At the highest frequency that is used in the frequency sweep or the lowest temperature used in the temperature sweep, run a strain sweep. Plot |G*| versus per cent strain and it will be seen that the value of |G*| is constant up to a certain value of strain after which it begins to drop. Choose the value of strain in the region where |G*| is constant - which is basically less than the critical value of strain after which the |G*| changes. Using this percent strain guarantees that you are generating data in the linear viscoelastic region.

When dealing with pastes (which essentially fall in the category of filled systems) getting reproducible data is a little more difficult than for fluids without any solid particles in them. The reason is that the fines in the paste can have agglomerates and the amount of agglomerates are often not consistent from one sample to another. A great deal of attention must be given to the preparation of these pastes and proper mixing methods and conditions should be used to get as close as possible to a good quality consistent mix. It is only then that the replicate data will be close to each other. Also, sufficient time needs to be given after gap closing in the rheometer to allow the residual normal stresses to relax completely before data is generated.

The following book may help in this regard to understand the various parameters that affect the rheological data of two-phase systems.

Aroon V. Shenoy, Rheology of Filled Polymer Systems, Kluwer Academic Publishers, Netherlands (1999).

Thank you very much Dr. Shenoy. Your insights have gained me a better understanding of the problem. There is more consistency in my results now.

Another method which is very similar to what Dr. Aroon Shenoy suggested is to run a oscillatory stress sweep (say 0.01 - 100 Pa) at the lowest temperature you intend to run a temperature sweep. At very low values of stress, you would notice a constant plateau in the graph of G' and G" against oscillatory stress. You may want to pay more attention to the G' curve. The constant plateau is the linear viscoelastic region. At some point, called the critical stress, the curve begins to go down, which signifies deviation from the linear viscoelastic region. Some authors have adopted a 10% deviation from the constant plateau to still be within the linear viscoelastic region. I speculate this 10% deviation to be because of the nature of their sample. You may not always see this constant plateau from experimental data.

Any information on particle size distribution and surface characteristics?

Hi Dr. Shenoy,

Happy to indulge in this discussion. If you have heard of this, please ignore it (as I don't wish to insult anyone's inelligence, while I often find it has been the solution).

The AIAG (American Industrial Aotomotive Group) back in the 80's came up with Measurement System (oterwise crudely known as GRR - Gage Reproducibility and Repeatability).

Reproducibility: Same Sample, Change operator to quantify output variation

Repeatability: Same Operator, Change sample to quantify output variation

In thi sparticualr testing given that it is a destrucive testing we would have to use what we call Cross-Technique.

So here are the factors that are playing against us when we measure for G' and G'' let alone viscosity measurements. Imagining a dynamic mechanical rheometer:

a - Temperature - Is a profile vs. a constant value across the test/sample area

b - Oscillatory Shear Rate Variation when kept constant

c - Operator error is almost insignificant as it is mostly programmatic

d - Torsion Measurement for Conversion to G' and G"

e - Since this is a destructive type testing, sample homgeneity

Please share your thoughts if you have looked up the GRR results and if you need any assistance I would be very happy to support with the test matrix, planning and interpretation and future steps to reduce it. This is one element of the lean six sigma methodology I have been practicing since the mid 90's.

Wishing you the best.

Cheers,

Dr. Nanguneri

Hi Dr. Nanguneri,

I am responding because your comments were addressed to me, though in reality, they should have been directed to the person who originally posed the question on the rheology of paste which is Valentine Cyprian Okonkwo. Since the dynamic shear rheometer is a well-established device for the measurement of rheological properties, most of the concerns (a – e) are generally taken care by the equipment manufacturer.

In the world of high sophistication and advanced electronics,

(a) there is utmost care taken to attain and maintain constant temperature across the sample area and there is an implicit guarantee that the variation has been minimized during data generation

(b) the input signal is well controlled to provide almost nil variation in the oscillatory shear rate when it is intended to be constant,

(c) most part of the experimentation is programmed in and computer-controlled in order to minimize human error,

(d) torque heads are designed to be robust to provide accurate torsion measurements for conversion to G’ and G’’ and other rheological parameters

(e) the idea of staying within the linear viscoelastic region is to make sure that the testing is not destructive to the sample during the measurement.

My response to the original question put forth by Valentine Cyprian Okonkwo was on the presumption that the repeatability issue that is being mentioned is not due to the operator error because reproducibility and repeatability would have been checked before posing the quandary. My answer therefore pinpointed the issue to the nature of the sample being tested. As for any assistance from you in this matter to provide support with the test matrix, planning and interpretation as well as future steps to reduce it, this is for Valentine Cyprian Okonkwo to decide since he is the one who had the repeatability problem and not me.

With best wishes

Thank you, Dr. Shenoy and it was my oversight on identifying who the original requetor was and my sincere apologies, however, great to see you response.

Back in 1990's while I was also responsible for acquiring rheological equipment for a global employer, I rarely found studies on GRR and vendors of such equipment claimed everything, instead produced nothing and were also reluctant to engage in an experimental effort. Nobody has published the GRR results as per the AIAG standards, yet want to claim they have done it.

Sorry to say, that simply doesn't work from a measurement system pointof view.

I then found one vendor willing to do it only because he knew then what we were talking about and clearly saw the effort as an advantage to his sales revenue as he could pitch that as a unique value proposition.

In my note, it is clear that operator error is irrlevant as most of these instruments are automaic (even in the 90's). I already shared that sample homgeneity is a possible source of error as it is a destructive type of testing.

I will copy and paste my feedback to Valentine Cyprian Okonkwo, the owner f his query. Have a wonderful day and am humbled you took the time to scribe your thoughts even if it was for someone else.

Thank you.

Best,

Dr. Nanguneri

Hi Mr. Okonkwo,

Excellent question.

Have you had a chance to review the discussion beween Dr. Shenoy (@Aroon Shenoy) and myself over the last two exchanges? If you have, please let me know of your thoughts.

I mistakenly responded to him without addressing it to you the original owner of this query.

Have a pleasant day.

Best,

Dr. Nanguneri