Blog post



I saw a really interesting paper in the AIP journal, Physics of Fluids, but I felt I'd missed the boat on writing it up as a news article. So instead, I decided to write a Twitter thread on it. You can read the whole thing there. Alternatively, I've pasted it below.

Last week, I came across this news story (by @beckyferreira),which reported on some research out of MIT’s Department of Mechanical Engineering. It was about how Oreos fail. So OF COURSE I had to download the paper 1/x

For those unfamiliar with Oreos, they’re a sandwich cookie –a creamy icing filling, flanked by two cocoa-flavoured wafers. People debate the ‘best’ way to eat an Oreo. Maybe you bite into it, dunk it in milk, or twist & separate the wafers and lick the crème. All are valid 2/x

However, if you’re in that third group, it’s likely that you’ve longed for a situation in which, post-separation, an equal amount of crème remains on each wafer. More often than not, what you get is a mostly dry wafer, and a fully loaded one. 3/x

Crystal Owens, a Ph.D. candidate at MIT turned years of loyal-but-somewhat-disappointing Oreo-eating into a research project. As she explains in the @vice article linked above, when she joined MIT, she learned how to use her lab’s rheometer. 4/x

A rheometer is a system designed to study the flow and deformation properties of a fluid, e.g. its viscosity (a measure of ‘gloopiness’).The basic idea is that you put a sample of a fluid between two parallel plates, set one rotating at a known rate (strain), and measure the stress. 5/x

Sandwich cookies are, scientifically speaking, rather like a parallel plate rheometer. Owens realised that if she could controllably twist the wafers apart, she could study the crème’s behaviour & investigate if it might be possible to evenly split it between the wafers 6/x

The team dubbed this area of study – whereby the flow and fracture of sandwich cookies is measured – as Oreology. It comes from the product name (Oreo) and the Greek term ‘rheo logia’ for ‘flow study’. Anew one for you @ologies @alieward! 7/x

Before I go any further, here’s the paper. It’s open access, so is free to read 8/x

One of the first measurements they made with the rheometer was the force required to separate an Oreo by pure rotation (no pull, just twist).This type of force is called torque. They found that on average, Oreos require a torque of 0.1 N.m. 9/x

Owens estimates that this is about one-sixth the force needed to twist a round doorknob, and one-twentieth the torque required to open a soda bottle cap. The thickness of the crème layer did have an influence –“double stuf” Oreos were the hardest to separate. 10/x

The crème itself is also relatively soft; considered “mushy” by food scientists. But how does it fail? Firstly, a cohesive failure of the crème would involve the crème itself rupturing. In those situations, you’d expect to see a bit of the crème on both wafers. 11/x

But what Owens found reflected her own experience in daily life – the crème stayed in one piece, but consistently detached itself from one wafer, leaving that wafer mostly dry. This is an adhesive failure. And there was a clear preference, related to how Oreos are made 12/x

On the manufacturing line, a pump dispenses a lil blob of warm crème onto the ‘bottom’ wafer (wafer 1), before sending it along the line to another robot arm, which then places a second wafer (wafer 2) on top. They’re then assembled into boxes with the same orientation 13/x

In this paper, the crème stayed stuck to wafer #1 80% of the time for newly opened boxes of Oreos. And while twisting the wafer more quickly resulted in higher yield stress and strain, it wasn’t found to make a significant difference to the crème distribution 14/x

Interestingly, in ‘old’ boxes of Oreos – left open for along time before testing – “nearly every cookie in the box split differently, leaving substantial crème on both halves.” Owens noted that in those instances, the crème tended to cover more of the wafer than usual…15/x

… suggesting that the crème had somehow flowed after it was deposited. This may have altered the bond between it and the wafers so much that the crème could then fail cohesively instead. Attempts to reproduce these conditions had inconsistent results 16/x

Owens and her colleagues also tested the influence of milk on the structural integrity of an Oreo. One wafer was found to fail after immersion of just 5s. A full cookie could withstand a 63s-long dunk without collapsing 17/x

Crystal Owens told @Vice, “The results validated what I saw as a child—we found no trick for opening up our Oreos. In essentially all possible twisting configurations, the crème tends to delaminate from one wafer...” 18/x

In the final part of the paper, Owens and her colleagues present a custom Oreometer – a 3D-printed torsion testing device that can be used outside the lab. It requires no power or fancy electronics, and they say it has a material cost is $6. 19/x

It uses rubber bands to ‘power’ a clamp that holds an Oreo(or any other similarly-sized sandwich cookie) in place. The torque is applied by adding pennies into two columns; one mounted on either side of the clamp. It’s simple, but genius 20/x

Want to build your own? Owens and her colleagues have madethe design, instructions and associated equations open source. Get experimenting! 21/x

I love research like this. It takes what might seem like a ‘silly’ idea and transforms it into something clever and interesting. And judging from the conclusions section, Owens has no shortage of ideas for future food science-related projects. Cookie manufacturers, take note! 22/x

Anyway, I hope you enjoyed this little thread as much as Ienjoyed reading this awesome paper. Here’s the link again in case you missed it earlier  23/23