Currently this issue is of major interest since the diffusion rates of all commonly used chemicals in almost all commonly used polymers (including Polyethylene, PPS, PEEK, PVDF, Polyamide etc.) are usually higher than the metal. Hence after a relative short time, the liner or coating becomes saturated with chemical. Then a combination of swelling and possibly temperature stress at the interface comes to play. The
stationary interface stress as a result of mass swelling can be calculated with formula like (for a long thin walled cylinder):
Scirc = [mass swelling of unrestrained sample in x direction] * [E]
For
stationary temperature stress (we have a major temperature gradient through the polymer since radiation through metal is much faster than radation through polymer):
Scirc = [0.5 (temperature swelling of unrestrained sample in x direction)] * [E / (1-v)]
These formula are similar for the interface stress in the longitudinal direction.
If an annulus is present, there is room to release the stress (at least to a certain extend), and to swell. If an adhesive is present the swelling stress must be compared with the adhesion stress, usually in the order of 0.1 J/m to 50 J/m, by setting up a proper energy balance. Hope this helps for now
Sincerely,
Sijmon
Composite Agency
Nick, maybe the following is helpful:
Titre du document / Document title:
Horizontal cylinder-in-cylinder buckling under compression and torsion : Review and application to composite drill pipe
Auteur(s) / Author(s):
WICKS Nathan ; WARDLE Brian L. ; PAFITIS Demos ;
Résumé / Abstract:
Available analytical results and experiments on the structural behavior of constrained horizontal cylinders subjected to axial compression, torsion, and gravitational loads are reviewed. Such configurations are of interest to the oil-drilling field and provide static design expressions for steel tubulars. The buckling problem is similar to restrained railroad tracks and submerged/underwater pipelines under thermal expansion. Due to outer cylinder constraint and gravitational loads, analysis has shown that long cylinders initiate buckling at loads significantly higher than classical Euler buckling loads. For these constrained long cylinders, buckling initiates in a sinusoidal mode that snakes along the lower surface of the constraining cylinder. Classic analytical expressions hold that as the axial load increases, the cylinder achieves an overall helically buckled state in which the buckled cylinder forms a helix spiraling around the inner surface of the constraining cylinder. Torsion is shown to have little effect on either buckling load but controls the sense/direction of the helical buckling. Little experimental data exist on constrained cylinder buckling, and it is unclear how the initiating sinusoidal mode transitions to the helical mode. Implications of the buckling progression for composite cylinder applications are described including the finding that composites perform poorly relative to steel on the metric of buckling due to lower density and axial stiffness; composites perform well on the metric of lock-up length when friction is considered. Based on this review and findings for composite cylinders, recommendations are made for further work.
Revue / Journal Title:
International journal of mechanical sciences ISSN 0020-7403 CODEN IMSCAW
Source / Source:
2008, vol. 50, no3, pp. 538-549 [12 page(s) (article)]
Langue / Language
Anglais
Editeur / Publisher
Elsevier, Oxford, ROYAUME-UNI (1960) (Revue)
Mots-clés d'auteur / Author Keywords
Helical buckling ; Sinusoidal buckling ; Cylinder ; Oil drilling ; Compression ; Torsion ;
Localisation / Location
INIST-CNRS, Cote INIST : 1321, 35400017369068.0150
