Abstract
The method presented by DNVGL in DNVGL-ST-F101 [1], “Submarine pipeline systems”, 2017, for calculating the collapse pressure of submerged pipelines is well-known for design of pipes intended to operate in very deep water. Such pipes are regarded as quite thick-walled with diameter to wall thickness ratio in the range of 15 to 30. There is now substantial experience in the practical manufacture, installation and operation of such pipes. Recently there has been a growing use of large diameter pipelines to transport high volumes of gas over great lengths at moderate water depths. The pipes are considered to be thin-walled with ratios of diameter to wall thickness in the range of 30 to 45. This paper assesses the validity of the DNVGL design method when applied to the design of such thin-walled pipes.
A particular aspect of the buckling pressure of large diameter pipes is the effect of the Bauschinger phenomenon. The phenomenon occurs when pipes made using the UOE method are subjected to internal pressure, to provide expansion of the pipe during manufacture, thus reducing the out-of-roundness of the pipe wall, and subsequently subjected to external hydrostatic pressure during pipeline operation. To date the Bauschinger phenomenon has been recognised as resulting in a reduction of the circumferential compressive yield of the pipe material. This reduction is accommodated in the DNVGL design formula. Recent research into material properties has shown that the Bauschinger effect also has the effect of reducing the modulus of steel materials over a range of values of applied circumferential compressive stresses.
The paper reviews the basis of the Bauschinger phenomenon and presents results from very detailed accurate testing of UOE pipe material. The tests determine the levels of modulus for pipes subject to circumferential compressive stresses. Although results for compressive stress-strain values have previously been available for pipes subject to high levels of hydrostatic pressure it has been considered that the Bauschinger effect is not generally significant for thick-walled pipes. The tests reported here consider the calculation of material modulus levels for low levels of stress that correspond to the buckling stress of thin-walled pipes. The calculated collapse pressure for such pipes is examined in this paper and compared to corresponding results from the DNVGL design formula to provide guidance on the effect of design levels of pipe wall thickness due to inclusion of the Bauschinger effect. The comparisons are for example pipe wall thickness and material conditions. Conclusions are drawn that including the Bauschinger effect in the calculated pipe wall thickness can have a beneficial effect with regard to pipe manufacturing and installation costs for pipe subjected to mild heat treatment.