Exposing the Myths of Tensile Seal Strength Testing
The controversies regarding tensile seal strength tests need to be resolved.
Donald S. Barcan, president & CEO, DBI Inc., Long Valley, NJ
There is a major debate regarding the best way to perform tensile testing, inviting reaction from both packaging material suppliers and end-users. Everyone embraces the use of ASTM F-88 as the test method of choice. However, the test method requires the user to choose between supported and unsupported test samples and between 90ï¿½ and 180ï¿½ approaches, as well as between peak, sustaining, or average values. And these are just a few of the possible variations.
Problems arise from this lack of standardization, both on the supplier end and/or on the user end. Some companies conduct the method one way, while others use another. Some companies even lack standardization and do it multiple ways within their divisions. As a result, industry is unable to compare numerical test values because there is no universal conversion factor between the test approaches. Also, in some cases, the specification values do not reflect the differences in test methods. The time has come for industry to choose a single method, one that can be verified by sound statistical evidence and that makes sense.
|Figure 1. Comparison of seal strength testing methods (click to enlarge).|
I would like to expose two myths that stand in the way of standardization. These are the proverbial one-pound-per-inch (peak) standard for determining tensile seal strength and the use of backing plates (supported, 180ï¿½) instead of the unsupported, free-tail method that uses no backing plate.
The one-pound-per-inch peak ï¿½requirementï¿½ came mostly from material suppliers of unconverted and converted products. This was the best value they could guarantee to medical device customers. Back in the 1960s, when surgical-grade paper was king, seal-strength values were governed more by the internal bond strength of the paper than by adhesive technologies. At that time, device manufacturers were equally concerned about clean peels, the opening of the package without generating loose fiber and encapsulating the device. The solution, though expensive at the time, was to provide a heat-activated coating on the paper.
If we move up to the present day, most, if not all, of the material suppliers of either roll stock or fabricated pouches indicate in their specifications that they will guarantee no greater than a 1-lb peak tensile seal strength for peelable seals. Depending upon the supplier, this value can be achieved with or without the backing plate. What I find remarkable is that this 1-lb specification does not vary, regardless of the materials involved. In their defense, suppliers will change the 1-lb peak specification if requested. However, have we come a long way from the 1960s? I think not.
I know from firsthand discussions, during those early days, that the pouch and material suppliers were looking for a test method that would maximize the numerical value of seal strength. In essence, the supported method was used because it produced numerical values that were between 1.5 and 3.0 times greater than those obtained without the use of the backing plate. This was adopted by most suppliers and has become a de facto standard.
The actual seal strength differences between methods depend on the specific material combination. I have specific experience with a 6-mil nylon film sealed to the same film with an adhesive layer coextruded on it. This combination produces seal strengths that are between 2.5 and 4.0 times higher with the supported backing plate than without. Unfortunately, the standard deviations are also significantly higher, between 2.0 and 3.5 times higher. This shows that the process variability indicator, standard deviation, was much greater for the supported tests than unsupported tests, yet the seals were identical. I suspect that this variation is dependent on material thickness. The thicker the material, the greater the seal strength and process variation will be.
I recently completed a comparison of different material combinations sealed under identical conditions. The purpose was to determine the effect on the numerical value of peak tensile seal strength using the various test methods. Figure 1 on page 26 shows that test speed changes produced insignificant changes in peak seal strength. However, the differences between unsupported, supported 90ï¿½, and supported 180ï¿½ were significant. As you can see, the supported 90ï¿½ approach produced the lowest numerical value and supported 180ï¿½ the highest.
Designing test methods usually takes into consideration how the item being tested is used. I donï¿½t know of any medical device practitioner who would use a backing plate to open a pouch package. The use of the backing plate artificially raises the value of seal strength, giving us a false sense of security.
The medical device industry, both material producers and users, should settle on one methodï¿½the unsupported method using no backing plate, also called the free-tail method. In addition, the setting of any specifications, including minimum seal strength, should be unique to the materials being sealed together. The specifications must be based upon a sound statistical rationale and validated process control, not the universal 1-lb ï¿½standard.ï¿½ This would create uniformity in the industry and conformity to a single standard.