Use this URL to cite or link to this record in EThOS:
Title: Modelling the interfacial degradation in adhesively bonded joints
Author: Liljedahl, Carl David Mortimer
ISNI:       0000 0001 3609 9776
Awarding Body: University of Surrey
Current Institution: University of Surrey
Date of Award: 2006
Availability of Full Text:
Access from EThOS:
Access from Institution:
The aim of the research was to develop predictive models for the interfacial degradation of adhesively bonded joints when exposed to aggressive environmental conditions. Four different joint configurations using the same adhesive system were exposed to a variety of conditions including immersion at 50°C, 96%RH at 50°C and 80%RH at 70°C. In addition data from joints for other adhesive systems were also incorporated into the investigation. Moisture has a degrading effect on the strength of adhesively bonded joints. Therefore the diffusion into the bulk material was determined by gravimetric experiments. However, the mobility of the water molecules at the interface between the adhesive and the substrate may be higher than in the bulk material. In order to assess this, the spatial moisture distribution in bonded epoxy laminates was detennined by a nuclear reaction analysis (NRA) technique. The moisture profile found experimentally and the modelling undertaken of the interfacial diffusion indicated that the ingress in the interfacial region was a few times faster than in the bulk material for the adhesive system investigated. Both hygroscopic (swelling) and thermal residual strains may affect joint durability. The thermal expansion was determined by means of a bi-material beam and the hygroscopic expansion was determined by measuring the expansion of bulk samples at various moisture levels. Creep properties for the adhesives studied were determined to investigate the relaxation of residual stresses during the aging process. The coefficients of thermal expansion and hygroscopic expansion were of the same order of magnitude for the adhesives investigated. Creep was seen to be enhanced in the presence of moisture. The AVl19 adhesive was seen to creep much more than FM73 and also absorbed more moisture. As a consequence, the residual stresses in the joints bonded with A Vl19 were seen to relax nearly totally whilst the residual stresses in the joints bonded with FM73 relaxed to about half of their original magnitude. Different interfacial fracture tests were carried out in order to assess which was most appropriate. Notched coating adhesion tests (NCA) were carried out initially. However, it was very difficult to produce a repeatable notch and the adhesive often cracked before the coating debonded. Good results were obtained then these samples were immersed in water. Another test investigated was a split beam specimen. However this test was of limited use as the secondary bond was weaker than the aged interface of interest. Finally, a mixed mode flexure specimen (MMF) was selected to determine the fracture energy of the adhesive systems in the 80%RH and 96%RH environments. The fracture energy degraded rapidly initially with moisture content and then at a slower rate as more moisture reached the interface. The fracture energy was found to be a function of the amount of moisture at the interface. No further degradation was found when the joints were held at equilibrium. The degradation and the progressive damage were simulated with a cohesive zone model (CZM). The model was extended from 2D to 3D. This was ~eful when predicting where 2 the crack initiated in the width direction and how the initiation site changed after aging for a L-joint configuration. When using a CZM the interfacial strength was defmed by a traction-separation law. The parameters governing the traction-separation law were determined using the interfacial fracture tests (NCA and MMF). The parameters were the tripping traction and the fracture energy. It was shown to be essential to incorporate elasto-plastic adhesive continuum behaviour in order to simulate the complete joint response correctly. The tripping traction was determined by correlating the deviation of the load-displacement curve with the simulated result. The fracture energy was then determined by correlating the experimental load-crack length response with the simulation. This gave a unique set of moisture dependent CZM parameters for various moisture concentrations. These parameters were then used to predict the response of other joint configurations. For most of the joints, the residual strength was predicted closely using the moisture dependent CZM parameters. However, in some cases other degradation mechanisms were active. These included stress enhanced degradation and cathodic delamination. When these mechanisms were included in the modelling, the prediction of the durability of all joint configurations was good.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available