used a KS-9812 AQ Soft Feel Coating donated by Kustom Group. During the initial setup of the experiment, I had considered adding alternate coating formulas to the list of variables by exploring UV coatings, but the high temperatures of our conventional UV curing unit caused the plastic film substrates to warp.
Ultimately, I produced six unique press setup scenarios through which I printed the samples. To assess the consumer feedback on the samples, I used a choice-based (also known as conjoint) analysis.
The statistical design outlined here resulted in three null hypotheses:
- There is no statistically significant difference among consumer preference of soft-touch coatings applied on PET vs PP
- There is no statistically significant difference among consumer preference of soft-touch coatings applied with different anilox technologies
- There is no statistically significant difference among consumer preference of soft-touch coatings applied using different tint sleeve hardness options
Because the samples generated in our pressruns were produced for a consumer survey, it was important to simulate a real-world product as closely as possible without distracting from the task at hand, which was to evaluate the feel of the surface texture. To do this, I created a fictional product with a simple, 1-color design using a dieline for a small standup pouch. With the help of SAS Institute’s JMP statistical software, I ordered the samples into six distinct pairings and fastened them into booklets, so participants could engage with the fictional products side by side. Participants were encouraged to handle the samples as if they were real, three-dimensional products, touching both the front and back. By doing this, I hoped to more closely simulate an in-store environment, rather than having participants simply run a hand over each sample.
Analysis & Results
As described, I conducted a choice-based study in which respondents were asked to select a preference for each of six paired sets of samples. Participants were selected as a convenience sample, and the majority of participants were recruited through graphics communication classes. In total, I had 91 respondents in the study. The data was analyzed in JMP using the conditional logistics regression, or conditional logit model developed by McFadden (1974) and the bias-corrected maximum likelihood estimator described by Firth (1993), with a confidence level (or p value) of 0.05.
Through the data analysis, I found all three variables were statistically significant; accordingly, all three null hypotheses were rejected. The results showed the PP substrate was preferred over the PET, and the 67 shore A tint sleeve was preferred, particularly when on the PP substrate. The anilox rolls had the least influence among the variables, but there was still a notable preference for the 12 bcm anilox for this application. The interactions between the variables were not considered statistically significant.
As expected, participants could discern between the different substrates used in the study and expressed a distinct preference. However, the preference for the harder durometer tint sleeve was somewhat surprising. The most significant finding of this research showed there is room for a better understanding of consumer preference for tactile coatings used in flexible packaging. In addition, it raises the possibility that the flexo industry could further optimize soft-touch coating results that appeal to the greatest number of consumers.
I am forever grateful to FFTA for making this project possible (and for trusting me to present on stage at Forum 2019). I’d also like to thank Matt Apke from Kustom Group for his help with selecting and sourcing the soft-touch coatings, as well as Mark Andy Inc, Harper Corporation of America, Apex International and Multi-Plastics Inc for their support. And a big thank you to my supervising faculty, Dr. Malcolm Keif, for having my back at every turn.