Volume 1 Number 11 Dennis R. Dinger 1 September 2003
An Update
Paperback and downloadable copies of Rheology for Ceramists and Particle Calculations for Ceramists are available at the Books and Downloads page of the web-site. Quantity discounts are available on the paperback books. If interested, please contact me for details.
Please forward any issues of the E-zine, or my web address, to any ceramists or materials engineers who might be interested. If this is the first issue you've seen, you can add your name to the mailing list by clicking HERE. All back issues can be accessed from the Publications page at the web site. Questions, suggestions, and/or requests for topics to be covered in future issues of this e-zine can be sent to QuestionsandComments@DingerCeramics.com .
The topic of this issue was a direct question from a reader. I expect it will apply to many of you.
"If I Could Buy Only One Viscometer, Which Do You Recommend?"
My Experience
To answer this question, I first need to describe my experiences with viscometers. First, I'll mention the viscometers that measure dynamic viscosity. On the coal/water slurry project at Alfred University years ago, we had a Haake Cup-and-Bob Viscometer and we tested an oscillating sphere viscometer. At Alfred University, Clemson University, and in many production companies, Brookfield viscometers were in abundance. I have also seen Gallenkamp-type viscometers used routinely in production companies. And over the years, I have tested some cone-and-plate viscometers.
Regarding the measurement of kinematic viscosities for fuel oil viscosities, I have used Saybolt viscometers. I have also seen other similar devices used in the ceramic industry (e.g., a cup with an orifice in the bottom mounted on a long handle.)
My apologies to the viscometer manufacturers whose brands were not mentioned by name. But these that I mentioned are the only ones with which I've had extensive experience. To those of you who are actually looking for viscometers for purchase, a search of the internet will bring up a variety of brand names in each category of viscometer. Do not ignore any other brands simply because I haven't mentioned them. Throughout the remainder of this article, I will speak of generic types of viscometers as much as possible.
Based on this, some may say that I haven't had much experience with a wide variety of viscometers. From some points of view, that would be correct. But the viscometers that I've mentioned have served me well over many years.
Several General Categories of Viscometers
The viscometers that are commercially available today fall into three categories. I will discuss each of the three below.
Research Tools
The most expensive viscometers are the ones I consider to be research tools. This includes concentric cylinder (cup-and-bob) and cone-and-plate viscometers. There are also some instruments in this category that have been designed to measure viscosities of high-solids plastic forming bodies. I consider 'expensive' to refer to viscometers that are $15k-$20k and higher. If your company has a substantial R&D lab facility, one or more of these viscometers can be a valuable tool for defining and understanding the fine details of body processing properties.
These types of viscometers have the precision that allows you to study the fine effects of particle size distribution, surface areas, mineralogies, particle morphologies, and chemical additives on rheological (and therefore processing) properties. If you have good particle size analyzers and SSA instruments in your lab, a research viscometer will be a useful and helpful addition.
Generally speaking, these types of viscometers have the capabilities to measure high shear rates (considerably higher levels than typical process viscometers) which is a valuable feature in their favor. Both the cup-and-bob and cone-and-plate viscometers are designed to produce constant shear rate within the whole measurement volume. For example, when the rpm is such that a shear rate of 100s-1 is produced, all of the suspension in the measurement volume should be subjected to that single, constant shear rate. Measurement shear rates increase as rpm increases and as the shear gap (the distance between the cup and bob, or the angle between the cone and plate) decreases. Some of these instruments can reach 1000 rpm (and higher), and when the shear gaps are small, shear rates can be really high (several thousands of reciprocal seconds).
I must mention, however, that in these types of viscometers, some rules are broken and short cuts are taken to accommodate particle/fluid suspensions. Such necessary adaptations prevent the measurements from being exactly as described. When the cup-and-bob gap is really small, or when the cone angle is really small, the whole measuring zone does see relatively constant shear rates (as mentioned above). But in ceramic suspensions, we want to measure the viscosities of fluids containing solid particles rather than simple fluids or mixtures of fluids, and particles generally don't like extremely small gaps or cone angles. When gaps are small and particles are large, particles can span the gap and cause problems. The same is true when cone angles are small. In such cases, cup-and-bob gaps are increased, cones are pulled back a distance from the plate, cones are truncated, and/or cones are replaced with flat discs, all of which alters the geometries, deviates from the simplicities that produced the constant measuring volume shear rates, and ruins the constancy of shear rates within the measuring zone. Such alterations probably produce only minor effects and minor deviations from constant shear rates, but when you're thinking that all of the sample is subjected to a constant shear rate when it's really not, that can affect the veracity of the results. This should be considered when purchasing or using such a viscometer system.
Depending on the way the instrument is designed, one might be able to measure low shear behavior and learn about yield stresses. But if the instrument reports the rpm of the drive motor, rather than the instantaneous rpm of the cone or bob, it is not guaranteed that the two rpm values will be constant. One would think they should be, and in most cases, they will be the same, but under acceleration, they will not be the same. I tested this on a cup-and-bob viscometer by holding the cup stationary in my hand while starting a slow measurement program (a slow, constant acceleration from 0 to some nominal value). As soon as the run started, the rpm display showed a positive value that was increasing (consistent with the program controller) but the bob was actually stationary in my fist (i.e., 0 rpm). I was simulating an excellent gel structure which would hold the bob stationary against the viscometer's attempt to move it. The results showed that a good gel structure that actually did hold the bob stationary would not be shown in the actual data. Apparently, the assumption was made that the difference in rpm between the bob and the drive motor would be negligible at all times, but in particulate suspensions under certain test conditions, this is not necessarily true. This type of design is understandable because it's certainly easier to monitor the rpm of the drive motor than the instantaneous rpm of the bob. And under steady-state conditions, the two should always be the same. This, too, should be considered -- especially if you are interested in the precise measurement of yield stresses.
Regarding viscometers for measuring high solids plastic forming bodies, they generally don't have the capability to impose constant shear on all of the sample body in the measurement chamber. Sigma blades, or a variety of other shapes of mixing blades, are used frequently in such instruments. Although they measure the relative ease or difficulty to stir and/or mix the body at a variety of shear rates, they don't impose constant shear on all body within the measurement chamber, nor can they reach particularly high values. These viscometers and rheometers will, however, measure rheological properties of plastic forming bodies because they can be run at a variety of shear rates (rpms) which is the primary requirement for measuring rheological properties.
Routine (Daily) Measurements
Suspensions
The infinite-sea type of viscometers (such as most Brookfields and their competitors), are valuable for performing routine, daily, process measurements. They consist of a rotating spindle that can be immersed in a beaker of suspension. The applied shear rate is highest at the spindle and it drops off towards zero at the beaker surface. Shear rates are not constant within the measurement chamber in this type of viscometer. Although this type of viscometer is well-suited for daily measurements, they can also be used for research. They are generally less expensive than the research instruments mentioned above, and they hold up well in the plant environment.
These types of viscometers tend to be limited to relatively low rpms (i.e., low shear rates), but they do measure dynamic viscosities (not kinematic viscosities). The various spindles define different complex geometries with each different measuring cup or beaker used, so for consistency, each lab should define a single measuring cup size that will always be used for all routine viscosity measurements. Since spindle/beaker geometries are complex, it is difficult to calculate the exact shear rate that corresponds to a given rpm value. Since it's difficult to do, I recommend you don't even try. When you take a measurement, just report the spindle type, the container in which the measurement was made, the rpm value, and the measured stress value or the viscosity. The fact that the actual instantaneous shear rate is not known is not a problem because the particular spindle, rpm, and beaker combination should reproducibly produce the same shear conditions again and again. This is sufficient specification of the measurement conditions so the test can be duplicated at a later time. We routinely use milkshake mixers (which produce high intensity dispersion conditions [HID] at medium and high speeds) prior to viscosity measurements, and the stainless steel milkshake mixer cups work well as viscometer measuring cups.
Some of the newer viscometer models of this type are computer controlled. With these models, we define several viscometer programs to be used routinely to measure both gelation behavior and rheological properties. A gelation program is simply a 20 minute run with the first 10 minutes at 100 rpm and the last 10 minutes at 1rpm. Result charts are plotted as viscosity versus time.
To measure rheology, we define a program that changes the rpm every 30 seconds (or some other constant increment such as every minute) from a low value (for example, 1 rpm) up to the highest rpm possible on the instrument. The rpm values should follow a geometric series (the ratio between each pair of adjacent rpms should be constant). This allows the rheology test results to be plotted as viscosity versus shear rate (rpm) on log-log axes.
This type of viscometer is frequently limited to relatively low maximum rpm values. Some of our older viscometers were limited to 60rpm max. Newer models can reach somewhat higher rpms. When limited in this way, it is difficult to measure viscosities at the shear rates that correspond to the onsets of dilatancy. For most suspensions, you want the onset of dilatancy to occur well above 100 rpm. So when you are dealing with a 'good' suspension and a viscometer that is limited to relatively low shear rates, you won't see the onset of dilatancy. It is good to know that your system is not dilatant within the range of the viscometer's measurement conditions. But you will be flying somewhat blind because you don't know exactly where they do begin. When you measure rheological properties as suspensions are being deflocculated or as solids contents are increased (or both), you will be able to see the dilatancy even at the low shear rates of this type of viscometer. And when you see dilatancy on these viscometers, you have reached conditions that will surely cause processing problems in the plant.
Also, don't believe some people who teach that "ceramic suspensions are generally Bingham in nature, and since Bingham rheograms are straight lines on shear stress vs shear rate plots, you can extrapolate the straight lines up to the shear rates of interest." This is baloney. Dilatancy is cause by particle/particle collisions. Just because collisions are not very substantial at 100rpm doesn't mean they won't dominate at higher rpm values. With viscometers that are limited to relatively low rpms, you can simply assume that viscosities at higher shear rates will (at best) be worse than any extrapolation would indicate. (See the chapters on dilatancy in Rheology for Ceramists if you're interested in more details of these types of phenomena.)
Viscometers in this category are less expensive than the research models in the first category. A model with a digital read-out that is used in the plant is essentially the same as a lab model that is computer-controlled. In fact, they will probably look similar. One will have a computer communications cable coming out the back. The other won't. But all else will be similar. You can probably buy several of this type of viscometer, including a computer-controlled model for the lab, for less than the cost of one research viscometer.
Fuel Oils
If you must routinely measure the viscosities of fuel oils, a Saybolt viscometer is recommended. These are available from the chemical supply companies. A Saybolt viscometer is simply a cylindrical container, with an orifice in the bottom of it, mounted in an oil bath. A cork blocks the orifice while the sample is heating. To measure the viscosity, the cylinder is filled with the fuel oil to be measured and the oil bath temperature is set to the desired temperature. When the sample has reached temperature, the cork is popped from the orifice simultaneously with the starting of a stop watch. The viscosity of the oil is equal to the number of seconds for a fixed volume to flow through the orifice. The greater the number of seconds it takes for the constant volume to flow, the higher the viscosity of the oil.
This type of measurement is a kinematic viscosity measurement. Kinematic viscosities are related to dynamic viscosities by the fluid density. Kinematic viscosities correspond to a variety of shear rates from higher at the beginning of the measurement to lower at the end. The 'head' which is the pressure produced by the height of fluid above the orifice is the only force causing the fluid to flow through the orifice. As it decreases, the flow rate through the orifice and the shear rate in the orifice decrease as well. In this type of viscometer, shear rates start relatively low and decrease even lower (towards zero) as each measurement is taken.
This is a standard measurement method for fuel oil viscosities. It is consistent from day to day, and once the viscosity at one measurement temperature is known, the viscosities of that same fuel oil at other typical process temperatures can easily be predicted.
It is not a good test for ceramic suspensions, however.
Spot Checks
The final category of viscometers are those used for what I consider to be 'spot checks.' Certainly, all viscometers mentioned in this article can be used for 'spot checks' but there are some that give little data other than a quick-and-dirty spot check.
These devices can be as simple as a stainless steel cylinder with a hole in the bottom, mounted on a relatively long handle. Such a device is similar to the Saybolt viscometer mentioned above for fuel oil measurements. With such a device, you reach into a tank, pull a sample into the cup, and time how long it takes to empty through the orifice. I'm sure some plant personnel swear by this type of device. I would feel more comfortable if process measurements were made on a lab viscometer, but if your people have, use, and like such devices, and their techniques calibrate adequately with lab measurements, I don't see any fundamental problems.
Such devices, however, are not of any value in a lab. They don't tell anything about the rheological properties of a suspension, and with suspensions, rheology is important. Such devices, used as a spot check, may show that today's viscosity in the process tank is the same as yesterday's, but it will not give any indication concerning whether the suspension today is shear-thinning or shear-thickening. It simply says, "The current viscosity is OK," or "The current viscosity is high," or "The current viscosity is low."
Another spot check is produced by Gallenkamp-type viscometers. These devices are used to show gelation behaviors after a certain length of time. In this type of viscometer, a bob is suspended in a sample of suspension. The bob is suspended on a length of spring wire, and the suspending wire is wound up 360degrees from its neutral point. After the desired length of gelation time, the bob is released by pulling a stop pin, and the bob unwinds by the 360 degrees and overshoots by a certain number of degrees. As gelation increases, the number of degrees of overshoot decreases.
The only drawback to this type of measurement is that it gives only one point of measurement. It doesn't show whether gelation is increasing or steady at the point of the measurement. If gelation increases quickly to reach a constant gel structure (representing 'good' gel properties), or if gelation starts very slowly and then climbs dramatically, or if it builds at a relatively constant rate, the reading after the designated time can be the same in each of these cases. These three gelation behaviors are all very different, but they all could produce the same reading on this type of device. In this way, the Gallenkamp-type viscometer gives incomplete information about gelation -- it gives a single spot check on the gel structure. But they are used routinely, and successfully, in many labs where gelation behavior is important.
Instead of using a Gallenkamp-type viscometer to measure gelation behavior, I recommend the 20 minute gelation test (10 minutes at each of two rpms) mentioned above.
My Recommendation
If I had only enough money to buy one viscometer, I'd buy a computer-controlled infinite sea viscometer (from Brookfield or one of its competitors). My experience has shown that such a viscometer will
(1) Measure viscosities over a wide range of values using the different spindles,
(2) Measure rheologies over the range of shear rates defined by the min-to-max rpm range of the instrument, and
(3) Measure gelation behaviors at constant shear rates with time.
With these capabilities, I could use the instrument to gain some information desired for R&D, and I could use it for daily process measurements of viscosity, rheology, and gelation behavior. As you can see, I would use it to provide some information in all three of the categories mentioned above. And all of this comes at what I consider to be a reasonable price for any ceramic production company, whether large or small.
Another way to look at my answer is this. I wouldn't buy the cheapest viscometer available (such as those for 'spot checking'), nor would I buy the top of the line model with all sorts of bells, whistles, and the capability to make the morning coffee. I'd buy the one with the greatest range of application at the most economical price. Then, later, if I needed to buy a similar model or a spot checker for the production people, or a top of the line model for the R&D people, those are always options. But between now and then, a lot of useful information would have been collected which can be factored into the decision process for any proposed new instruments.
Miscellany
I assume, as with everything else concerning fluids, that most viscometers were originally designed for simple fluids. Their application to particle/fluid suspensions came later. The differences between simple fluids and particulate suspensions is the subject of one of the chapters in Rheology for Ceramists. The book also contains a brief discussion of the different types of viscometers. If you're interested in such topics, the book will be a valuable addition to your personal library.
I continue to look for suggested topics for future columns. I look forward to hearing from you.
See you next time. Thanks.
Thanks for signing up to receive Ceramic
Processing E-zine
To unsubscribe, click the "Remove Me" link below.
Copyright © 2003 Dennis R Dinger
103 Augusta Rd, Clemson, SC 29631 (864) 654-3155
All Rights Reserved.
All Rights Reserved.