TWI Knowledge Summary

Comeld ^TM

by Faye Smith

Introduction

Historically, the composite-to-metal joint has presented significant design challenges to achieve high levels of mechanical performance. This has meant that designers have been reluctant to design structures incorporating joints or have adopted highly conservative designs which increase weight and thus negate some of the benefit of using composite materials.

TWI has developed Comeld, a proprietary material surface treatment technique and joining process which offers the potential for joints to be made between fibre reinforced plastic (FRP) composite materials and metals, with enhanced performance.

Steps to create a Comeld joint

Comeld is the application of a proprietary material surface treatment technique ^[1] called Surfi-Sculpt ^® . This technique uses a power beam (e.g. an electron beam) to create protrusions (known as 'proggles') and cavities in the metal onto which the composite is laid and cured, forming the Comeld joint.

The steps followed in creation of a Comeld joint are as follows:

1. The metallic part of the joint is machined to the correct shape for the style of joint to be formed. Examples are shown in Figure 1.

Fig.1. Examples of shape of machined metal parts for creating Comeld joints

2. An appropriate Surfi-Sculpt treatment is then applied to the required part of the machined metal part. Examples are shown in Figure 2.

Fig.2. Examples of Surfi-Sculpt treatment on different metals

3. If an extra adhesive layer is required in the Comeld joint, adhesive is applied to the treated surface of the metal. In the case where the resin of the composite is to act as the adhesive in the joint, this step is not necessary.

4. The composite material is laid-up onto the treated surface of the metal. The exact details of this will depend upon the composite material and the processing technique being used. Figures 3 and 4 show glass fabric being laid up onto a single step stainless steel joint, and a schematic of the lay-up covered with a vacuum bag in order that the composite can be created using the vacuum infusion process.

Fig.3. Glass fabric being laid up onto treated stainless steel

Fig.4. Schematic of manufacture of a Comeld specimen via the vacuum infusion process

5. The composite material is cured, consolidating the Comeld joint. The cure process will depend on the resin being used and may require heat and pressure.

Figure 5 shows Comeld joints between stainless steel and glass fabric reinforced vinylester and polyester respectively. Joints have also been prepared between carbon fibre reinforced plastic (CFRP) and aluminium and CFRP and titanium. Figure 6 shows a high magnification image of the interaction between a titanium proggle and CFRP plies in a laminate (carbon/epoxy UD prepreg laid up in 0/90 configuration).

Fig.5. Double step Comeld joints between stainless steel and glass fibre reinforced plastics
	a) Stainless steel/non-crimp glass fabric reinforced vinylester Comeld joint
	b) Stainless steel/woven glass fabric reinforced polyester Comeld joint

Fig.6. Interaction between a titanium proggle and CFRP plies

Figure 7 shows the results of tensile testing on the joint shown in Figure 5b, along with control joints which were made using exactly the same material and processing but did not have the Surfi-Sculpt treatment applied. All specimens had a surface treatment applied that was appropriate to aid adhesion between the polyester resin and the stainless steel.

It can be seen that the Comeld joints failed at a much higher load than the control joints and absorbed three times as much energy (corresponding to the area under the load-displacement curve) as the control specimens before failure. The Comeld specimens also failed in a more progressive manner than the sudden failure seen in the control joints. This could have benefits in terms of in-service detection of deterioration in a structure before failure occurs.

Fig.7. Results of tensile testing on GFRP/stainless steel double step joints

Figure 8 shows the results of tensile testing on control and Comeld single step joints between titanium and CFRP (combination shown in Figure 6). It can be seen that in this case the Comeld sample and the control sample failed at approximately the same load, however, as with the previous material combination, the Comeld joint absorbed three times as much energy as the control joint before failure.

Fig.8. Results of tensile testing on CFRP/titanium single step joints

An explanation as to why the stainless steel/GFRP Comeld joints demonstrated a higher failure load than the control specimens and, in contrast, the titanium/CFRP Comeld joints only reached the same failure load as the control joints, can be obtained by observing the way in which the specimens failed. It can be seen from Figure 9 that the control specimens for both material combinations failed at the interface between the composite and the metal.

Fig.9. Photographs of failed control joints
	a) GFRP/stainless steel control joints failed at interface
	b) CFRP/titanium control joints failed at interface

Figure 10 demonstrates that the GFRP/stainless steel joints failed within the composite material. The mechanical interaction of the proggles with the composite material meant that the interface between the composite and the metal was no longer the weakest part of the joint. It can be seen that after failure there is still composite material attached to the proggles and that the predominant failure mechanism was shear failure of the composite in the region where the proggles finished. Inspection of the failed specimens also showed that the proggles had undergone plastic deformation during failure of the specimen.

Fig.10. GFRP/stainless steel Comeld joint failed in composite

Figure 11 demonstrates that the CFRP/titanium joints failed within the titanium. As with the GFRP/stainless joints, the mechanical interaction of the proggles with the composite material meant that the interface between the composite and the metal was no longer the weakest part of the joint. However, in this case the metal failed before the composite. Closer inspection of the Surfi-Sculpted titanium sections revealed that the high level of Surfi-Sculpt treatment applied to the section to be joined meant that the metal had significantly reduced cross-section. This decreased the load carrying capability of the titanium, which resulted in the unexpectedly low failure load.

Fig.11. CFRP/titanium Comeld joint failed in metal

The different failure modes demonstrated by these two different Comeld joints show that, with further research and understanding of Comeld joint design, it should be possible to design Comeld joints with a predetermined failure mode in a specific section of the joint.

The information described in this TWI Knowledge Summary was the result of early work to demonstrate the feasibility of Comeld joints. In-house, collaborative and single client work is currently being performed to generate data, understand and model this technology and apply it to industry-focused applications.

Further information

For more information on Comeld please email Ewen Kellar ewen.kellar@twi.co.uk at TWI.

Reference

1. Dance B G I and Kellar E J C: 'Workpiece Structure Modification'. International Patent Publication Number WO 2004/028731 A1