Paul,

You are correct in stating that for the typical buried pipeline applications, characterizing the soil restraint using a bi-linear soil model in accordance with the ASCE guidelines along with representative soil strength parameters from the site specific geotechnical report is usually adequate. I would also venture to say that more detailed pipline stress analyses which have been performed using other more rigorous finite element softwares for plastic analysis (non-linear) of buried or seabottom supported piplines subjected to large displacements also have characterized the soil response using bi-linear soil models, and have achieved reasonable results.

There have been some major advances in research during the 1980's to date regarding how to better characterize soil restraint functions for buried pipline movement problems using bi-linear soil models as previously referenced in the ASCE publications for example. The main point here is that present CAESAR II software (and operating Manual equation descriptions) generate soil response models that are based on somewhat dated information referenced to Dr. Peng's May 1978, Part 2 publication appearing in Pipeline Industry magazine. In addition, Dr. Peng's paper only provides sufficient information to completely define the bi-linear soil model characteristics for the transverse-lateral direction of pipline movement only.

The information required to completely describe the bi-linear soil response curves for the remaining important directions of buried pipeline movement (vertical uplift, vertical bearing and longitudinal-axial directions)are not provided in Dr. Peng's paper. Although these other directions of pipeline movement are generally mentioned in his paper, only an equation for estimating the soil response against longitudinal-axial movement is further suggested with no corresponding equation for the soil-pipeline diplacement at which this soil response is mobilized. In addition, Dr. Peng's paper does not provide any information on how to quantitatively define the bi-linear soil model the vertical uplift and vertical bearing directions of buried pipline displacments. Make no mistake, Dr. Peng's paper was an excellent contribution to the Industry at the time, but the information was somewhat limited, did not completley address all important directions of buried pipline movement, and the technology and research has moved on from there.

The subsequent research summarized in the more recent ASCE technical publications do provide all the information that is needed to properly generate the bi-linear soil models for all directions of pipline movement. Therefore it is recommended that these procedures (equations) be followed, and the bi-linear soil model parameters be user-defined (i.e. user specified) in CAESAR II accordingly as opposed to CAESAR II autimatic generating the bi-linear soil models especially for the vertical uplift, vertical bearing, and longitudinal-axial directions of pipeline displacements.

For buried pipline problems involving lateral pipline movements (i.e. sidebends or pipeline exposed to lateral movement due to earthquake loadings or mud slides for example), if the user desires CEASAR II automaticly generate the parameters for the lateral restraint soil springs on either side of the pipeline, the user should be cautious in the selection of the "overburden compaction factor" (OCR)to be used. It is understood (based on CAESAR II output)that when CEASAR II automatically generates the ultimate lateral soil restraint in accordance with the equation given by Dr. Peng's paper, the results are then multiplied by the OCR. This can lead to very unrealiticly high values of lateral pipline restraint. If one is studying movement of sidebends for example, it is conservative to underetimate the lateral soil restraint rather than to overestimate it by many fold.

If it is desired to provide a reasonably "high" estimate of soil response, this is conventionally accommodated by using reasonably high etimates of soil density and soil strength (soil friction angle for granular soils or shear strength for cohesive soils)in the constitutive equations to predict ultimate soil restraint rather than to mutiply the end result by some general factor.