Using Melt Flow Rate to Determine Change in Molecular Weight 2

[RichGeoffroy]

If there is one key property for polymers, it is molecular weight --- actually it’s the presence of the very-high-molecular weight fraction. It’s these very long molecules which give the polymer its high elongation, toughness, impact strength, long-term creep resistance and resistance to stress cracking. In semi-crystalline polymers, these long-chain molecules get captured in one or more adjacent crystalline sections and act as links (tie molecules) between the crystallites. In amorphous polymers, the chains are so long that they get severely entangled with the other molecules. Unlike short chains that can easily slip through the entangled mass (a term we define as flow), the long chains are tied together such that pulling on one causes all the others to resist the movement. These long chains form the “mortar between the bricks” which gives the strength to the plastic part.

Of course, it’s the high molecular weight fraction which is also responsible for high resin viscosity which shows up as low extrusion rates, short injection flow lengths, and poor knit-line strengths. Also, when you “force” the material to flow, these long molecules which are carrying a disproportionate amount of the load, tend to break, therein, slightly reducing the overall molecular weight average.

When there is a failure, one often wants to determine if the molecular weight of the material is the same as that which was specified, or has it been altered during processing or in use. After all, small reductions in average molecular weight can be responsible for significant reduction in tie-molecule density, i.e., performance.

There are a number of methods by which one can determine different average molecular weights --- melt rheology, solution viscosity, membrane osmometry, gel permeation chromatography (also called size exclusion chromatography), and light scattering techniques. Each has its own advantages as well as limitations, however, the one most widely used is also the simplest and often the most sensitive measurement --- melt flow rate.

As discussed earlier, a very small change in molecular weight can result in a very significant alteration of a material’s properties. Most of the other methods measure properties of the polymer which are directly proportional to the average molecular weight. Small changes in molecular weight therefore can be lost in the normal variability of the measurements. Melt flow rate, on the other hand, is exponentially related to the average molecular weight of the material:

MFR =?M_w^3.4​

The power factor makes melt flow rate an extremely sensitive tool for measuring small changes in molecular weight.

Don’t overlook the power of simplicity.




Rich Geoffroy
Polymer Services Group
POLYSERV@aol.com

 

[jmw]

Viscosity is a good indicator of molecular weight.
One of the commonly referred to relationships is the Mark-Houwink equations where:
[η]=KM_w^a where K & a are constants and [η] is the dynamic viscosity.

This appears to be the form of expression referred to here as the MFI.

Melt flow index is often found using a variant of the capillary viscometer. Of course, a capilary is ideally suited to Newtonian fluids, which polymer melts are not, but they can be corrected. However, it is more usual to simply calibrate the output in terms of the MFI.

What is not said is that this relationship is very temperature dependent.

One of the main problems with viscosity measurement is accounting for temerature effects. It is necessary to discriminate between a change in viscosity due to temperature, from a change in viscosity due to a change in molecular weight. Changes due to temperature are often as significant, if not more so, than changes due to moelcular weight.

Most viscometer based methods usually depend on measuring the viscosity at a typical constant(ish) operating temperature and deriving an empirical correlation between the viscosity at that temperature and the molecular weight. This is a normal scenario for MFI sensors, especially in continuous production reactors where the temperature is usually sufficiently stable for measurement, a least, for control purposes.

An alternative approach is to take samples of the pellets or solidified polymer and disolve a known mass in a fixed volume of solvent and then measure the viscosity of the resultant solution. An accurate and controlled measurement but hardly suitable for reactor control as this is usually time displaced by as much as 8hrs.

Satisfactory alternative solutions to the MFI meters include rotational viscometers operating at the reactor efflux and measuring the melt in a similar manner to the MFI meters, though because they are suitable for non-newtonian fluids and have a controlled shear, some approach to the actual viscosity is achieved.

Whatever is used as an instrument, the dependance on constant measurement temperature restricts them to continuous reactions and to operating at quite high temperatures.

However, in batch reactions, such as in methyl methacrylate polymerisation end point spotting, digital viscometers have proven extremely good.

These measure the viscosity at the operating conditions and derive the viscosity at 40[°]C or 100[°]C or any other suitable reference temperature.
This base viscosity value is then correlated with the average molecular weight.

In this batch application temperature is critical. Initially heat is applied but as polymerisation progresses, the reaction becomes exothermic and the polymerisation rate tends to become exponential. Hence the need to account for temperature.

Tradditionally end point spotting depended on taking samples at frequent intervals and measuring the viscosity using a crude cup measurement where the efflux time is a measure of the viscosity (crude because accuracy is sacrificed for speed of measurement). These readings would then be used to plot a reaction curve from which the time at which to quench the reaction would be inferred.

Most viscometers lack the necessary accuracy for any temperature correction, however the digital viscometers have sufficient accuracy which, when combined with methods based on ASTM D341 (temperature Vs viscosity) can provide a sufficiently accurate measure of the reference temperature viscosity to enable accurate end point spotting.
These provide a continuous reading of the base viscosity and can be configured to show the MFI or molecular weight based on simple algorithms.

The end point is now detected rather than predicted.

In batch applications the process can take around 2-2.5hrs with only a 20sec window in which to quench the reaction. The ability to derive the reference temperature viscosity is crucial to the success of this operation.

The digital viscometer has also been successful in pipline continuous polymerisation reactions also.

Given the value of the viscosity relationship and the advances in the technology, many more processes can be controlled in this way.


JMW

http://www.viscoanalyser.com

Eng-Tips: Pro bono publico, by engineers, for engineers.

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[gambro]

I would like to point out that, due to the short length of the capillary die used in the standard MFI measurements, the MFI value not only reflects the shear viscosity at a certain shear rate (single-point shear viscosity measurement), but also reflects to a certain degree the so-called normal stress behavior resp. the elastic properties of the melt. In effect, the MFI expresses a conglomerate of various melt properties (shear viscosity, elongational viscosity, elastic effects). This makes is very difficult to deduct absolute values of molecular properties from the MFI measurement.

When comparing two grades of the same plastic type and assuming similar molecular weight distributions (MWD) for both grades, MFI can indeed be employed to give an initial, albeit not very refined, relative indication of the (average) molecular weight of these two grades.

Things become difficult when comparing two different plastic types, say PE vs PS, with very different MWD. Here, care must be taken to avoid false interpretation of results.

 

[jmw]

Good points, Gambro.

Many process viscometers are beneficial for their real time repeatabile measurements. Repeatability, under repeatable condition is a valuable measurement if correlated with laboratory measurements. A different correlation may be required for each specific fluid.

I don't know if the term "complex viscosity" can be applied to polymers as it is to hyper colloids

(

http://www.iop.org/EJ/abstract/0957-0233/14/4/308/)

but, as this artice suggests, some of the modern vibrational viscometers may offer more benefits than conventional viscometers, whether digital (bandwidth measuring) or analogue (amplitude measuring). It should be said that many are used rather more conventionaly.

This link to hydramotion demonstrates the principle for a high frequency oscilatory type (non-fluid displacing hence no density value and hence dynamic viscosity only):

http://www.hydramotion.com/tech.html

This is a similar viscometer (though I don't know what frequency it operates at, some only operate at around 50/60Hz)and we should include Nametre:

http://www.galvanic.com/nametre.htm

and this newer instrument:

http://www.viscositymeter.com/products.html

Some digital sensors are also displacing and can measure the density as well as the dynamic viscosity and hence the kinematic viscosity:

http://www.solartronmobrey.com/products/viscosity/index.php

to see how this works, refer to this link:

http://www.viscoanalyser.com/technology2.html

At high frequencies the apparent shear rate of vibraional viscometers is very high but for the same absolute viscosity this apparent shear rate can differ for different non-Newtonian fluids so again, the value of the laboratory measurement to calibrate the readings is important.

Even if we use a rotational viscometer, where the shear rate is controlled and constant, as a process instrument it also must have some flow shear to account for and it too benefits from cross correlation with laboratory measurements.

In any complex fluid viscosity measurement a controlled flow rate is important. Even though non-displacing vibrational viscometers claim no sensitivity to flow, they refer to the effect of flow on the sensor and not to the effect of flow on the fluid.

It would be nice to discover from users of these technologies, in polymer applications, what their experience is.



JMW

http://www.viscoanalyser.com

Eng-Tips: Pro bono publico, by engineers, for engineers.

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[jmw]

For anyone who doubts the value, even of relativey crude viscosity measurement, this article on Nylon manufacture illustrates just how valuable viscosity measurement realy is, especially in industries currently dependent on laboratory measurements.

I should add that while I am rather fanatical on the subject of process viscosity measurement and its benmefits, this is not an application that I have had any association with nor do I know which viscometer type or manufacturer has been used. It is just that too many people focus on cost instead of payback. These sorts of paybacks make these no-brainer investments. I personally have had experience of systems where the payback of much more extensive investment could be measured in weeks.

http://www.emcentre.com/unepweb/tec_case/chemical_24/process/p12.htm

JMW

http://www.viscoanalyser.com

Eng-Tips: Pro bono publico, by engineers, for engineers.

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.