« BACK TO EXAMPLE OF FORENSICS ANALYSIS
E10 GRP process plant piping
Background
In recent years, the use of glass fibre reinforced composite piping has been gaining favour in a number of industries. These systems offer benefits in the form of reduced weight, improved corrosion resistance and longevity, as well as reduced costs. Unlike conventional metallic pipework, however, it is not possible to weld the composite pipe. Conventional compression, or gasketed flange connections are less desirable in view of the increased potential for leakage. As a result the preferred method for joining composite pipes is by adhesive bonding.
In general there have been few problems associated with the longevity of the adhesive bonding. Most problems are identified during the initial pressure testing, and can be traced back to haphazard preparation of the joint prior to bonding. The samples obtained for this investigation came from the pipework for the flue gas desulphurisation plant at Drax, where extensive use of bonded glass fibre reinforced pipe has been made.
What was Done
The spool piece which was used in this investigation came from a section of pipework located in a protected external location, and thus saw normal ambient temperature conditions, but not direct sunlight or rain. The pipe was designed to carry a relatively chemically benign, heavily particulate loaded fluid, at a relatively high flow rate. This particular section of pipe was located in a region of significant turbulence, and was removed along with a nearby pipe bend after six months duty following rapid wear of the bend. During this time there were no reported problems associated with the adhesive bonds on this section of pipe.
Both ends of the spool piece were inspected using an ultrasonic transmission imaging method.
Joint Details
Adherends
Glass fibre epoxy composite
Adhesive
Filled cold curing epoxy
Pre Treatment
Surface abrasion
Bonding
Most systems available make use of bell and spigot type joints. Pipes and fittings are formed with bell ends into which is inserted a mating tapered section of pipe. The spigot end is either formed in the factory or prepared on site using a mechanical shaving device. The surfaces of both the spigot and the bell end are abraded, prior to mixing and applying the adhesive. The pipes are then brought together and pulled tight using winch pullers if necessary. Curing the adhesive is accelerated by the use of heating blankets, which are wrapped around the pipe, and covered with insulation.
Cure Cycle
Room temperature cure
Location
The bonding was carried out at the field location.
Results
Due to the fact that the adhesive layer is very thin, and has a similar density to the adherend material, it is not easily identified. Nonetheless, the NDT was able to identify regions of less good and above average bonding. The inspection confirmed that other than in a few localised positions, principally at the lower edge of the joint, corresponding to the open edge of the bell, the remainder of the joint appeared to be sound. These results were confirmed through the subsequent mechanical testing of sections of the joint.
The first of the samples to be tested did not have saw cuts to reduce the overlap length. As a result, the sample failed within the composite tube, rather than at the joint. Whilst this result does not give any indication of the condition of the adhesive, it does confirm that if correctly bonded in the first instance, adhesive jointing of composite pipes can result in joints at least as strong as the pipe itself. It should also be noted that, according to the NDT results, this sample came from a region where the bonding quality was average, rather than particularly good. A simplistic calculation of the total joint longitudinal load capacity indicated that the joint could sustain a load of twice the allowable longitudinal stress in the pipe.
The mechanical test results were compared with test data generated as part of the original development of the piping system using conventional lap shear tests with varying bondline thicknesses. This suggested that there might have been around an 18% reduction in joint strength. Care must be exercised in drawing this conclusion, however, as the adherends used in the two sets of samples were different configurations of GRP, with differing stiffnesses and strengths, and the joint geometry was slightly different due to the curvature of the pipe. Also, it is considerably easier to manufacture a perfect joint when the surfaces to be joined are small and flat.
Inspection of the fracture surfaces revealed that in the less strong specimens the failure was predominantly on the interface between the adhesive and the fitting. This was confirmed using XPS analysis of the surface nano-layers, although there was some evidence for partial cohesive failure of the adhesive close to the interface in places. There was no evidence of contamination of the bonding surface. In the stronger samples, the principle locus of failure was within the surface veil layer of the pipe, probably due to incomplete wetting of the veil fibres. Where the main pipe filament wound fibres were exposed, they had been well adhered to the adhesive, resulting in significant fibre pull out during failure of the joint. There was evidence that there had been adhesive starvation in some areas, possibly due to unevenness of the surfaces to be bonded.
Discussion
The results of this study clearly illustrate that, in order to achieve the maximum long term strength from an adhesive joint, the adherends must be carefully prepared. In practice the near perfect joints achievable in the laboratory are not possible in the field. Nonetheless, strengths are achievable which compare with those of the underlying adherend material.
Lessons learnt
- In order to achieve the maximum long term strength from an adhesive joint, the adherends must be carefully prepared.
- In practice the near perfect joints achievable in the laboratory are not possible in the field.
- Strengths are achievable which compare with those of the underlying adherend material.
Source
MTS Project 3 Report No 9 Forensic Studies of Adhesive Joints. Part 1 - General Introduction and Conclusions Part 8 - GRP Process Plant Piping
NPL/ ESR Technology Limited
Colour image Courtesy Petrobras
« BACK TO EXAMPLE OF FORENSICS ANALYSIS
