Mooney stress testing for SBR processability.

The processing characteristics of an elastomer are very important to the producer of finished rubber products. Processing includes all the steps from the first mix through vulcanization. Obviously, there are many different and varied steps involved. The first step mix which incorporates the filler often sets the tone for the whole process. Several rheological tests have been used to predict the first step mixing characteristics of a raw elastomer. However, for the most part, these tests require a complex instrument, a great deal of care in sample preparation, careful operation and/or sophisticated interpretation of the data. Thus they are not useful for routine quality control testing.

The addition of the stress relaxation feature to the Flexsys Mooney Machine 2000e, however, offers an opportunity to get rheological data for a raw elastomer relatively fast. In an earlier paper (ref. 1), the correlations between Mooney stress relaxation and several other processability tests were developed in the R&D lab. Good correlations were obtained between Mooney stress relaxation time to 80% decay or Mooney stress relaxation slope and Brabender black incorporation time (BIT), a lab test which simulates and correlates with factory mixing charateristics in many cases. In that study, delta Mooney and Mooney did not discriminate very well between normal SBR 1502 and some off-spec 1502 with wide molecular weight distributions and long BIT. On the other hand, Mooney stress relaxation t-80 showed a correlation with both the molecular weight distribution data and the black incorporation times. The SBR 1712 types in that study showed a good correlation (r2 = 0.92) between Mooney stress relaxation t-80 and BIT as shown in figure 1. Mooney stress relaxation data have now been generated in the production quality control laboratory to begin to establish a typical background for several products and to determine test reproducibility.

Experimental

Two types of instruments were used: the Monsanto (Flexsys) MV 2000E and a Brabender PL 2000 mixer. This Mooney machine can be programmed to run a stress relaxation test after the standard viscosity test. After the normal viscosity test, the machine records the stress as the sample relaxes. The time for the stress to relax or decay to a specified percentage of the original stress (final Mooney) is determined as the decay time. Also, the coefficients for the log time versus log stress regression equation are reported as the slope and intercept. While the temperature and times can be varied, these were set to conform with the standard ASTM Mooney test for SBR: 10 mill pass sample preparation, 100[degrees]C, one minute pre-heat, and four minutes running time. Two additional conditions were set for the stress relaxation portion of the test: time to 80% decay and two minutes total relaxation time. Based on our experience, a one minute relaxation time may be satisfactory.

The lab mixing studies were done with a Brabender PL 2000 instrument equipped with a 60 cc internal mixer mixing head and charging chute. The initial temperature was set at 60[degrees]C and the rotor speed at 60 rpm. Although the mixing chamber jacket was air-cooled, the mix temperatures did rise. Forty grams of polymer were added to the mixer over a 30 second time period and masticated another 30 seconds before 20 grams of IRB #6 carbon black were charged. The torque and temperature were recorded continuously. The mixer was stopped eight minutes after the black addition.

From the torque versus time graph, the black wetting time (BWT) and BIT were determined. The BWT is the time from the black addition to the first minimum torque; the black incorporation time is the time to the first maximum following the first minimum. The mix was removed from the Brabender and pressed flat. Then the Mooney viscosity was run and reported as the Brabender processability index (BPI).

All the rubber samples are production material. However, special runs and a few samples of off specification material made under upset conditions are included.

Results

One study involved the comparison of some overaged samples of SBR 1502 with fresh material. The data are summarized in table 1. Two of the overaged samples (No.'s 1 and 9) clearly gave different MSR t-80 values than the others. These were the only samples which showed so much gel they could not be run through the GPC. Sample 1 had a high Mooney result while sample 9 had a normal Mooney result. Neither sample 1 nor sample 9 showed what would be considered an abnormal delta Mooney result. This indicates that Mooney stress relaxation can discriminate between gelled and nongelled SBR 1502 better than delta Mooney.

    Table 1 - comparison of several different SBR 1502
                              samples

        ASTM Mooney tests        Mooney stress relaxation
Sample  Age   ML-4    DML    ML-4   t80     Slope     Gpc
                                    mins             Mw/Mn
  1     Old   62.6   -19.6   62.5   0.46   -0.3304   gelled
  2     Old   56.0   -20.6   56.2   0.30   -0.3566     5.9
  3     Old   57.6   -20.3   58.1   0.34   -0.3467     5.9
  4     Old   56.9   -20.6   57.5   0.37   -0.3398     5.6
  5     Old   51.0   -20.3   51.0   0.31   -0.3564     6.0
  6     Old   56.9   -21.5   58.2   0.33   -0.3468     6.7
  7     Old   49.0   -17.1   49.4   0.29   -0.3512     5.7
  8     Old   52.6   -19.2   52.9   0.33   -0.3444     6.5
  9     Old   50.3   -20.2   50.7   0.44   -0.3244   gelled
 10     Old   51.4   -20.0   51.7   0.25   -0.3705     4.8
 11     New   52.4   -19.4   52.4   0.35   -0.3455     8.5
 12     New   54.8   -19.8   54.9   0.32   -0.3465     4.9
 13     New   55.8   -20.6   55.4   0.32   -0.3487     5.5
 14     New   53.1   -19.5   52.7   0.32   -0.3525     7.3
 15     New   50.6   -18.9   49.7   0.27   -0.3623     6.3

Several studies were made in which the production lab obtained Mooney stress relaxation data on a routine basis on specific polymers. The regular operators assigned to that shift ran the samples on the Mooney machines normally used for finished product testing. The purpose was to determine the typical variation in Mooney stress relaxation properties for several high volume products. Based on these results, samples from rubber produced at about the same time were selected and tested in the R&D laboratory for Mooney stress relaxation properties and BIT.

The first such study involved 240 samples of various SBR 1712 oil masterbatches produced over about a four month period of time. The Mooney viscosity ranged from 31 ML to 58 ML. Of these, 30 samples were evaluated in the R&D lab. Table 2 shows the various correlations obtained from this data. As in the earlier work, there is a fairly good correlation between the MSR time to 80% decay (t-80) and the BIT. Figure 2 shows the correlation between the plant run MSR t80 and BIT. The difference between the plant correlation and the R&D correlation may have several causes. Multiple machines and operators in the plant lab could contribute to these differences. Also, while all of the R&D data were generated on exactly the same samples, the BIT data were not run on exactly the same sample used in the production lab. The delta Mooney test is used routinely to evaluate SBR 1712 for processability characteristics. The combined plant data for Mooney (ML-4, 100[degrees]C) and delta Mooney correlate moderately with both the plant MSR t80 and BIT.

      Table 2 - SBR 1712 production testing

Correlations with black incorporation time
                        [r.sup.2]         n
Plant MSR t-80            0.84           30
R&D MSR t-80              0.93           30
R&D MSR slope             0.92           30
R&D BPI (ML-4)            0.76           30
Plant ML + DML            0.84           30
R&D ML + DML              0.93           30

Correlations with Mooney plus Delta Mooney
                        [r.sup.2]         n
Plant MSR t-80            0.85            30
Plant MSR t-80            0.83           230
R&D MSR t-80              0.90            30

Using these data and correlations, a set of estimated specifications for MSR t80 were generated for the three major SBR 1712 products: low, intermediate and high Mooney. The range for each product is 0.21 minutes for the MSR t80 and 0.05 for the MSR slope. This information will be used later in discussing the test reproducibility.

A study of SBR 1712 produced on other lines was also run. Ten samples were selected for R&D testing out of 112 samples tested in the production lab. The correlations are similar to the first study, as shown in table 3.

    Table 3 - second SBR 1712 production testing

Correlations with black incorporation time
                        [r.sup.2]         n
Plant MSR t-80            0.83           10
R&D MSR t-80              0.86           10
Plant MSR slope           0.71           10
R&D MSR slope             0.90           10
Plant ML + DML            0.83           10
R&D ML + DML              0.78           10
R&D Brab. proc. index     0.56           10

Correlations with Mooney plus Delta Mooney
                        [r.sup.2]         n
Plant MSR t-80            0.77            10
R&D MSR t-80              0.92            10

Similar work, though less extensive studies, were made when SBR 1778 and 1713 were being produced. These lightcolored naphthenic oil extended masterbatches are produced less frequently than SBR 1712. The SBR 1778 data are summarized in table 4. The SBR 1713 runs, however, were fairly consistent showing little variation in Mooney, Delta Mooney, MSR decay time and BIT.

Table 4 - SBR 1778 production lab testing for
                    MSR properties

Correlations with black incorporation time
                        [r.sup.2]         n
Plant MSR t-80            0.63           10
R&D MSR t-80              0.91           10
Plant MSR slope           0.56           10
R&D MSR slope             0.94           10
Plant ML + DML            0.72           10
R&D ML + DML              0.79           10
R&D Brab proc. index      0.92           10

Correlations with Mooney plus Delta Mooney together
                        [r.sup.2]         n
Plant MSR t-80            0.71            10
Plant MSR t-80            0.71            76
R&D MSR t-80              0.90            10

Reproducibility studies

With the realization that the Mooney Stress Relaxation testing correlates with processability, its utility as a control test must be considered. The question about test reproducibility logically follows. Two studies were conducted in our production quality control lab: one involving SBR 1502 and the other SBR 1712 oil masterbatch. The 1502 study was by far the more extensive.

Over a two month period 136 samples from a "1502 Standard" production lot were tested using six different Mooney machines. The Mooney machine choice was left to the operator. The bulk of the data came from two machines. No attempt was made to determine differences between operators. Of the 136 samples tested, the MSR slope data were recorded on only 85 samples. Standard deviations were calculated for Mooney (ML-4), MSR time to 80% decay (t-80) and MSR slope for each Mooney machine and as a whole. To be able to compare the reproducibility of the MSR data with Mooney, a way of "normalizing" the units is necessary. The procedure chosen here is to determine how much of the specification range is covered by the standard deviation. This is expressed as a percent. While formal specifications have not been set on the MSR t-80 or slope, for the sake of this exercise, the 1712 production data were studied and likely "specification" ranges were determined for each product in three Mooney ranges. The specification range for ML-4 is 9, the range for MSR t-80 and the MSR slope probably would be 0.21 and 0.05, respectively. Table 5 summarizes the data. On the basis of the specification range covered by the standard deviation, MSR t-80 is close to the Mooney data. MSR Slope is much poorer than either Mooney or MSR t80.

    Table 5 - SBR reproducibility study

                     Mooney       MSR        MSR     No. of
                     ML 1+4       t80     -slope      tests
Mooney machine #6                                     n = 4
Average               53.98     0.288     0.3742
Standard deviation     0.56     0.015     0.0077
% Standard dev./
 average               1.03      5.22       2.06
% Standard dev./
 spec. range*          6.18      7.14      15.40

Mooney machine #7                                     n = 22
Average               55.19     0.321     0.3598
Standard deviation     0.77     0.018     0.0074
% Standard dev./
 average               1.40      5.59       2.06
% Standard dev./
 spec. range*          8.56      8.54      14.83

Mooney machine #9                                     n = 5
Average               54.24     0.304     0.3631
Standard deviation     1.34     0.030     0.0126
% Standard dev./
 average               2.47     10.03       3.48
% Standard dev./
 spec. range*         14.90     14.52      25.25

Mooney machine #10                                    n = 4
Average               54.93     0.335     0.3481
Standard deviation     0.58     0.010     0.0035
% Standard dev./
 average               1.06      2.99       0.99
% Standard dev./
 spec. range*          6.44      4.76       6.92

Mooney machine #11                                    n = 51
Average               54.52     0.293     0.3679
Standard deviation     0.63     0.019     0.0090
% Standard dev./
 average               1.16      6.66       2.44
% Standard dev./
 spec. range*          7.00      9.27      17.94

Mooney machine, all                                   n = 85
Average               54.67     0.302     0.3649
Standard deviation     0.78     0.024     0.0100
% Standard dev./
 average               1.42      7.86       2.75
% Standard dev./
 spec. range*          8.65     11.31      20.04

Mooney machine, all                                   n = 136
Average               54.71     0.303
Standard deviation     0.76     0.022
% Standard dev./
 average               1.39      7.39
% Standard dev./
 spec. range*          8.44     10.67

* Estimated spec. range   9      0.21       0.05

For the 1712 study, 25 samples from a single bale were tested by three operators over a ten day period using six Mooney machines. These data are summarized in table 6. Again, the MSR t-80 shows a little more variation than the standard Mooney test based on the percent of the specification range covered by the standard deviation. The MSR slope is much poorer than Mooney.

Table 6 - SBR 1712 reproducibility study

All data                 ML 1 + 4      MSR        MSR
                                       t80     -slope     n = 25
Average                     46.31     0386     0.3583
Standard deviation           0.72    0.024     0.0104
Estimated spec range            9     0.21       0.05
% standard dev./
 spec. range                  8.0     11.3       20.8

Oper. 1                                                   n = 7
Average                     46.10    0.379     0.3613
Standard deviation           0.63    0.025     0.0144

Oper. 2                                                   n = 7
Average                     46.69    0.391     0.3572
Standard deviation           0.79    0.025     0.0068

Oper. 3                                                   n = 11
Average                     46.20    0.386     0.3571
Standard deviation           0.69    0.023     0.0099

Machine 11                                                n = 5
Average                     46.42    0.374     0.3644
Standard deviation           0.31    0.029     0.0107

Machine 10                                                n = 4
Average                     46.45    0.388     0.3550
Standard deviation           0.95    0.032     0.0103

Machine 9                                                 n = 5
Average                     46.76    0.402     0.3502
Standard deviation           0.87    0.019     0.0054

Machine 8                                                 n = 3
Average                     46.73    0.387     0.3527
Standard deviation           0.49    0.006     0.0031

Machine 7                                                 n = 3
Average                     46.07    0.380     0.3584
Standard deviation           0.29    0.030     0.0144

Machine 6                                                 n = 5
Average                     45.52    0.382     0.3661
Standard deviation           0.41    0.019     0.0088

Thus, MSR t-80 (time to 80% decay) is about as reproducible as Mooney and can be considered as a quality control test. Note that the Mooney machine can display the MSR t-80 result as either seconds or hundredths of a minute. While the reading in seconds is a nice convenient whole number, it is also less precise than the reading in hundredths of a minute.

Reference [1.] Male, F.J., Rubber World 210-5, 73-76 (Aug. 1994).