American Honda Motor Co.’s 2018 Accord sedan is the company’s first to use adhesive bonding in body structures.

Manufacturers and suppliers are moving to thinner-gage composites for exterior body panels to reduce vehicle weight. Because of this, the industry has sharpened its focus on the causes and mitigation of adhesive bond-line read-through (BLRT) in sheet molding compounds (SMC).

BLRT is the transference of the bond path through the exterior body panel to the show surface. It shows as a slight depression, best seen from a glancing angle across a glossy painted surface. BLRT is a deal-breaker for customers, and its presence causes part rejection, rework, and additional cost. Eliminating BLRT, or reducing it to an acceptable level, allows designers to move beyond aluminum and high-strength steel.

Stress from differential thermal expansion coefficients between the substrate and adhesive or dissimilar substrates, can result in relative movements of the joint element1. Adherents with different thicknesses or heat capacities will heat or cool at different rates. During cooling, one substrate may cool and contract at a faster rate than the second, exposing the adhesive to deformation forces. A larger volume of adhesive will thermally shrink more than a smaller volume as it cools, so the differential shrinkage will upset the adhesive-bead uniformity and create an area prone to BLRT.

Other BLRT causes are bond-line squeeze-out and net-out, both associated with differences in the bond gap. Expect a higher degree of BLRT at thicker sections of adhesive. Bond-line standoffs may telescope through to the show surface.

Mitigating BLRT in SMCs requires a three-pronged approach involving design, process, and adhesive. In practice, the approach entails:

  • Identifying the required adhesive properties
  • Creating an adhesive with the optimized properties to mitigate BLRT
  • Designing the part and production process to work best with the adhesive

Simply using a drop-in replacement adhesive in existing processes will yield an unsatisfactory result.

Identifying the right adhesive

To identify an adhesive that reduces BLRT and demonstrates fast cure and heat resistance through automotive electrocoating (ecoat) ovens, Ashland initiated a research program to find a solution. The research team formulated several structural adhesives, including model epoxy and polyurethane formulations, as well as hybrid products combining both technologies.

The study focused on bonding glass-filled polyester SMC. Two different SMCs were bonded together to form the test panels. One side representing the exterior body panel was a low density, 1.2g/ml, toughened Class A formula and was molded at 1.8mm. The interior panel was a higher strength structural SMC molded at 2.8mm with density of 1.77g/ml. Continental Structural Plastics of Auburn Hills, Michigan, provided both SMCs.

First, researchers needed to ensure that the adhesives and assembled panels could pass ecoat, typically a 40min. exposure at 204°C, followed by a lap shear strength of 0.3MPa at 204°C. If the adhesive cannot survive ecoat, any benefits derived from fast cure or BLRT reduction is negated.

After ecoat testing, the Ashland team settled on five adhesives – four two-part polyurethane/epoxy hybrid systems and one epoxy – with the mechanical properties (Table 1). Hybrid systems containing both epoxy and polyurethane chemistries offer lower modulus and higher elongation, plus higher temperature stability, all creating a more balanced structural adhesive.

The adhesives met the research team’s goal of an 8-minute working time and heated strength build of 90 seconds. They had working times from 8 minutes (H2, with the lowest tensile strength and modulus) to 60 minutes (the epoxy, Ep2, which had the highest tensile strength and modulus).

Measuring BLRT

Ashland used an ABISOptimizer surface analyzer from Zeiss Industrial Metrology to measure BLRT. The analyzer takes an electronic slice across the surface topography where a stone is placed, producing graphical altitude and profile data. The altitude data subsequently converts to its second derivative to emphasize the change in slope or curvature – the lower the better.

Researchers used flat 25cm x 30cm panels, with the 1.8mm thick panel bonded to the 2.5mm thick panel and an adhesive bead drawn down the panel center from edge to edge. The adhesive-panel assembly was cured in a fixture press preheated to 127°C for 90 seconds, unless stated otherwise, and post-baked 40 minutes at 204°C. The thick and thin sides of the panels were measured. There was less curvature on the thick side with every adhesive tested. Ashland conducted a regression analysis plotting curvature vs. Young’s modulus, that showed the lower modulus adhesives helped reduce curvature.

Researchers used a Zeiss Industrial Metrology ABISOptimizer surface analyzer to measure adhesive bleed through.

Ashland researchers then investigated the lower-modulus adhesive H2 for robustness by varying its fixture cure time at 127°C. Except for the 1-minute cure time, as cure time increased so did the curvature of the control panel. A first regression analysis, including all cure time data, did not have a high level of correlation. A second analysis without the shortest cure time data, showed reduced curvature with a shorter time in the cure fixture down to 1.5 minutes. Fixture times less than 1.5 minutes likely produced adhesives that were not set when the panels were removed from the fixture, and this may have disrupted the bond line and resulted in elevated curvature measurements.

Using the low-, medium-, and high-modulus formulas – H2, H3, and Ep2 – Ashland evaluated bond gap thickness versus curvature, with bond gaps from 0.43mm to 3.00mm. Curvature increased as bond gap did, so reducing bond gap reduces curvature.

The scientists also evaluated the effect of adhesive bead width using the low modulus H2 formula. They varied bead widths from 13mm to 38mm. They found no significant relationship between adhesive bead width and curvature, which is counter to previously published data in the literature2.

The team tested H2 for the relationship between curvature and bond gap with and without post bake at 204°C. They varied the bond gap from 0.42mm to 3.00mm and found that with and without post-bake, curvature worsens as the gap increases. At every thickness the curvature deteriorates with the addition of the post bake.

Even with a 2-minute cure at 127°C, thermodynamic events take place within the adhesive or SMC. The data confirm the benefits of lowering the temperature of the post-bake oven and show the importance of additional work adjusting the adhesive and composite cure profile. At the minimum tested bond gap, curvature remains less than the threshold target value of 0.6/m, with and without the post bake. Therefore, bond gap is a key factor in controlling BLRT.

Ideal adhesives

The Ashland research team determined the following relationships:

  • The shorter the cure time, the lower the curvature
  • Curvature increases as adhesive bond gap increases
  • The lower the modulus, the lower the curvature
  • Adhesive bead width has no effect on curvature

These observations were made from adhesive formulations developed by Ashland. To get the maximum return from the research, it is important to select an adhesive with the properties defined exclusively for this work, available from Ashland.

Using Ashland’s high-elongation, low-modulus adhesive, auto engineers can focus more on composite substrates. Composites now provide engineers and stylists with more design freedom – the ability to create sharper lines than with a steel stamp.

Composite manufacturers must focus on process and part design in addition to adhesive selection. First, eliminate bond-line squeeze out and ensure uniform bond gap across the assembly. Second, ensure uniform heat across the bond fixture. Process engineers must account for the coefficient of thermal expansion of heating blocks when designing bond fixtures to provide uniform pressure and heat transfer across the assembly. This helps minimize bond gap variation and in return, BLRT. Parts should mate together correctly prior to the bond operation and not be forced into shape in the tool.

BLRT followed the path of least resistance and was more severe on the thin side of the assembly. Increasing the gage of the show side while reducing the interior panel gage will have positive effects. Engineering constraints may preclude this design change.

Following these design, process, and adhesive parameters and eliminating BLRT at the design stage provides the industry with tools to reduce weight through composite gauge reduction while mitigating adhesive-related surface defects.

Ashland Inc.

www.ashland.com

Continental Structural Plastics

www.cspplastics.com

Zeiss Industrial Metrology

www.zeiss.com

About the author: Michael J. Barker, a research fellow at Ashland Inc. in Dublin, Ohio, can be reached at 614-790-1894 or mjbarker@ashland.com

1. Durso, S., Howe, S., Pressley, M., Adhesive Bond-Line Read Through: Theoretical and Experimental Investigations, SAE Technical Paper Series, SAE 1999-01-0984 (1999); Hahn, O., Orth, T., Avoiding Bondline Readthrough On Thin Steel Elements by Modification and Optimization of Processes, Adhesives and Sample Shape, International Body Engineering Conference, pp 111-115, Stuttgart, Germany (1997). 2. Fernholz, K. Lazarz, K. Wang, C.S., Preliminary Results from an Experimental Evaluation of the Root Causes of Bond-Line Read-Through, The 32nd Annual Meeting of the Adhesion Society, Savannah, GA, 15-18 Feb. (2009), pp. 285-287.