Burst Test

INTRODUCTION

The objective of Phase II was to learn more about the thermal welding process of PVC geomembranes and to develop a window of appropriate conditions for welding.  More specifically, a wider range of welder temperatures and speeds were to be evaluated along with the effects of sheet temperature.  There was also interest to further explore the relationship between bursting the seams from the air channel out and peeling the seams from the outside in.  If this relationship is known, one can eliminate the practice of cutting holes in seams to determine seam strength.

SEAM PREPARATION

The 72 prepared seams were made in a single day in Austin, Texas on an asphalt subgrade.  There were two crews, one using a hot air welder and the other a hot wedge welding machine.  The hot air machine was the same one used in Phase I, namely a Leister Twinnie Model CH6056.  The hot wedge machine was a “Mini-Wedge” made by Plastic Welding Technologies (formerly Columbine).

The crews each used three welder set points and three welder speeds based on their “normal” conditions, and their experience.  Each crew made a set of 30 mil and 40 mil seams in the shade in the morning. Then, they each made an identical set of seams in the sun in the afternoon.  The sheet temperatures range from 50 to 100°F.  The temperature was monitored by a thermocouple attached to the sheet.  The effect of nip roller pressure on seaming was not examined during this study.  Both welders had a typical pressure pre-set and this was maintained throughout the seaming operation.  The following table shows the different conditions used.


 

                                                   Table 1 - Seaming Parameters Used

SEAM EVALUATION

The seams were evaluated by the standard peel test at 20 in/min at 73°F and by a burst test developed for this project.  The burst test was performed by sealing off one end of a seam length and pressurizing the other end with compressed air.  The basic procedure was to select a starting pressure, hold there for 30 seconds, then ramp 5 psi at a time, holding for 30 seconds for each 5 psi step.  The 5 psi was applied in a 5 second time period.  This went on until failure occurred.  Most of the failures were peels that occurred during the 30 second soak.  However, there were some seams that burst during a 5 psi step.

The burst test done at room temperature was done on a 6 feet length of seam.  More seam length was useful to determine the relationship between peel and burst.  The burst test was also performed at two higher temperatures.  These tests were performed in a constant temperature room set for 100°F and 120 °F.  The actual sheet temperatures were 95°F and 116°F.  These elevated temperature tests were performed on 4 feet lengths of seam. The seam was clamped in the center, then both 2 feet halves tested to produce duplicate results.  None of the individual test strips that were pressurized were also peel tested. All tests were performed on as-made strips from the original 30 ft length.

RESULTS

The results from testing all of the 72 seams for peel strength and for burst strength at three temperatures are given in Tables 2-9.  The two values listed for peel strength are the two weld tracks.

The burst values at room temperature were done once while the higher temperature bursts were done in duplicate. The values presented are the averages of the two burst tests.  Typically, the two results were within 5 psi of one another.  Also, in the hot air seams (“A” series) the numbering is not consistent from 1 to 36, as it is in the hot wedge seams (“W” series).

There were two instances where a particular seam burst at a very low value due to specific weak spot in the seam.  They occurred at the temperatures of 800° F and 900° F, and looked like “burn through”.  This is the type of thing that will occur in the field since 100% of the seams will be tested.  In the field, these weak spots would be locations for a patch.

Otherwise, the burst behavior was as one might predict.  As more heat got into the seam, the peel  and burst values increased.  Of course, more heat gets into a seam from higher set point temperatures, slower speeds, or an increased sheet temperature.  Experienced welders will often adjust their welding speed to the ambient conditions.  For example, they might increase their speed during the day as the temperature gets warmer.  Or, if clouds suddenly appear, they may slow down because the sheet is cooler in the shade.

The strongest seams exceeded the pressure gauge used.  It had a scale up to 120 psi, so above this pressure, the operator made an estimate of the pressure.  The pointer still went higher, but there was no scale to read above 120 psi.



 

                   Table 2 - Peel and Burst Results of 30 mil Seams
                               Made with a Hot Air Welder at 50°F

                  Table 3 - Peel and Burst Results of 30 mil Seams
                              Made with a Hot Air Welder at 80°F

                   Table 4 - Peel and Burst Results of 40 mil Seams
                               Made with a Hot Air Welder at 60°F

                   Table 5 - Peel and Burst Results of 40 mil Seams 
                               Made with a Hot Air Welder at 90°F

                  Table 6 - Peel and Burst Results of 30 mil Seams 
                          Made with a Hot Wedge Welder at 50°F

                   Table 7 - Peel and Burst Results of 30 mil Seams
                         Made with a Hot Wedge Welder at 90°F

                  Table 8 - Peel and Burst Results of 40 mil Seams
                          Made with a Hot Wedge Welder at 60°F

                  Table 9 - Peel and Burst Results of 40 mil Seams 
                          Made with a Hot Wedge Welder at 100°F

Effects of Welder Speed, Welder Temperature, and Sheet Temperature on Peel Strength

Since there are so many sets of results, a series of bar graphs was prepared to examine trends in the results.  Figure 1 shows the peel strength results for all the seams made by the Hot Air Welder.  The temperature in the header was the sheet temperature when the seams were made.

There are a number of observations one can see in these plots:

1. The effect of speed seems to be greater than the effect of temperature.  This suggests that one should be able to increase the strength under a given set of conditions by slowing down the welder.

2. A speed of 10 ft/min produced seams significantly weaker than those made at 7 ft/min under the same conditions, within the temperature range studied.  In fact, of the 4 seams that did not meet 15 ppi in strength, all were made at a speed of 10 ft/min.  Two other seams made at this speed had peel strengths of exactly 15 ppi.

3. The effect of sheet temperature was greatest at the lowest welder temperature for 30 mil sheet.  This suggests that a temperature of 608 °F is too low for good welds.

4. All these seams made at 734 °F have similar strengths as a function of speed and welder temperature.  This suggests that a temperature near this one (like 750 °F) would be a good starting point for setting up a welding operation.

5. The upper temperature of 824° F produced excellent seams at 4 and 7 ft/min and good seams at 10 ft/min.  This suggests that the temperature can be raised further to provide more heat for a better seam at 10 ft/min.  It is likely that a welder temperature of 850 or 875 °F would produce higher strengths at 10 ft/min.  Of course, burn through and acidic corrosion increase at high temperatures also.

6. Seams that do not peel are obtained at peel strengths around 40 ppi.

The observations one can make about these results include:

1.  The 30 mil seams had a maximum peel strength around 25 ppi. Also, the difference in strength with different conditions was small.  This seems to suggest that there is a ceiling on peel strength in this case.  The results for seams welded at 80°F show a difference of only 6 ppi between the highest and lowest strength.  This is odd, in light of the burst test results, which will be discussed in the next section.

2.  As before, the effect of speed is greater than the effect of temperature.  A speed of 19 ft/min is obviously too fast for 40 mil but made reasonable 30 mil seams.

3.  The effect of ambient temperature was largest in the 40 mil seams, except for the slowest speed.  The temperature difference was also the greatest (40°F).  These results also suggest that good seams can be made at a set point temperature as high as 900° F, but again, the increased possibility of “burn-through” and corrosion increase as the temperature is increased.

4.  The only acceptable 40 mil seams made in cooler temperatures were at the slowest speed of 3 ft/min.  This might involve the sheet’s contact with the wedge surface.  A cooler and stiffer sheet may not contact as well under the roller pressure used.

Effects of Welder Speed, Welder Temperature, and Sheet Temperature on Burst Strength

The burst test is essentially a peel test from the air channel out.  It should be sensitive to the same conditions as the peel test but may be more indicative of seam quality since it will find the weakest area of the seam.
 

These results are similar to the peel strength results. Speed has a more dramatic affect than welder temperature, and there is just a small increase in strength when the sheet temperature is changed from 50° F to 80°F. Also, the burst strength of the 40 mil seams is significantly higher than the burst for 30 mil seams. 

These results show some significant differences in burst strength for 30 mil seams that did not appear in the peel test results.  Otherwise, the same trends were seen. Speed has the greatest effect and weaker seams are made at cooler temperatures at speeds of 10 ft/min or more.

One of the most interesting findings from the last two sections involve the sets of 30 mil seams. Notice that the seams made by the wedge had maximum peel strengths around 25 ppi while the hot air seams had maximum peel strengths of 45 ppi.  Conversely, the hot air seams showed burst strengths less than 80 psi while the wedge seams had maximum burst strengths over 100 psi.

This seems to indicate that bursting from the inside out is not the same as peeling from the outside in. If there is a difference, it must be in the initiation of peeling. It looks like the inside of the wedge seams is stronger than the outside. And, the opposite seems true for the hot air seams. Their outsides seem stronger than the insides. More information should be obtained by exploring the relationship between the peel strength and the burst strength.

The Relationship Between Peel Strength and Burst Strength

It seems reasonable that there would be a strong relationship between these two properties because they both involve peeling the seam apart, assuming the seam peels.  This section will look at this relationship for all of the seams prepared.  The relationship between peel strength and burst strength for 30 mil seams made with hot air is seen in Figure 5.  This plot uses the lowest peel value of the two tracks and the average burst value.

The value of the slope noted on the graph is the slope after the four points furthest from the original line were removed. The purpose of removing these points is to get a better value for the slope of the line.  Notice that there was very good linear correlation as displayed by the r2 value.  A correlation of 1.0 would indicate a perfect fit.

This time, three points were omitted to better define the line. The selection of these point is largely arbitrary. However, this part of the analysis will give an estimate of the Burst Strength vs. Peel Strength relationship. It will still need to be proven that the relationship is correct.

The relationship for the hot wedge welder are somewhat different then the ones just seen. 
 

This plot clearly shows that these seams reach a maximum peel strength even though the burst strength indicates stronger seams. Therefore, only five points were used to define the relationship for this set of seams. Similar behavior is seen in Figure 8, which shows the plot for the 40 mil, hot wedge seams. 

Once again, it is observed that a maximum peel strength has been reached. Both of the plots for the wedge welder indicate that the seams are stronger from the air-channel out than from the outside towards the air channel. It is not known at this time if this is specific to the particular welder used or if its a general phenomenon with wedge welders. It is possible that the wedge itself might have a slightly different profile at the air channel than at the outside edges.

If one takes an average of the four slopes, the overall slope is 0.448 ppi/psi.  Alternatively, one can combine all the data for a general line.  Figure 9 shows the combined results before and after the “outliers” are removed.

Relationship Between Burst Strength and Sheet Temperature

It is well known that plastics soften with increasing temperature. Not only is the tensile strength lowered, but increased temperature also results in lower peel strengths and lower burst strengths. Since the burst test might be used as a field test one day, it is necessary to know the relationship between burst strength and temperature. Ultimately, one would have to know what burst strength is required at any given temperature to ensure a particular peel strength. This is why a good relationship between peel and burst must be developed.

The same exercise that generated the plots in Figure 9 was done for the seams burst at 95°F and those burst at 116°F. The results are shown in Figures 10 and 11. This provides a good estimate for the relationship between room temperature peel strength and the burst strength at different temperatures.

Before and After Removal of “Outliers”

Since the room temperature peel strength is common in Figures 9-11, a single plot can be prepared showing the relationships between burst strength and peel strength at all three temperatures. This is seen in Figure 12.

The slopes of the lines are rates of change which makes it appropriate to use the Arrhenius model to determine the temperature dependence of this process.  The first step in this process is to prepare a plot of ln Rate vs. 1/T, where T is the absolute temperature.  This is shown in Figure 13.

This line can now be used to prepare a plot of the required burst strength to equal a given peel strength at any temperature. For example, Figure 14 shows the minimum burst strength required to ensure a 15 ppi peel strength.

So, if this model is correct, one could simply apply the required burst pressure and if the seam holds, then the peel strength was over 15 ppi. This could be done in place of destructive seam testing.  It has the advantages of no cut holes, no patches, and 100% testing.  It also can be done onsite, regardless of the temperature.  Similar curves can easily be prepared for other peel strength values.
 

COMMENT: 

EPI is continuing research on burst strength at lower sheet temperatures. As more points of data are developed, the relationship between burst strength and sheet temperature will likely be a straight line graph.

Evaluation of the Model

Now that the model has been developed, it can be checked by seeing if all the 72 seams burst at three temperatures fit the model.  Using the pass/fail criteria defined by the model, all the seams will be evaluated for pass/fail.  The results are shown in tabular form in Tables 10 - 13.

                   Table 10 - Pass/Fail Results for 30 mil Hot Air Seams

Each of the three seams less than 15 ppi peel also failed to meet the appropriate burst requirement.  Two seams (A9 and A18) had peel strengths greater than 15 ppi but failed the burst requirement at one temperature.  Notice that both of these seams were near the 15 ppi requirement.  Seams that fail to meet the burst requirement but meet the peel requirement are considered false negatives.

                   Table 11 - Pass/Fail Results for 40 mil Hot Air Seams

This time, the one seam that was less than 15 ppi in peel (A30) passed one of the three burst requirements.  This would be considered a false positive.  There was  also one false negative (A39) which failed the burst requirement at 116°F.  So, for this set, one seam had a false positive and 1 seam had two false negative values.

              Table 12 - Pass/Fail Results for 30 mil Hot Wedge Seams

This set showed one false negative (W2).  Notice that the burst value at 73°F (25 psi) was actually lower than the peel value at 95°F (31 psi).  This indicates that the portion of the seam tested at 73°F had a weak spot that was captured by the burst test.  Interestingly, this was the set that had the fewest results used to construct the model.

              Table 13 - Pass/Fail Results for 40 mil Hot Wedge Seams

One seam showed two false negative values (W29).  There is little to be said about W29.  It easily passed the peel test, barely passed the burst test at 73° and failed the burst test at 95°F and 116°F.

Table 14 shows a summary of all the false positives and false negatives.

                        Table 14 - Summary of Inconsistent Results

This summary shows that 5 seams had false negatives while one had a false positive.  One can also consider how many total seams would have failed for each of the different pass/fail criteria.  Table 15 shows this information.

      Table 15 - Numbers of Failures for Each Different Requirement