Level: Alchemist
Read time: 20 minutes
This question came up on a forum I’m on and I totally own I completely underestimated the question and gave a shite answer. I gave a shite answer because in reality I had only a passing familiarity with viscosity measurements (I performed some 3 decades ago as a very green chemist) and almost no familiarity with it as it related to chocolate. Have a read.
A quick question for any chocolate makers here selling to chocolatiers or using an enrobing machine or chocolatiering yourselves. Do you know how to calculate the rheological properties/viscosity of your chocolate varieties? Is it possible to estimate with a math equation or would we need to have it analyzed in a lab? It’s a question that’s come up a couple of times for us and we would be curious if anyone has a solution.
It was and is way more complex than I thought it was. I really should have known better.
At this point, I have been researching and studying viscosity for the last 5 hours and think I finally have my head wrapped around the concepts enough to talk about and not 100% embarrass myself (again). To that end, my sincere apologies if I do get some of it not exactly right. The intent here is not so much answering the question (which I can now, sort of) but laying out the basics of simply talking about viscosity and how it relates to chocolate in an intelligible fashion so you too can do the same if and when the questions come up.
We have been playing with AI and this was a fun take on viscosity and its units as if explaining it to a 5 year old.
Imagine you have a big, thick, gooey slime. Now, if you try to push your hand through the slime, it resists and feels sticky. That’s because the slime doesn’t like to move quickly when you apply force to it. We call this “viscosity.”
Pascal-seconds is like a special way grown-ups measure how “sticky” or “gooey” a liquid or slime is. It helps scientists and engineers understand how hard or easy it is for liquids to flow or move. If something has a big number of pascal-seconds, it means it’s really sticky like honey or molasses. But if it has a small number, it means it’s more runny like water.
So, pascal-seconds is like a special number that tells us how “thick” or “thin” a liquid is, just like how we might say honey is thick, and water is thin!
Whether you know it or not, you know all about viscosity, even before that technical description above. Analogy time.
You are in grade school and playing around with glue and rub it all over your desk (oh, naughty, naughty) and then put a piece of plywood on top of it, but it isn’t in the right place so you start pushing or pulling on it to make it slide. You could either pull with a constant force and measure the velocity, or you could pull at whatever force is necessary to achieve a certain velocity. Either way, it's the same idea. Call the force F and the velocity v.
Now, just think about this for a moment. If the glue is really viscous then the force will be a lot. That means viscosity will be proportional to force. Also, because the glue is viscous, that plywood is not going to move fast so viscosity is also inversely proportional to the velocity. For any given force, a high velocity means a runny, less viscous material compared to a low velocity, because the less viscous material puts up less resistance to the drag.
Ok, a more technical definition.
Viscosity in its simplest form is a measure of a fluid’s (or liquid, in this context you can use them interchangeably) resistance to being moved, i.e. how hard you had to push against that glued up piece of plywood.
Also, if you made the piece of wood really big (in surface area) then you would have a lot more stickiness pulling against you which would mean a greater force. That force is greater not on account of the viscosity being greater but because the surface area is larger.
You finally push hard enough and long enough and the wood slides off the desk. Now you put on an extra thick layer of glue and this time the wood moves around much easier. That just means that layer thickness. Another way of say this is that if I have two different materials that I'm comparing viscosities of (I can hardly believe you rubbed your vanilla pudding all over your desk), and they both move with the same velocity for the same applied force, but one forms a sandwiched layer that is twice as tall as the other, then that one must have a higher viscosity. That just means that viscosity needs to be proportional to the layer thickness which we will call m (for meters).
Maths for the Geek Contingent
We said that viscosity (µ) was directly proportional to force F and layer thickness m, and inversely proportional to area A and velocity v. It can be represented like this.
µ = F *m / (A * v)
With a little rearrangement we get
µ = F/A * m/v
F/A has units of Pascals (Pa by definition), i.e. force divided by area which is just another unit of pressure like pounds per square inch.
Velocity is defined as distance over time or m/s (meters per second)
That means m/v has units of (meter)/(meter/s). The meter cancels out so
m/v =s.
Pulling it all together
µ = F/A * m/v = Pa-s
So there you have it. Viscosity has a unit of Pascal-seconds.
This is a case where it doesn't really make sense to think about it as Pascals times seconds -- that's just what things happened to work out to after all the cancellation.
I hope that makes sense and you get you kind of already knew it.
For those of you that want to really dig into some units and relationships, check out the aside.
Viscosity’s resistance is generally measured in Pascals. This is just a version of pressure. You don’t need to deeply understand this but I want to get the concept across. You inflate your tires to a given pressure, usually in psi or Pounds per Square Inch or lb/in2.. Another version of this is Newtons per square meter.(N/m2 ) and bringing it home, one Pascal (Pa) is just one N/m2. Viscosity itself is measure in Pascal seconds.
Finally, and I promise, this is the last set of units, because a Pascal second does not really make intuitive sense (see the Math's for the Geek Contingent for why), the powers at be have defined a Pascal second as a Poise (P), the unit of viscosity.
The next thing to understand is that viscosity is not one quantifiable measurement. It is the flow (velocity) relationships between the substance (glue) and the container it is in (like a desk and board) and the pressure the substance is under. It is also related to the temperature.
Whether you know it or not, most people have been exposed to the concept that viscosity is not just one number. You saw that in grade school with your glue and you see it now with your car.
Motor oil is literally advertised with viscosity ranges. 5w-30, 10w-40. Those are measures of how the oil flows and lubricates whether it is warm or cold. 5w-30 motor oil has a winter viscosity grade of 5 (w stands for winter). That means it is less viscous (less thick) at low temperatures compared to 10w-40. Likewise, 10w-40 has a hot viscosity grade of 40, which means it is less viscous at high temperatures than an oil such as 5w50.
As simple as those familiar examples makes it seem, viscosity is not just thick or thin. Teasing that all apart and remembering what I had in school is what took up most of those 5 hours. Given our focus is chocolate I’m going to just skim over a LOT of what you would find if you were to dig in. I still want to touch on it so you at least can recognize what you don’t really need to pay attention to.
Let’s get to it.
There are a number of parameters that go into actually describing viscosity (it’s rheostatic properties)
There is:
Shear viscosity (also called Dynamic viscosity)
Apparent viscosity
Plastic viscosity
Kinematic viscosity
Yield rate
Flow velocity
And others.
The first one, shear viscosity, is a measure of the amount of pressure it takes to move a viscous fluid and how the fluid behaves under that pressure. It is how hard you had to push on that glued board before it started moving.
In this context there are two types of fluids. Newtonian and non-Newtonian fluids. If the shear viscosity is constant over a wide range of shear rates then you have Newtonian fluids. This just means if you push it twice as hard, it will move twice as much. These kinds of fluids can usually be described with one value, i.e. it’s apparent viscosity.
The most common Newtonian fluid is water. If something behaves normally is it probably a Newtonian fluid; drinking or rubbing alcohol, thin motor oil, gasoline, cooking oil and many common solvents. Most normal liquids we come into contact with.
But we come into contact with lots that things that act…..strange. They kind of seem like a liquid but they don’t act like water. They are generally thick or viscous and that viscosity changes when you do things to them.
In science geek speak, fluids and liquids without a constant viscosity are non-Newtonian fluids and cannot be described by a single number. Instead we have to define them in terms of shear stress and shear rate.
Let’s connect them back to our glued board.
Shear stress is how hard you are pushing on that gluey board to make it move (and keep it moving).
Shear rate is how fast the board and glue are moving.
Non-Newtonian fluids are really funny creatures. Dialants, thixotropics and Bingham plastics oh my!! They are defined by how shear stress affects shear rate and vise versa. You don’t really need to deeply understand or memorize these. I present them because I find them fracking interesting.
Shear-thickening (dilatant) liquids. Their viscosity increases with an increase in shear strain (classic cornstarch slurry that turns ‘solid’ when you hit it hard but otherwise flows)
Shear-thinning liquids become less viscous as the rate of shear strain increases (lotions and blood are good examples)
Thixotropic liquids become less viscous with pumping, for example: glues, non-drip paint, greases, cellulose compounds, soaps, starches, gelatinous food stuffs, and tar. Thixotropic fluids may take a considerable time to settle back to their original viscosity. (this is also like a corn starch slurry)
Pseudoplastic fluids become less viscous with pumping, like thixotropic fluids, but immediately return to their original state when static again. Rubber, latex, and mayonnaise all fit here.
Rheopectic liquids become more viscous over time when shaken, agitated, or otherwise stressed. This is seized chocolate.
Irreversible liquids do not resume their original viscosity after agitation. Cheese, yogurt, and marmalade are good examples.Bingham plastics require a minimum shear stress (called yield stress) to initiate flow. The apparent viscosity decreases from an infinitely high value (like a solid) as the flow increases.
After all that, finally we can get to what we need to as Chocolate is a Bingham plastic. I know, it is a funny name.
Just like mayonnaise, you can let it set there and it can have ripples and waves and will resist flowing but you can stir it (stress it until it yields) and it will flow and pour. This is why you have to rap, tap or vibrate a chocolate after it goes into a mold. Until you stress it to flatten out, it sits there like a solid.
And how much do you have to stress it to flatten it out? I’m glad you asked.
That ability to resist flattening out is called yield rate and one of the key parameters you need to define the rheostatic/viscosity properties of a non-Newtonian Fluid.
Of course, “what is the viscosity of chocolate?” does not have one answer. It is literally the wrong question.
So what are the right questions? Even that is hard to define and what I think the main issue the OP had. Their customer asked them what the viscosity of their chocolate was. That really isn’t what they wanted to know. The question should have been “Can I use your chocolate in my depositing machine?” But it could have been, “Can I use this chocolate in my enrober?” which is a different question with different answers. And likewise, “Can I use your chocolate to make wafers?” has yet another answer.
To all of these questions, the answer can be yes or no (or probably) but what you can’t do is give one viscosity value because each of those cases require a different SET of viscosities. A better question would be “What is the apparent viscosity, yield rate/stress and plastic viscosity of your chocolate?” but even that does not work because Apparent viscosity is a relative, not absolute measurement. That is why it says apparent.
What it is really asking for is ‘what is the viscosity at the temperature and flow rate I need to use it at?”
Thankfully that has mostly been standardized but is also rather complicated. Check this out.
Definition:
“The apparent viscosity at a particular shear rate (20 RPM) is measured at a standardized temperature, 40° C for real chocolate and 50° C for compound coatings. Apparent viscosity is a relative measurement used as a singular data point throughout the confectionery industry. In application, the chocolate may be utilized with different shear rates. Chocolate is shear thinning, which means the viscosity decreases as the rate of shear increases. Therefore, to get a more complete picture of the product, viscosity is measured at multiple shear speeds. This allows the calculation of the other two important viscosity measurements: yield value and plastic viscosity “
The rub of all this is you have no good way to measure these values.
A rheometer is a laboratory device used to measure the way in which a viscous non-Newtonian fluid flows in response to applied forces (shear) at a set temperature. Some setups even look like glorified boards glued to tables.
Although it sounds simple, achieving the high accuracy and precision required by the industry is an extremely formidable task. There are many factors that affect the precision of this test method and sadly it is not as simple as measuring how fast chocolate pours or how far it spreads at some temperature.
The most important factor affecting the value of a home is location, location, location. In viscosity measurements, the most important factor affecting the quality of a viscosity measurement is temperature, temperature, temperature and the amount chocolate cools just during the test can vastly change the measurements to the point of making them useless.
So what is one to do? Do you send off your chocolate to have it tested at a cost of hundreds of dollars? Do you spend many thousands of dollars to get a good rheometer? I would suggest neither due to what the OP further offered up:
“They need a chocolate that measures 4 droplets on that flow scale according to the manufacturer, and I don’t know where this would fall on it.”
That right there is the rub. I’ve spent over 2000 words about viscosity, industry standard and all of that, and some enrobing manufacturer has come up with their own flow scale where the chocolate has to have a value of “4 droplets”
I tried to dig in and use my google foo and find that scale and utterly failed. What I did find is that there is a fine graph that shows fine working ranges of different applications and for the two values we have been talking about.
Clearly there is no one perfect set of rheostatic properties for all applications. With that information now in hand, you have what you need to make choices and can actually do something on a very practical level.
You can take your chocolate, use it, evaluate how it performed and then methodically change the property to need to adjust to give you the result you need. Basically, you will iterate your chocolate until you get to the flow characteristics you need. Have a look at this representation of how chocolate might deposit into drops/wafers, how it might enrobe a center and how it could fill a mold.
Plastic viscosity determines how well the chocolate will flow into a mold once the Yield viscosity has been overcome by an appropriate yield stress. Generally speaking, you are not going to have an issue with too low of a plastic viscosity. The chocolate is going to fill the cavity of the mold very easily. The only time you might have an issue is if you are using machinery (like the OP’s customer) that has a need for minimum viscosity parameters. As Plastic viscosity goes up the chocolate is going to have a harder time filling the mold evenly and even with a rapping may still not settle the chocolate and in extreme situations my leave air bubbles in the chocolate.
Yield viscosity presents in a different way. As shown in the lower portion of the graphic above, the greater the shear viscosity the more of a doming effect the chocolate has. Sometimes, as in the case of certain drops and wafers, this could be a desirable trait, up to some point. When enrobing, too high or too low of a yield viscosity can cause the chocolate to puddle, cause feet, or be too thick because there is insufficient self-weighted pressure to allow it to flow off properly or under a directed air stream.
In both cases, all is (usually) not lost. Sometimes though you might need to remake a chocolate.
Plastic viscosity adjustments.
Fat Concentration: Luckily this is pretty easy to address by adjusting the amount and/or type of fat in your chocolate. It might be the difference in 34% fat vs 36% cocoa butter or changing the ratio of cocoa butter to milk fat from 35/5 to 32/8. Generally speaking, the greater the fat content, the lower the plastic viscosity.
Emulsifiers: The addition of lecithin and PGPR (Polyglycerol polyricinoleate) can decrease viscosity with concentrations up to a maximum point at which point further additions do not have a continued effect. Studies seem to show an effect in the 0.1-0.6% range but that can vary depending upon chocolate formulation and water content. The main take away is there is a maximum useful amount in adjusting plastic viscosity.
Technically speaking, from a non-Newtonian fluid standpoint you could also increase the temperature but due to the constraints of tempering this is not really an option.
Yield viscosity adjustments.
The main culprit of a high yield viscosity is particle size. Basically, this is the dreaded over refining. Remember how great surface area increased how much pressure was required to move that glued boards? That is what happens with over refining. By decreasing the particle sizes past some critical level, the effective surface area increases until negative effects in yield viscosity are seen. The good news Is that short of refining very small amounts of chocolate (1 lb or less) it is relatively difficult to overrefine in current melangers. Mind you, it is not impossible, but it often takes many days.
Like plastic viscosity, the yield viscosity can be adjusted with emulsifiers. As you increase the amount f lecithin, the yield viscosity decreases....up to a point. Lecithin concentrations around 0.4-0.5% show the greatest decrease in yield viscosity; after that adding more causes a significant increase of the yield viscosity.
Added fat usually has a very minimal effect on Yield viscosity.
In both viscosity adjustments, it may not be possible to fully adjust the chocolate you have already made so that it is suitable for what you need. Sometimes you have make a new batch.
Water: This is not so much an adjustment you can make but a non-reversible effect. Water introduced by under roasted beans or ingredients that have taken up water can significantly increase Plastic and Yield viscosity that emulsifiers and the addition of fat may not be able to correct. If too much is incorporated, the chocolate will turn into a Rheopectic fluid that will only get thicker the more you work with it. In other words, it seizes.
Alright. I think that covers the basics. I realize it is rather frightening that that is only the basics. I’ve certainly taken liberties in nuance but hopefully it leaves you with more of an idea of what viscosity is and how you can work with it.
Summary:
Chocolate is a non-Newtonian fluid that can’t be described by one viscosity value.
Plastic viscosity and Yield viscosity defines how chocolate behaves rheostatically.
There is no perfect set of viscosity values. They will vary upon your needs (bars, dipping, enrobing, etc).
Plastic viscosity can be adjusted with fat content and emulsifiers.
Yield viscosity can be adjusted by particle size and emulsifiers but not fat content.
Both can be negatively affected by water content
Oh, right. There was a question up there I did not really answer.
Is it possible to estimate with a math equation?
Technically that answer is yes….have fun.
If you want something to really dig into about how all this changes with temperature and maybe melt your brain, try this