Rotational rheometers have great functionality and are arguably the most versatile rheology characterization tool available. They are used for a broad range of sample types and can be configured for a number of different rheometric methods. Two of these methods are measuring the viscosity and viscoelastic properties of a material.
Recently, we set up a two part webinar series, with the goal to give a basic introduction to rheology and rheometry Part 1, presented by Technical Specialist Philip Rolfe, was about viscosity and viscometry. Part 2, presented by Dr. Shona Murphy, focussed on viscoelasticity. A great audience joined this two part complimentary web seminar series. Thanks to all for your participation! As usual, we have made the recording available on the website for those who could not make it for the live session You can watch the recordings on our website at any time here:
Many questions were asked during and after the webinar, and these are listed below along with answers for your interest. If you have any further questions, please don’t hesitate to contact us directly here.
Q: Could you please explain the slip test again? Do I put a sample between two glass slides and see the trail?
Yes, although this isn’t really a rigorous or official test method. But, if you place the sample on a clean piece of glass or metal and tap it vertically on a bench, does the sample flow down or does it slide down? If it slides down this is a clear example that it will likely slide between smooth geometry surfaces. Here’s a photo of hand cream that slips – you can clearly see that the thinner layer of the cream on the surface at the top did the bulk of the shearing, while the rest is moving as a unit. This sample would therefore need to be tested with roughened or serrated plates to eliminate wall slip and measure its bulk rheology properties.
Q: Which system would you use for cement and mortar?
One can use several different geometries for cement and mortars, depending on the particle size of the sample:
– If very fine, we can use roughened or serrated plate geometries (1) to prevent slipping.
– If coarser, we may need to go to splined cup and bob geometries (2).
– Next would be a small vane in a large cup (3).
– And lastly the twin orbital ball geometry in a large cup (4).
Q: What will be the surface roughness of the plate to eliminate the slip flow? Has it constant roughness?
The roughness of the surfaces needed to eliminate slip, depends on the sample type. If the sample phase separates a lot, i.e. exuding a lot of water when a shear stress is applied, as with tomato ketchup, then a very rough surface is needed to accommodate that volume of water. Similarly, if the sample is like a hand cream or grease, then even distribution of the stress onto the sample requires roughened surfaces to prevent surface slippage.
Normally, tests are initially made with smooth geometries and if a double knee is seen in the flow curve (second area of shear thinning) this is an indication that the sample might be slipping. If the use of a roughened and then serrated plate geometry eliminates that double knee, it is clear that the sample was slipping, but that this has been resolved with roughened plates. So, we would use the smoothest plates that we can, while still eliminating the slip.
Malvern Panalytical have a well-controlled manufacturing process so that the geometries should have even roughness across the surfaces. Serrated plates are machined to give a serration depth of ~600 µm.
Q: For the plate surface roughness, how rough a surface can you use before losing accuracy?
If the sample is truly slipping, then it is better to use a roughened surface and lose some gap accuracy, than to get completely artifact data from slippage in most cases. I would say for most samples a roughness equivalent to 150-200 grit sandpaper is good, or maybe even rougher for phase separating samples.
A sample that slips will have an apparent yield stress, as there needs to be a stiffer body that resists flowing, but when stress is applied, either it is dramatically shear thinning and part of the sample starts to flow easily, (eg brittle gel, hand sanitizer), or it separates to form a slipping layer, (eg. tomato ketchup). The key is that the surface roughness needs to have sufficient void volume to accommodate the liquid that separates from the bulk (in the case of the phase separation) or enough surface area spread the stress sufficiently into the sample before the material yields.
Q: I am measuring yield stress for a gel at 20 °C. I think what I am getting is the data from when the plate slips, so is it just a case of having rough surfaces on my plates?
Yes, this would resolve the slip issue if sufficiently rough and give you the true yield stress of the sample.
Q: I need to predict if I am going to have laminar flow problems. How can I do it?
Other than sample slip (which I described above) other non-laminar flows include turbulence at very high shear rates and the Schrag gap loading limit for high frequency oscillation.
The onset of turbulence at high shear rates calculations vary depending on the geometry type being used but generally can be calculated using the Taylor and Reynolds numbers. For cup and bob geometries, generally, when Ta exceeds ~1800, turbulence is likely (1, 2, 3).
For cones or parallel plates the critical Reynolds number should be less than 3, i.e.:
1. S. Chandrasekhar, Hydrodynamics and Hydromagnetic Stability, Dover, New York, 1961,
2. P. G. Drazin and W. H. Reid, Hydrodynamic stability, Cambridge University Press, 2nd Ed.,
Cambridge, England, 2004, 69-123.
3. P. Chossat and G. Iooss, The Couette-Taylor Problem, Springer-Verlag, 1994.
Q: How do I determine the best geometry to use, plate or cone?
There are different best geometries to use for different sample types. In an ideal world, we’d use a cone and plate set up for viscometry, although the sample may become turbulent at high shear rates if too low viscosity, in which case I’d use a parallel plate and a narrower gap which will prevent this. If the sample has large particles and very high shears aren’t necessary, obviously we need to avoid jamming or boundary concentration depletion, so a wider gap geometry is used such as plates with a wider gap, or a Couette system. A review of the presentation again may give some good pointers, as we did discuss this a little.
Q: You sometimes showed shear rate data and at other times shear stress data – where would each best be used?
If the rheometer is being used to simulate a rate driven process (e.g. pumping through a specific diameter pipe with a controlled flow rate, or coating with a known diameter roller and thickness, etc) then a controlled shear rate test makes the most sense. If the process is controlled stress, (e.g. sagging, pouring or sedimentation, yield stress etc), then a stress-controlled test may seem more appropriate.
Q: Can you briefly explain “complex viscosity” and why it is useful?
Complex Viscosity (more correctly, Shear viscosity (complex component)) is dimensionally the same as the steady shear viscosity, but the measurement is made in oscillation mode with increasing frequencies, instead of rotational steady shear with increasing rates. This is particularly useful if a sample doesn’t stay within the gap for very long during a steady shear experiment, as happens with stiff polymer samples or semi solids.
For samples that aren’t highly filled, there is an empirical relationship known as the Cox-Merz rule which relates directly the complex viscosity and the steady shear viscosity, if the angular frequency and the shear rate are the same. i.e.:
Q: What are the main uses for creep studies?
Creep studies are useful for characterizing the viscoelasticity of a sample over longer timescales. Historically, creep testing was carried out as it could generate very precise data on even low-resolution instruments due to the larger strains incurred in the length of the test. One of the main applications of the creep test was to apply low shear stresses / rates and determine the properties when a sample reached steady state flow. With modern rheometers however, we can apply low shear rates / stresses very easily, and use the live data to determine when a sample reaches steady state flow. The approach to steady state is recorded (similar to a creep test), but repeated for every point of an equilibrium flow curve.
Creep does, however, have its uses with modern rheometry, some like to use it because it can be used to simulate certain processes / situations for example, for simulating things like slump where gravity is acting on the sample.
Now we have a high-resolution test with a high-resolution rheometer. The small stresses that a modern rheometer can now apply are a lot smaller than what can be naturally be stored in the sample (stored energy) which means we can just be measuring this noise at this energy is released from the sample. So, the stress we apply needs to be larger than the stored energy/stresses in the material so care must be taken is selecting appropriate conditions
Instead, these days we will typically use oscillation measurements as a good alternative to creep. The fact that the geometry is oscillating in both directions makes it even more sensitive to any changes occurring in the sample at rest. We can use the phase angle to give us information if the sample is likely to flow/deform (high phase angle) or maintain its shape / deform less etc. (low phase angle).
Q: Do you prefer parallel plate or cone for oscillation test?
This is a good question. The advantages of a cone and plate geometry are, that no matter where the sample is in contact with the surface of this geometry, it will be exposed to the same amount of deformation. However, in Small Amplitude Oscillatory Shear (SAOS), the sample is kept withing the linear viscoelastic region and therefore the stress is always proportional to the strain at any diameter under the parallel plate anyway, so the benefit of the cone is reduced. The advantage of a parallel plate system is, that gap can be changed according to the nature of the sample (small gap – for low viscosity, large gap – for high viscosity) or if the sample has particles present.
As a general rule of thumb, we usually recommend that the measuring gap size should be set to around 10 times that of the largest particle in your sample. When pushing to work at higher frequencies, using a plate might be advantageous as it means the measuring gap can be reduced. This supports avoiding the higher frequency Schrag gap loading measurement limit which in turn enables us to achieve accurate results at higher frequencies.
Q: Can you please comment on utility of doing oscillation measurement if LVER is not clearly discernible, such as having a low G’?
If a material has a low G’, for example a highly viscously dominated sample then we can use the G” or phase angle to assess the LVER (Linear Visco-Elastic Region).
Often these sorts of materials can possess a long LVER. In these instances, we can select a higher strain (still within the LVER) to use for other oscillation test (such as a frequency sweep or single frequency) so that we obtain nice smooth data. Although this depends on the nature of the sample and the oscillation test used, for example during a single frequency measurement to monitor a sample curing, the LVER would be expected to shorten during the test, so the strain selected would need to be low enough to accommodate for the sample changing. However, the rSpace software is flexible enough to allow us to transition between selecting different strains in accordance to the sample changing.
If the material is completely viscous, such as with a standard oil, the Cox -Merz rule can be assumed and either viscometry or oscillation tests can be implemented. See above question: “Can you briefly explain “complex viscosity” and why it is useful?”
Q: Are there some assumptions made when you model spraying?
Yes there are several assumptions… Firstly, many consumer products are packaged in tubes or bottles where product application involves pumping or spraying the product through a nozzle. Such products tend to be shear thinning products where the viscosity drops during the extrusion process due to the increasing shear rate, and then recovers on exiting the orifice as the shear rate is reduced.
The shear rate encountered during this process is related to the radius, r of the orifice and the volumetric flow rate Q by the following expression:
The parameter n is the power law index, which is 1 for a Newtonian liquid and between 0-1 for a non-Newtonian fluid. This value can be readily attained from a variable shear rate test by fitting a power-law model to the resultant data.
Secondly, rheology tests usually only consider the viscous, viscoelastic, time and temperature dependent properties of the sample, but not the solvent loss during spraying that a highly volatile paint for instance may experience. So while the thixotropy of the sample may indicate that the coating could sag or even run after being sprayed onto a non-absorbent substrate, if the material flash dries, this may not happen.
As a samples drying time is weather dependent (varying with atmospheric humidity, temperature and air flow), it’s a rather variable quantity, so this is difficult to reliably replicate. One may also say that a well formulated coating should work well in a variety of conditions, so formulating the rheology to work irrespective of evaporation is a good approach.
Q: Can you please recommend a good oscillation standard other than the PDMS
The principle instrument calibrations (i.e. torque, gap, temperature, displacement) are all calibrated robustly. PDMS can be used as an oscillation verification which has been found to give very good, reliable data when needing to do an instrument check. The loading conditions for PDMS are particularly important, but this is one of the reasons why Kinexus uses configurable sequences to carefully control, and ensure all users do the same procedure each and every time (the loading conditions).
Although not a UKAS/NIST material, our sample of PDMS does enable the consistency of the instrument (and operator loading). It is available to purchase from the Malvern Panalytical Web store.
The difficulty in using a lower viscosity sample to verify oscillation values is that there is a higher variance in the inertia when using lower viscosity samples which if incorrect will affect the measured phase angle, particularly at high frequencies. We use a UKAS certified 1Pa.s oil to verify this in oscillation.
A 1 Pa. s oil (or any Newtonian oil) should have a linear viscosity and have a 90-degree phase angle at all frequencies. (This can also be purchased through the Web store)
- A basic introduction to rheology
- Overcoming and quantifying ‘Wall Slip’ in measurements made on a rotational rheometer