Corus RD& T in the Netherlands have on a number of occasions undertaken analysis to help customers solve their problems with adhesion on particular composite systems. An example of the kind of analysis Corus have undertaken is given in the following example.
In this example an analysis was performed using optical microscopy on a sandwich panel locally suffering from lack of adhesion.
The client produces sandwich panels by adhesively bonding colour coated steel skins to a Styrofoam core. The steel skins are manufactured by Corus Colors and are supplied via Multisteel.
Lately problems were experienced with occasional lack of adhesion in certain areas of some panels. Therefore, an example of one of these panels was offered to the Joining group of the Corus IJmuiden Technology Centre to be analysed.
Although the exact details are not known, the basic principle of the manufacturing process of the panels is sketched in figure 1. Adhesive (a 2 component polyurethane) is applied in beads on the two colour coated steel skins of the sandwich (left in figure 1). The foam core is placed between the skins and the sandwich is laminated in a press at an elevated temperature (80°C). As a result, the adhesive is spread out and a continuous joint is created over the complete surface area of the panel (right in figure 1).
|
|
Figure 1 Sketch of a sandwich panel, before (left) and after (right) the laminating process
The following two conclusions can be drawn from the investigation.
The following four possible causes for this spread in geometry were identified and recommendations are given to test which of these is the actual cause of the problem.
Based on this advice the Client was able to resolve the manufacturing problem found in the sandwich structure.
In this example an analysis was performed using optical microscopy on a sandwich panel locally suffering from lack of adhesion.
Adhesively bonding colour coated steel skins to a Styrofoam core
Colour coating of skin. Other processes not specified.
2 component polyurethane
A 2 component polyurethane is applied in beads on the two colour coated steel skins of the sandwich (left in figure 1 above). The foam core is placed between the skins and the sandwich is laminated in a press at an elevated temperature (80°C). As a result, the adhesive is spread out and a continuous joint is created over the complete surface area of the panel (right in figure 1).
Not specified. Pressure cure during 80ºC laminating process
The bonding was carried out at the manufacturing facility
Figure 2 shows a picture of the investigated panel. The panel has been pulled apart after the adhesive was cured. Only one of the steel skins has been received and could be investigated.
In this specific case, the dimensions of the skins have been larger than of the core. As a result, on the left side of the panel the adhesive hasn’t been pressed and is still present as beads.
Two distinct different failure modes can be observed in the remainder of the panel. By fare the largest part has been properly bonded which resulted in failure through the blue foam core (being the weakest link). Giving a blue surface in Figure 2. However, in the bottom right area of the panel the adhesion has not been good. Areas with good adhesion (blue) are intermixed with areas where apparently no adhesion has taken place. Figure 3 shows a detail of part of this area (detail 2 in Figure 2).
To study this phenomenon more closely 2 specimens were cut from the steel skin of the panel (detail 1 and detail 2 in Figure 2) and offered for an optical microscopy investigation. The findings from this investigation are discussed below.
Several cross sections were prepared using the specimens from detail 1 and 2. Figure 4 shows a cross section of an ‘undisturbed’ adhesive bead (using material from detail 1). At a higher magnification (Figure 5), the steel base, the zinc layer, the paint layers (primer and top coat), the adhesive, and the resin used for embedding the samples can clearly be distinguished.
Figure 6 shows a cross section of an area with good adhesion (using material from detail 1). (The area in the middle of the picture is a ridge that has been machined into the foam prior to the laminating process. These ridges are also visible in figure 2.) Again, at a higher magnification the different layers are visible (Figure 7). Note that a separate adhesive layer is hardly distinguishable but that the adhesive is absorbed into the foam up to a thickness of approximately 150µm.
Figure 8 shows a cross section (using material from detail 2) where an area of ‘good’ adhesion (on the left) changes over to ‘bad’ adhesion (on the right). In contrast to figure 6 and 7, in this case of ‘good’ adhesion (left in figure 8) a separate adhesive layer is clearly visible. The adhesive layer is actually even thicker, 200 to 300 µm, than the layer of absorbed adhesive in figure 7 (approximately 150 µm). At a higher magnification it can be seen that apart from the separate adhesive layer, adhesive is absorbed into the foam up to a thickness of approximately 50 µm (figure 9). Two pictures with a higher magnification from the ‘bad’ adhesion area (right in Figure 8), show that the steel is completely covered with a thin layer of adhesive (compare Figure 10 and Figure 11 to Figure 5).
An interesting result is the fact that the steel in ‘bad’ adhesion areas is coated with a thin layer of adhesive (figure 10 and 11). That is a clear indication that it is very unlikely that the current problem is an adhesion problem at all. A problem that can occur when bonding a surface that is not clean (or rather: that has a surface energy that is too low) is that the adhesive will not be able to wet the surface completely (does not spread out over the surface). If that is the case, in some areas no adhesive will be present and as a result locally no joint will be formed. However, in this case the adhesive is present over the complete surface area of the steel and thus the cleanness of the steel (or the type of coating used) simply cannot be the cause of the problem.
A large difference in adhesive thickness was observed between areas of good adhesion (having a very thin separate adhesive layer, figure 7) and areas of ‘bad’ adhesion (where the distance between steel and foam is much larger, figure 8). That indicates that the ‘adhesion problem’ is actually more likely to be a geometry problem. The mechanism that causes areas without a joint seems to be the following: if the gap (the distance between steel and foam) becomes too large, the volume of the adhesive is insufficient to fill it completely and areas without a joint are formed.
It is not clear from these results what causes the creation of a gap. The dimension of the base materials may be involved (thickness and flatness of steel and foam) but also the production parameters may play a role. Underneath an overview is given of possible causes for the creation of these gaps and recommendations are given to test which of these is the actual cause of the problem:
1. Dimensions of the base materials.
If the thickness of either the steel or the foam is not constant, gaps may be formed. By performing random checks on the materials used in production this can easily be verified.
2. Flatness of the press.
Obviously, if the plates of the press that is used are not parallel a gap may be formed on one side of the sandwich panel. By verifying the alignment of the press this can easily be checked.
3. Insufficient pressure.
If the pressure used to press the panels is insufficient to completely counteract any possible inaccuracies in flatness of either the foam or the steel, it may results in gaps. By performing some tests with increased pressure this can be verified. However, it seems actually rather unlikely that insufficient pressure is the cause of the problem since it doesn’t take much force to flatten a thin sheet of steel.
4. Insufficient pressure time.
If the panels are removed from the press before the adhesive is sufficiently cured, gaps may form due the inaccuracies in flatness of the base materials or due to temperature effects. The panels are pressed at an elevated temperature, due to differences in the coefficient of thermal expansion between steel and foam, stresses may occur in the panel during cooling down. If the adhesive is not sufficiently cured it may not be able to withstand these stresses and gaps may be formed.
Performing some test using different pressure times (e.g. curing times) will shed more light on this subject.
Optical microscopy provided a useful understanding of the causes of the ‘adhesion problem’ in the sandwich structure
Adhesion is not always the issue. In this case the issue was varying separation distance of the two adherends.
Based on the advice arising the Client was able to resolve the manufacturing problem found in the sandwich structure..
From the investigation performed using optical microscopy on a sandwich panel locally suffering from lack of adhesion. The following two conclusions could be drawn:
The following four possible causes for this spread in geometry were identified and recommendations are given to test which of these is the actual cause of the problem.
Courtesy Corus Research and Development Technology, IJmuiden Technology Centre, Netherlands 2002
Contact names: R. Teunissen, J. Vrenken
Corus Research, Development & Technology
IJmuiden Technology Centre, Product Applications, Joining Technology
P.O. Box 10.000, 1970 CA IJmuiden, The Netherlands, T + 31 251 494785, F + 31 251 470432
