A



PROJECT FINAL REPORT

GROUP NUMBER____T2_____

PROJECT TITLE: Designing a Procedure Protocol Using Optimal Programming and File Transfer Features of the Brookfield Viscometer

DATE SUBMITTED_12/18/00

ROLE ASSIGNMENTS

ROLE GROUP MEMBER

FACILITATOR………………………..__Greg Miller___________

TIME & TASK KEEPER………………__Wynter Duncanson___

SCRIBE………………………………..__Danielle Antalffy

PRESENTER………………………….__Jehann Biggs_____________

SUMMARY OF PROJECT CONCLUSIONS

In order to best utilize the programmable features of the Brookfield Viscometer, the following conclusions were made. The minimum time allowance between continuous measurements was found to be 15 seconds. The viscometer was determined to be most accurate and precise when run from low to high speed.

The data averaging function revealed no significant difference in the shear stress vs. shear rate measurements. Viscometer was determined to be more accurate and precise at higher the higher viscosities of the solutions tested (i.e. percent differences are low and reasonable). The viscometer was also found to be more accurate at higher torques within each viscosity. A final program that will give values within + 3.0 % of the literature values for each of the solutions was developed.

Objectives

The initial objective of the experiment was to learn how to utilize the Brookfield viscometer programming and file transfer property capabilities. Once this initial objective was achieved, we advanced further in the experiment with the actual use of several of these capabilities. The main objective of this experiment was to optimize the Brookfield viscometer for use in future student laboratory experiments. The abundant programming capabilities of the Brookfield viscometer along with the Wingather software, were used to develop an optimal and more effective program for future applications.

Specific Aims

•Find the minimum time allowance between continuous measurements.

•Determine whether the procedure should run from low to high speed, high to low speed, or include both.

•Determine Critical % Torque for each mass percent sucrose solution (20%, 30%, 50%, and 60% mass percent solutions).

•Determine usefulness of the data averaging function.

•Develop a program that will give values within + 3.0 % of the literature values (as determined by the class values) for the viscosity of the 20%, 30%, 50%, and 60% solutions.

Background Information

The CRC1 was referenced in order to obtain the literature values of the viscosities of the 20%, 30%, 50%, and 60% mass sucrose solutions. The literature viscosity values of the respective mass percent sucrose solutions were significant in determining the accuracy of the results. More Solutions to Sticky Problems2 provided significant background information on the viscometer and was useful for explaining the specific functions and parameters of the viscometer. Specifically this source revealed that the display on the Digital viscometer may fluctuate by 0.1 or 0.2% even after equilibrium is reached. It also revealed that the Brookfield measures only the torque required to rotate an immersed element (spindle) in a fluid and then converts the measurement to shear stress. Lastly, it noted that overshoot may occur since momentum is gained by the spindle during acceleration. This information proved to be significant in the achievement of the initial experimental objective of learning how to utilize the programming and file transfer capabilities of the Brookfield viscometer.

The Brookfield Viscometer Manual3 also provided specific instructions on the use of the Brookfield viscometer, which proved to be particularly important in the use of the data averaging function. Finally, the Wingather Software Manual4 provided detailed instructions on the use of the Wingather software in order to create, download, and run programs on the Brookfield viscometer. The Wingather Software Manual4 gave sample programs and sample raw data for each respective program, which aided in the selection of specific programming properties in the development of an optimized final program.

Materials, Apparatus, Methods

Materials

▪ 20%, 30%, 50% 60% wt. Sucrose solutions in DI water with 0.01% NaN3

Apparatus

▪ Brookfield DV-II Programmable Viscometer

▪ SC4-18 small sample adapter with spindle

▪ Refrigerated water bath (OR HEATED???)

Software

DVLoader Software

▪ Utilizes a scripting language that allows for the creation of programs to control the Brookfield viscometer.

▪ Loads program into the Brookfield viscometer.

Wingather Software

▪ Collects data from viscometer and displays it on PC computer.

▪ Able to set Brookfield viscometer at speeds ranging from 2-200 RPM.(SSN)

▪ Able to make Brookfield wait an allotted time interval once speed is set.(WTI)

▪ Able to print data points at specified time intervals. (SPI)

Procedure

▪ Place 8mL of each solution in the sample holder per trial.

▪ Download preliminary program (speed range: 10- 200 RPM) to Brookfield Viscometer using DVLOADER.

▪ If speed ranges were too high more appropriate program written with new speed range.

▪ Four solutions tested from low to high and high to low speed run directions (4 trials each direction).

▪ In addition, 30% wt. solution tested using data averaging feature on the Brookfield.

▪ Collect % Torques, Shear stress, shear rate, viscosity from Brookfield using WinGather.

▪ Final program written for each solution with 8 speeds and one data point collected 15 sec after speed change.

Results

Time Interval:

To determine the minimum time allowance between continuous measurements, which is the time needed for the viscometer readings to reach steady-state, the Viscosity was plotted against the Time, as can be seen in Figure 1 below. The peaks appear when the speed changes.

Figure 1: Determining the minimum time allowance between measurements.

[pic]

Run Direction:

To determine whether the viscometer should run through the speeds starting at the lowest and proceeding to the highest, or starting at the highest and proceeding to the lowest, the Shear Stress was plotted against the Shear Rate, as in Figure 2 below.

Figure 2: Sample Plot of Shear Stress vs Shear Rate for determining the run direction

[pic]

A Linear Regression was then performed on the data points and for Newtonian solutions, the slope of the graph is Viscosity, and the y-intercept should be zero. After a regression was performed, the lowest speed was removed from the graph, and another regression was performed. This continued until there were only three shear rates. The results of these regressions were then compared to known literature values to determine the accuracy and precision of each direction. Sample Trials for 30% Mass Sucrose Solution for each run direction can be found in Tables 1 & 2 below.

Table 1: Sample Trial for 30% Mass Sucrose Solution: Run Direction Low – High

[pic]

Table 2: Sample Trial for 30% Mass Sucrose Solution: Run Direction High - Low

[pic]

Data Averaging:

To determine the usefulness of the Data Averaging feature of the Brookfield Viscometer, no calculations were needed. A comparison of the raw data was made between trials using the Data Averaging feature and those with out. Sample data points for the 30% Mass Sucrose solution both with and without the Data Averaging feature can be found in Tables 3 & 4 below.

Table 3: Sample Trial 30% Mass Sucrose Solution without Data Averaging

[pic]

Table 4: Sample Trial 30% Mass Sucrose Solution with Data Averaging

[pic]

Method 1:

Method 1 entailed performing regression analysis on the shear stress vs. shear rate line plots. Because there were inconsistencies in the shear stress and shear rate readings at low speeds (low % torques), the first point (which coincides with the lowest speed) was eliminated and a regression was performed on the remaining points. The second point was then eliminated and a regression was performed on the remaining points until the 3 points representing the highest speeds remained. The viscosities obtained from the slopes of each regression step were compared to literature values, and the percent differences were calculated. The % torques that corresponded to the lowest speed in the speed interval used in the regression that yielded the lowest percent difference from the literature value was considered to be the critical % Torque. These % torques values were then averaged to obtain the Critical % Torques for each solution. Figure 1 and Tables 1 and 2 provide an overview of the regression analysis performed for the 60% sucrose solution.

Figure 3: Shear Stress vs. Shear rate Plot of 60% wt. sucrose solution (Trial 1)

[pic]

Figure 3 above is a linear regression plot for the 60% weight sucrose solution, trial 1. This regression included all speeds in the speed range of the 60% weight sucrose solution, which included speeds form 2 – 50RPM. As displayed, the viscosity of this solution is 58.84cP, which is calculated from the slope of the line.

Table 5: Calculated viscosities, % Torques and % differences from the literature value for 60% sucrose solution.

[pic]

Table 5 above displays the calculated viscosities from a sequence of regressions performed on the graph displayed in Figure 3. The viscosities listed were calculated from the slope of the curve and multiplied by a conversion factor of 100 to obtain units of cP. In addition, the average % torques listed corresponds to the lowest speed within each speed interval, i.e. the first torque of 3.10% corresponds to a speed of 2 RPM in the 2-50 RPM speed interval. The % differences listed in the last column are the percent differences from the literature value (58.37 cP) of viscosity of the 60% sucrose solution. The highlighted line indicates that, for this particular trial, the % torque was chosen because the measured viscosity had the lowest percent difference from the literature value. This % torque was considered to be the cutoff % torque or the critical % torque for this particular trial.

Method 1 was performed for all four trials of the 60% mass sucrose solution, and an average % Torque was calculated which represented the average % cutoff Torque for this solution. Method 1 was also performed for four trials of each mass percent sucrose solution, i.e. 20%, 30%, and 50%. Table 3 below displays the chosen critical % torques for the four different trials of the 60% sucrose solution.

Table 6: Chosen torques with the lowest % difference from literature values for all four trials of 60% sucrose solution.

[pic]

Table 6 above displays the lowest percent differences of the viscosity from literature values for four trials of 60% solution. The bold numbers at the bottom of the table represent the average and the standard deviations for the cutoff torques and the calculated viscosity values.

Method 2:

Method 2 entailed the further analysis of the linear regressions, specifically paying close attention to the 95%confidence limits. The regressions from Method 1 for Trial 4 of the 60% and 30% sucrose solution with their confidence intervals are shown in Tables 7 and 8 below. Using the regressions obtained from Method 1, the confidence limits of the slope of the shear stress versus shear rate graphs were compared to each another. The percent change of each value to the first value, and then from each value to the last value were calculated. The cutoff point was determined to be the point at which the 95 % confidence limit changed by more than 3%.

Table 7: Trial 4 60% Sucrose Solution Regression Statistics

[pic]

Table 8: Trial 4 30% Sucrose Solution Regression Statistics [pic]

Method 3:

Figures 3 and 4 below are two graphs of Viscosity vs. Shear Rate for the 50% and the 30% mass percent sucrose solutions. As in Methods 1 & 2, due to the differences in the shear stress and shear rate readings at low speeds (low % torques), the first point (which coincides with the lowest speed) was eliminated and a regression was performed on the remaining points. The second point was then eliminated and a regression was performed on the remaining points with the continuation of this process until only 2 data points remained. The slope in the viscosity vs. shear rate is expected to be zero since sucrose is Newtonian and should have the same viscosity at every speed.

Figure 3: Trial 3 50% sucrose solution Viscosity vs. Shear Rate

[pic]

Figure 4: Trial 3 30% sucrose solution Viscosity vs. Shear Rate

[pic]

The statistics for the regression analysis are shown in the tables 9 and 10 below. With the regression analysis, the point when zero fell within the confidence limits was taken as the cutoff point. At this point, the slope would then no longer be significantly different from zero. For the 50% solution, zero fell between the confidence limits at a speed range of 150-180 RPM, giving a Critical Torque of 75.9%. For the 30% sucrose, there was no cutoff torque as the zero never fell between the confidence limits. The % cutoff torque is shown in yellow in the regression table.

Table 9: Trial 3 50% sucrose Solution Regression Statistics

[pic]

Table 10: Trial 4 30% sucrose Solution Regression Statistics

[pic]

Method 4:

Method 4 was a revision of Method 1. Again we performed regression analysis on the shear stress vs. shear rate line plots. The lower speeds were removed one at a time, and a regression was performed on the remaining points. The viscosities obtained from the slopes of each regression step were compared to literature values and the percent differences were calculated. These % difference values of each trial were then averaged to obtain the % difference for each solution. Instead of choosing the regressions that yielded the smallest percent difference in viscosities from literature values, the % torques that corresponded to the lowest speed in the speed interval that first came within 3% of the literature value was considered the critical Torque. Table 11 shows the % difference as a result of the regression analysis for the four trials of 30% mass Sucrose solution. The highlighted area indicates the Critical % Torque that first came within 3% difference for this solution. Table 12 below is a summary of the Critical Torques as determined through Method 4.

Table 11: % Difference results of Method 4 Regression Analysis of 20% Mass Sucrose

[pic]

Table 12: Summary of Critical Torques determined using Method 4

[pic]

Analysis

Time Interval

By zooming in on the Viscosity vs. Time graph (Figure 1), the minimum time allowance was chosen as the time when the change in viscosity with respect to time equaled zero. The peaks in the graph represent the point in time when the speed changed which caused an increase in the momentum. At the speed change “momentum gained by the spindle during acceleration may cause the reading to initially oscillate about the final equilibrium value”2. The viscosity settled after 10 seconds, but the time allowance between continuous measurements was chosen as 15 seconds, to eliminate any possible error.

Run Direction

The low to high run direction gave more accurate, and precise values. The overall percent difference from the actual value for the low to high was smaller than the percent difference from the actual value for the high to low, which deemed them more accurate. The range of the absolute value of the percent difference for the high to low for the 30% solution was 3.74-15.69%, and the range of the absolute value of the percent difference for the low to high was 2.17-5.94% (Tables 1 and 2).

Data Averaging

The data averaging gave the same values for speed, torque, viscosity, shear stress, and shear rate for each measurement and calculation. Without data averaging, the values for the torque, viscosity, and shear stress fluctuated. The values for speed and shear rate remained constant. The shear rate is solely speed dependent, so since the speed did not change the shear rate should not have changed. Specifically for the 30% solution (Table 3), the viscosity values fluctuated between 3.17 and 3.20 cP. When data averaging was performed on another sample of 30% sucrose solution, the viscosity reading remained constant at 3.09 cP (table 4). Although the data averaging gave a constant value, the shear stress, viscosity values for the data averaging, and without data averaging did not significantly differ from one another.

Method 1:

Method 1 was inconclusive in determining the % critical Torques for all four trials of each mass percent sucrose solution. The main problem with this method was that as the lower points in the shear stress vs. shear rate graphs were dropped, the percent differences from the literature values increased. For example, when all the points were included in the regression for the 60% solution as seen in Table 5, the viscosity was 0.81% below the literature value. Because the viscosity was not significantly different form the literature value, all the points could have been used and the critical torque would be as low as 3.10%. However, because this method called for the lowest percent difference of the viscosity from literature values, the speed range between 15 and 50 RPM (highlighted line in Table 1 in the Results section above) was chosen because the viscosity was only 0.57% lower than the literature value. Since this method of analysis was performed on all of the trials of the 60% solution, when the lowest % differences were chosen for all the trials, the average % torque was found to be 12.83 + 14.00 (Table 6). The large standard deviation was due to trial 1, where the cutoff torque was taken when percent difference was 0.57% (cutoff torque was 29) because it was lower than 0.81% (cutoff torque was 3.1). (see Table 5)

The regression results for the lower concentrations such as the 50 % and 20% solutions were even more difficult to interpret. Figure 5, below represents the linear regression fit for the shear stress and shear rate for the 50% solution, Trial 1.

Figure 5: Shear Stress vs. Shear Rate regression curve for 50% sucrose solution, Trial 1.

[pic]

The viscosity, which is also the slope of the curve, when all points were included in the regression is 15.15 cP, with an R2 value of 0.9998 and a cutoff torque of 5.35% (see appendix). This viscosity was only 1.62% higher than the literature value (15.54 cP). However, when the first 6 points were dropped and a regression was performed, the resulting viscosity was 0% different from the literature value as seen in Table 13 below. This regression was performed for points between 105 and 180 RPM, but it yielded a high cutoff torque of 53.2 %. Because of the methodology of choosing the Torques with the lowest percent difference in viscosity, the cutoff torque of 53.2% was chosen instead of the 5.35% which included all the points in the regression (see appendix). Table 13 below, gives a summary of all the cutoff torques chosen for the four trials of the 50% mass sucrose solution.

Table13: Chosen torques with the lowest % difference from literature values for all four trials of 50% sucrose solution.

|Trial |Speed Interval |Cutoff |R^2 |slope |y-intercept |Viscosity (cP) |Lowest %difference|

| |(RPM) |Torque | | | | | |

|1 |105-180 |53.20 |0.999 |0.15 |-0.35 |15.40 |0.00 |

|2 |90-180 |46.00 |0.999 |0.15 |-0.21 |15.43 |-0.19 |

|3 |120-180 |60.40 |0.998 |0.15 |-0.11 |15.33 |0.45 |

|4 |50-180 |24.50 |1.000 |0.15 |-0.23 |15.15 |1.62 |

|Average |46.03 |- |0.999 |0.15 |-0.22 |15.33 |0.47 |

|SD |15.51 |- |0.001 |0.00 |0.10 |0.13 |0.82 |

As seen in Table 13 above, the average % cutoff torque found for the 50% mass sucrose solution was 46.03 + 15.51, which is unusually high. When all the points (speed interval between 10-180 RPM) were included in the regression for trial 1, the viscosity was only 1.62% different from literature value (see appendix). But the regression between 105 and 180 RPM was chosen because it yielded the lowest percent difference from the literature value. For this trial, it corresponded to a high cutoff torque of 53.2%. In addition all the regressions performed for the 50% solutions (see appendix) contained less than a 5 percent difference from the viscosity literature value. This low percent difference indicates that the viscosity values obtained from each regression were too accurate to distinguish a cutoff torque.

For the lower concentrations of sucrose solutions, a different problem arose. The linear regression line does not fit the 20% sucrose solution data seen below in figure 6 as well as the line fits the 50% or 60% sucrose solution plots. The R2 value is 0.9776 because the points are more scattered than before.

Figure 6: Shear Stress vs. Shear Rate regression curve for 20% sucrose solution, Trial 1.

[pic]

When a regression was performed on all the points for Trial 1 of the 20% mass sucrose solution (Figure 6), the viscosity was 4.69% greater than literature value. However, when the first three points were dropped, the percent difference dropped to –0.46%, as seen in the table 14 below. This percent difference is more accurate and the corresponding cutoff torques was 1.20%. This low cutoff torque was a trend seen for all the four trials in the 20% solution and the average cutoff torque was determined to be 1.88 + 0.52 %(Table 14). The problem seen here is that the cutoff torque is supposed to be higher for lower dilutions. In our case, the lower dilutions (50% and 60%) had higher cutoff torques and higher dilutions (20% and 30%) had smaller cutoff torques. This problem lies in the inconsistencies in choosing the cutoff torques according to this particular method.

Table 14: Regression statistics for 20 % sucrose solution

|Trial |Speed Interval (RPM)|Cutoff Torque |R^2 |slope |y-intercept |Viscosity (cP)|Lowest |

| | | | | | | |%difference |

|1 |17-100 |1.70 |0.916 |0.02 |0.09 |1.95 |-0.46 |

|2 |12-100 |1.20 |0.991 |0.02 |0.01 |1.99 |-2.52 |

|3 |20-100 |2.30 |0.985 |0.02 |0.28 |1.97 |-1.49 |

|4 |20-100 |2.30 |0.940 |0.02 |0.02 |1.98 |-2.13 |

|Average |- |1.88 |0.959 |0.02 |0.10 |1.98 |-1.78 |

|SD |- |0.52 |0.037 |0.00 |0.13 |0.02 |1.08 |

In all of the solutions, as more of the lower points were dropped, the uncertainty and % percent difference in the viscosities increased. This was unusual because at higher speed, and thus higher torques, the viscosity values should be more precise and accurate, not less precise and accurate as our results indicate. However, in further evaluating this method, dropping off more points provided fewer points at higher speeds to be included in the regression. Fewer points led to an increased uncertainty. Therefore, even though points at higher speeds are supposed to lead to more accurate viscosities, we concluded that there were not enough points in the regression to give a more reliable and accurate viscosity measurement. If the machine allowed us to select speeds above 200 RPM, higher torques would have been obtained and it would have been easier to determine the critical % torques, especially for the lower mass percent sucrose solutions.

Method 2:

In general, the higher speeds should yield more accurate results. The viscosity should be more accurate as the lower speeds are dropped off, however, they became less accurate. Rather than the uncertainties decreasing, they steadily increased. The results were not consistent, or decipherable, so the torque could not be determined. The slopes changed significantly, for example the 95% confidence interval increased by 29% when the lowest speed was dropped off, and 40% when the second speed is dropped. A such , the cutoff could not be determined. The uncertainty of 3% was chosen as a % difference from the actual value, so it was not really applicable in this case as there was no comparison to literature values in this method.

Method 3:

For the 50% mass sucrose solution data, the values of the critical % torques determined are in Table 14 below. The point at which zero fell within the confidence interval possessed a very large torque. With the 30% data, there was no cutoff point, as zero never fell within the confidence limits. Due to the great amount of fluctuation in the data and the inconsistency of the results, the critical percent torque could not be determined. Following is a table of the critical % torques determined through the use of this method for the 50% mass sucrose solution over four trials.

Table 14: Critical % Torques for 50% Mass Sucrose Solution using Method 3

[pic]

From the data, one can see that the values are not consistent, and declaring a critical torque and speed range would be unscientific and incorrect. The standard deviation is large, and the values are significantly different. For the 30% mass sucrose solution there would not be a critical toque within the speed ranges used in this experiment, and it could be declared that there was no critical torque for the 30% solution; however, since the 50% data was incorrect, the 30% solution data is also assumed to be incorrect.

Method 4:

Method 1 was also determined to be inconclusive for finding the % critical Torques for all four solutions. The main problem with this method was that for both the 50% and 60% Mass sucrose solutions it was determined that the inclusion of all points fell within the 3% cutoff. As such, there would be no Critical Torque. This however cannot be true. As can be seen in plots of the viscosity vs shear rate in the results section. As for all Newtonian solutions, the slope of the line should be zero, and as it can easily be seen it is not. The results of Method 4 which determine that there is no Critical Torque for the 50% and 60% mass solutions is wrong. Another problem with Method 4 arose in trying to determine the Critical Torque for the 30% Mass sucrose solution. As can be seen in Table 15 below, the inclusion of all points (range 10-180) fell within 3%, and as each speed was dropped, fell above the 3% cutoff, and then fell within again. Here, the highlighted region indicates the selected Critical Torque.

Table 15: % Difference results of Method 4 Regression Analysis of 20% Mass Sucrose

[pic]

Final Program:

A final, optimized program was developed for each mass percent sucrose solution using the data analysis performed via Method 1. This final program included a speed interval that began at the critical % torque as determined via Method 1 for each mass percent solution and ended at the maximum possible torque for that particular solution that could be achieved by the viscometer. Each final program initially set the speed at the speed at which the critical % torque was achieved via the command SSN, waited a time interval of 15 seconds as determined through data analysis of the viscosity versus time graph as discussed above via the command WTI, and printed a single data point after the 15 second time interval at that particular speed. Thus, the final program produced a total of eight data points for analysis, as opposed to the 370 data points produced by the original program. This final, optimized program also required only 120 seconds (2 minutes) as opposed to the approximate 6 minutes required by the original program. Finally, this optimized programs produced data points within the critical percent torque range for each mass sucrose solution according to the data analysis performed via Method 1. As discussed previously, however, this data analysis is inconclusive, and a true critical % torque was unable to be obtained for any mass percent solution. In Table 16 below, the final program for the 60% sucrose solution is given with a speed interval of 17-50 RPM as the critical % torque for the 60% mass sucrose solution was determined to be achieved at a speed of 15 RPM via data analysis using Method 1. The final programs for the additional mass sucrose solutions, i.e. 20%, 30%, and 50% mass sucrose solutions, are given in the Appendix.

Table 16: 60% Mass Sucrose Solution Final Program.

[pic]

This final program was tested on two samples of the 60% mass sucrose solution with the analyzed, compiled results given in Table 17 below.

Table 17: Final Program Results over 2 Trials of the 60% Mass Sucrose Solution.

|Final Program Results |  |

|Trial |Calc. Visc. |Lit. Visc. |% Diff. |

|1 |61.9 |58.37 |0.65 |

|2 |58.62 |58.37 |0.42 |

As can be seen in Table 17 above, the final program produced viscosity values within 0.65% of the literature values, which was a main objective of this experiment.

CONCLUSIONS

1. The minimum time allowance between continuous measurements was found to be 15 seconds.

•

•

2. The viscometer was found to be most accurate and precise when run from low to high speed.

3. The data averaging function revealed no significant difference in the shear stress vs. shear rate measurements.

4. Viscometer was determined to be more accurate and precise at higher viscosities within the range of solutions tested (i.e. percent differences are low and reasonable). As such, the 50% and 60% percent solutions were found to be more accurate than the 20% and 30%.

5. Viscometer was found to be more accurate at higher torques within each viscosity.

6. A program that will give values within + 3.0 % of the literature values was determined.

References:

(1)“Concentrative Properties of Aqueous Solutions,” Handbook of Chemistry and Physics. 63d Edition, 1982-1983. p D-227

(2) More Solutions to Sticky Problems A guide to getting more from your Brookfield viscometer. Brookfield Engineering Laboratories INC. Middleboro MA p6-11

(3) Brookfield DV-II+ Programmable Viscometer Operating Instructions Manual No. M/97-164-B299 Brookfield Engineering Laboratories INC. p33-39

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download