Mooney viscosity measurement without mill massing.

Mooney viscosity measurement without mill massing

Measurement of viscosity with the Mooney instrument is still the most important measurement in the release and acceptance testing of rubber polymers.

The main thing required of test methods used in release testing is that they should have good repeatability and, above all, good between-laboratory reproductivity.

Measurement of viscosity with the Mooney instrument is also favorable in several other respects, such as the fact that it is in general use, that it has been standardized, and that it is rapid and not very costly.

A Mooney viscosity value is the result of two consecutive steps: sample preparation and the actual measurement with the Mooney instrument.

If there are differences between the results of different laboratories, the cause generally lies in one or both of these steps. These test errors can then cause apparent production problems.

It is in the interest of product manufacturers and processors to reduce these chances for error. The purpose of this article is to examine if omission of sample pre-treatment will result in a substantial reduction of test errors.

Reason for sample preparation

Preparation on the mixing mill, to which one was accustomed as the first step in mixing, was introduced in the early days of rubber technology as a means of making samples homogeneous. The milled sheet was compact and therefore a favorable source of Mooney specimens, regardless of supply forms (bales, chips, granules, etc.). In addition, the production of thin sheets facilities drying (in the case of rubber containing residual moisture) and visual inspection for impurities and inhomogeneity.

Finally, milling reduced the polymer viscosity, which is necessary if very hard rubbers are to be measured in the Mooney instrument.

These problems are today largely non-existent, because rubber polymers are now very homogeneous and their viscosity levels so specific that the only remaining question in Mooney viscosity measurement relates to the filling of the chamber by the various supply forms of rubbers.

Problems of preparation on the mill

For decades it has been customary to prepare samples for Mooney viscosity measurement on laboratory mixing mills. Directions for this exist in company specifications and standardized procedures (table 1).

For Bayer, as a polymer manufacturer, two methods have been important until now:

* The Bayer method, which in the case of CR is practically identical with the standardized procedure;

* The standardized procedure (DIN, ASTM, ISO).

The Bayer method has generally been used for polymer release testing.

Material changes caused by milling

Milling basically causes polymers to undergo thermomechanical changes. Investigations[1] have shown that these changes may take the form of positive or negative viscosity shifts. As a processing test, this observation is an interesting by-product of the Mooney measurement procedure; if, however, the mechanical and methodical parameters are not strictly adhered to at this stage, these alterations must be looked upon simply as sources for error, the result being the aforementioned differences between individual laboratories[2].

The dependence of Mooney viscosity on milling is shown in the case of chloroprene rubber in a three-dimensional diagram (figure 1) in which the roll nip, temperature and number of passes appear as parameters (from ref. 1).

Figure 2 shows how different polymers can react to the various preparation techniques.

In this figure the levels of the Mooney viscosities given by the DIN and Bayer methods have been normalized with reference to those determined after vacuum compacting. It should be added that similar results are obtained with and without compacting. It will be seen that the differences between the values determined after vacuum compacting and those determined after preparation by the other methods range from + 15% to -25%.

Method- and equipment-related parameters

Different milling methods have different parameters, the maintenance of which may be critical or less critical in release and acceptance testing according to established polymer specifications.

The method-related parameters are: roll speed; roll temperature; sample quantity; number of passes and foldings; and subsequent storage.

Equipment-related parameters (roll diameter; roll length; deviation from cylindrical shape [caused by wear]; roll surface finish; and nip width, non-parallelism) must also be mentioned here. Whereas adherence to method-related parameters is a question of internal plant organization, some of the equipment-related parameters present more difficult inherent problems. The first difficulty, which we have encountered repeatedly in cross checks and discussions with other laboratories, is the availability of the correct or standardized roll size and the adjustment and controllability of the various parameters. These requirements are often not satisfied in practice. Another problem is that of the roll nip: although nip width measurement with lead slugs is standardized, it does not generally reveal the nip width that actually develops under the force exerted by the passage of the material and hence the nip specified for sample preparation.

Numerous measurements have shown that every mixing mill has individual nip enlargement characteristics. Bearing and axle play, roll bending and the load-bearing capacity thickness of lubricants give different actual nip widths at given forces exerted by the sample, even where the rolls are equal in size and design.

Figure 3, depicting two laboratory mixing mills of equal size and design, shows the nip enlargement caused by differing loads, through lead slugs of varying cross-sectional areas.

The two mixing mills just mentioned differed not only in nip enlargement but also with respect to roll surface finish quality, the rolls of one mill having been recently refinished. Figure 4 shows how these differences between the mills affected the Mooney viscosity readings obtained after preparation of a chloroprene rubber specimen.

The graph shows that mixing mill A had to be adjusted differently from mixing mill B to obtain the same Mooney viscosity readings. Tests with other rubbers have shown, however, that this procedure cannot be generalized.

It is clear from this example that, at the present time, different laboratories can hardly prepare samples in an exactly uniform manner on their existing laboratory mills. On the other hand, we do not consider the manufacture of high-precision laboratory mixing mills only for this purpose as justifiable.

Time and costs

According to the standard (DIN 52323, Part 1) the milled sample must be stored for at least 30 minutes before its Mooney viscosity is measured. This delay is particularly inconvenient in production control. The maintenance of mixing mills in compliance with the standards, and performing the actual milling procedure, require considerable time and expense.

Alternatives to milling

As alternatives to milling, the following will be considered: vacuum compacting and omission of pretreatment (except in some cases minimal compacting).

Vacuum compacting

Prompted by experience gained in the production of samples for Defo testing[3], Bayer has examined whether vacuum compacting is also suitable for preparation of samples for Mooney testing.

The rubber is cut into slices, preheated for several minutes to 105 [degrees] C, freed from air under vacuum and compacted by a piston in a cylindrical chamber.

The result is samples that are largely free of trapped air and intrinsic stresses. The method has already been described in detail[1] and now constitutes a part of a draft standard (DIN 53529, Part 1).

Omission of preparation

As vacuum compacting likewise entails delay and some expense, attempts were also made to eliminate preparation completely except for minimal compacting in some cases.

Polymer chips can be placed in the chamber of the Mooney instrument directly. Bales can likewise be tested directly after being sliced. Slight compacting can be applied to granules and powdered rubber: the granules or powder are preheated for five minutes in an oven at 50 [degrees] C and then compacted in a cylindrical mold with the aid of a manual press.

When the chamber of the Mooney instrument is filled directly, the volume of the material used must always be about 30% greater than that of the chamber itself. It is advisable to introduce 25 [cm.sup.3]. Air inclusions are then generally non-problematical, as the pressure within the chamber rises to about 50 bar. Any air within the sample is then compressed to about 2% of its original volume.

Cross-checks and Bayer's investigations

Since 1986 we have organized or taken part in Mooney viscosity cross-checks in collaboration with a substantial number of laboratories. The results have been supplemented by investigations conducted within our own organization.

Tables 2 and 3 indicate the scope of these investigations. The results obtained were based on some 3,000 measurements.

Details of the cross-checks are as follows:

* 3rd Mooney cross-check by the IISRP-ES-TC, 1986 (International Institute of Synthetic Rubber Producers - European Section - Technical Committee) with 11 participants from Europe.

* Comparative Mooney test in NMP 435, 1987 (Standards Committee for Material Testing) with 9 participants from the Federal Republic of Germany.

* ISO cross-check - TC 45/SC 3/WG 1: Interlaboratory Test Program - Effect of mill massing on Mooney viscosity via ISO 289, 15 participants worldwide.

Table 2 - investigations of the influence of sample preparation

Cross-check 1. 1986, IISRP, Europe

7 + 1 polymers

11 laboratories

3 (or 2) preparation methods:

milling, vacuum compacting, none

3 individual measurements each 2. 1987, DIN, FRG

4 + 1 polymers

9 laboratories

3 preparation methods: milling, vacuum, none

3 measurements each 3. 1988, ISO, world

3 + 1 polymers

15 laboratories

2 preparation methods: milling, none

3 measurements each

Table 3 - investigations of the influence of sample preparation

Company investigations 1. 1987/88, Perbunan N, nitrile rubber, Leverkusen

2 polymers (2 lots N 3310)

1 laboratory

3 preparation methods: milling, vacuum, none

15 measurements each after 6 storage times

period: 1 year 2. 1988, Baypren, chloroprene rubber, Dormagen

4 polymers (B11O, B124, B230, B610)

1 laboratory

3 preparation methods: milling, vacuum, none

15 measurements each after 5 storage periods

period: 1/2 year 3. 1988, Perbunan N, nitrile rubber, flakes

12 polymers

1 laboratory

1 pretreatment method

15 measurements

Evaluation

A uniform procedure, as described in DIN/ISO 5725[4] was used for evaluation. In this connection it is important to mention that precisely specified methods (the Cochran and Dixon tests) were used to eliminate outliers:

Repeatability

r = 2.8 [s.sub.r] and

between-laboratory reproducibility

R = 2.8 [s.sub.R]

If these target functions are used, then, according to DIN/ISO 5725, Section 1/3 "the repeatability r is the level below which the absolute value of the difference between two individual findings obtained under repeat conditions can be expected with a probability of 95%."

This applies to the mean degrees of repeatability within the individual laboratories, whereas R represents analogously the variations expected between the individual laboratories.

Results

Here it should first be pointed out that, as shown in figures 1 and 2, differences in the levels of the viscosity measurements result inherently from the preparation (on the mill). These differences will be discussed separately later. In view of the general differences between the levels of the different polymers the two quantities r and R will always be stated below as percentages and relative to the respective mean value.

Repeatability r

The repeatability results are presented in figure 5 and table 4.

In figure 5 the results of the three cross-checks for these target functions are presented as bar charts. In each case the mean values for the preparation variants "milling," "vacuum compacting" and "without preparation" are presented side by side for the individual polymers.

Differences between the degrees of precision obtained for the tested polymers are immediately apparent. They range from about 2 to 8%. Except in a few cases (DIN cross-check), however, the differences between the measurements obtained for the samples prepared in the different ways are not significant. Therefore, as in table 4, mean values for r can be stated, irrespective of the polymer. Depending on the preparation method their ranges are 3.8 to 4.6% in the case of the cross-checks, 1.1 to 3% in that of the control polymer NBS butyl and 2.3 to 2.8% in that of the internal investigations. These differences were not significant, however.

The result of the investigation as a whole is that the method of preparation has no influence on the degree of repeatability. Although this is not improved when pretreatment is eliminated, there is no disadvantage either.

Table 4 - repeatablity (r)

Results

Investigation m v o m/v m/o

A) Repeatability r (%)

Cross-checks             4.6   4.0     3.8     1.21     1.22
NBS butyl                            1.1-3.0          1.5-4.2

Internal investigation 2.6 2.3 2.8 1.28 0.94 m = milling; v = vacuum comp.; o = without prep.

Between-laboratory reproducibility R

The results of the investigations of between-laboratory reproducibility are again presented as bar charts and numerical values (in figure 6 and table 5).

Here, also, much depends on the polymer. On the whole, however, the laboratory differences are considerably greater than the variations seen when the individual laboratories repeated their measurements. They ranged from a minimum of 4.8% to a maximum of 23.6%. In six cases out of 11 preparation on the mill gave a considerably less favorable result than vacuum compacting and omission of preparation. In the remaining five cases there were practically no differences.

Hence, as depicted in the right half of table 5, it is appropriate to calculate so-called improvement factors in terms of "vacuum compacting versus milling" (w/v) and "omission of pretreatment versus milling" (m/o). Admittedly these factors are not uniform; altogether, however, they show that when milling is omitted, the differences between laboratories can be reduced by about 41% (m/v = 1.70 and m/o = 1.66), which means that the differences between laboratories are about 70% greater with milling than without.

In the case of the control polymer (without pretreatment) good between-laboratory reproducibility levels of 3.2 to 4.1% are obtained.

In two cross-checks remeasurement by Bayer of samples prepared by the participants was carried out on one Mooney instrument only; this gave "theoretical" improvement factors of 3.1 and 4.0, which, however, was only achieved because there were no differences between instruments. These figures show that additional and more precise standardization of the Mooney instruments could reduce the problem of between-laboratory reproducibility still further.

Table 5 - results - between-laboratory reproducibility (R)

Results

Investigation m v o m/v m/o

B) Between-laboratory reproducibility (%) Cross-checks

IISRP SBR              5.8    4.9    4.8    1.18     1.21
      NBR              8.7    7.7    5.8    1.13     1.50
      CR               5.5    4.5    6.9    1.22     0.80
      CR              15.5   14.3   16.3    1.08     0.95
DIN   SBR             23.6    5.9    5.2    4.00     4.54
      CR              20.3    7.6   10.3    2.67     1.97
      NBR             12.9    8.6    6.5    1.50     1.98
      BR               6.5    7.7    7.1    0.84     0.92
ISO   SBR              6.8           7.1             0.96
      SBR             11.3           6.1             1.86
      CR               1.7           5.0             1.56
                      11.3    7.7    7.4    1.70     1.66

NBS butyl (Bayer repeat measurement)

IISRP                         3.2   (3.1)
DIN                           4.0   (4.0)
ISO                           4.1

m = milling; v = vacuum comp.; o = without prep.

Influence of storage time on the viscosity of differently prepared polymers

Bayer's investigations included a study of the effects of storage time.

Two examples (referring to two different chloroprene rubbers) show that the directions of the changes during the six months of the test are the same for the different preparation methods (figures showing these results are available upon request).

Summary, situation and prospects

As shown by the results presented here, Mooney viscosity measurement without sample preparation on the mill is definitely advantageous. Vacuum compacting and direct testing show no drawbacks with respect to repeatability, and lead to a marked reduction of the laboratory differences (by 41% on average). As vacuum compacting and the omission of pretreatment give equally good results, but vacuum compacting is more labor-intensive and time consuming, preference should be given to the omission of pre-treatment procedures.

It must be pointed out, however, that testing without preparation entails the possibility of shifts in the Mooney viscosity values measured. It may therefore be necessary to alter the values stated in the specifications despite the fact that the materials themselves will have undergone no change.

Additional advantages, both for polymer producers and for their customers, are a reduction in the testing workload and equipment needed, as well as an increase in the number of tests that can be performed in a given time period.

Since the last quarter of 1989 Bayer has altered polymer specifications where necessary, and during a transitional period the Mooney viscosity values, obtained both with and without sample preparation on the mill, will be reported.

DIN Standards Committee NMP 435 has revised part 1 of the DIN Standard 53523 to the effect that in the future, although preparation on the mill remains permissible, priority should be given to testing without pre-treatment. Vacuum compacting is reserved for reference purposes and for difficult compacting problems. Similar changes are envisioned by ISO. [Tabular Data 1 Omitted]

PHOTO : Figure 1 - chloroprene rubber, mercaptan type

PHOTO : Figure 2 - influence of preparation on Mooney viscosity (normalized)

PHOTO : Figure 3 - characteristic nip width curves of two mixing mills of equal size

PHOTO : Figure 4 - influence of the roll nip width on Mooney viscosity

PHOTO : Figure 5 - repeatability (r) in three cross-checks

PHOTO : Figure 6 - between-laboratory reproducibility (R) in three cross-checks

References

[1]H. Kramer and R. Koopmann, "The effect of thermomechanical action during specimen preparation on polymer viscosity," (German) Kautschuk + Gummi, Kunststoffe 37 (1984) Nr. 10. [2]H. Kramer, "The problem of comparability of Mooney viscosity measurements between different test laboratories," (German) Kautschuk + Gummi, Kunststoffe 33 (1980) Nr. 20. R.D. Stiehler, Standard and Standardization ASTM STP 553, ASTM 1974, p. 87. [3]R. Koopmann and H. Kramer, "Improvement of standard rheological test for better material characterization," paper presented at the ASTM-Symposium on processibility and rheology of rubber, 1982 (Toronto, Ont.) R. Koopmann, "Improvement for Mooney-viscosity test," Kautschuk + Gummi, Kunststoffe, 33 (1985) Nr. 2. [4]DIN/ISO 5725, "Precision of repeatability and reproducibility of standardized test producers through |Round Robin' testing," (German).