Page 200 - Handbook of Adhesives and Sealants

170 Chapter Four

4.6 Prototype and Non-standard
Test Methods
One must keep in mind that the final product will never be a T-peel
specimen or a lap shear specimen. ASTM and other standard test
methods are excellent tests to offer relative comparisons of adhesives
and/or bonding conditions. They also offer a valuable starting point
for adhesive evaluation. However, there may be better ways to obtain
information for predicting performance of the actual joint in service.
The standard test methods are well defined and, therefore, can be well
controlled. However, their relationship, if any, to performance in the
product must be established by trial and error or advanced analytical
means.
The methodology of the test and evaluation program requires care-
ful thought. A critical feature is that the tests must have a known
relationship to the final product. Often this requires either developing
creative, non-standard tests that stress the part in a mode that is more
indicative of its service load, or producing actual prototype specimens
with the adhesives and bonding conditions that are intended to be
used in production. These prototype specimens then would be sub-
jected to simulated service environments. The environmental exposure
can be accelerated to reduce testing time. Caution needs to be exerted
so that the acceleration does not cause reactions or mechanisms
within the materials or bond-line that would not actually be present
in the intended real environment. These ‘‘non-standard’’ test methods
should be controlled so that the tests are repeatable and the variabil-
ity is low. Among the obvious variables that need to be controlled are
surface cleaning, joint geometries, method and extent of material mix-
ing, method of application, fixtures utilized, and cure conditions.
To develop a joint with an adequate service life and with a realistic
23
design margin, the use of a ‘‘Mathes’’ ladder is suggested to establish
a testing hierarchy. In this process, shown in Fig. 4.23, testing pro-
ceeds from simple, standard tests of basic materials where well-
defined test specifications are available to increasingly complex tests.
Depending on the application and the type of information available
from the lower rungs of the ladder, the need for more complex testing
may be reduced or even eliminated. However, the need to completely
understand the simpler tests is mandatory. Unexpected failure in ser-
vice is often associated with a lack of understanding of the effects of
the service environment on the basic materials or on a lack of under-
standing of the test variables (e.g., rate of loading).
The most difficult failure situations to predict are those that result
from interactive effects. Thus, it is important to consider and evaluate
the adhesive or sealant joint as a ‘‘system’’. There is a thought pro-