Sliding Connections |
 
|
Sliding Connections |
|
Sliding connections are similar to standard ones, with the added advantage that the connections are interchangeable. This is appropriate for modelling scenarios where there is significant relative axial motion between primary and secondary structures.
When setting up sliding connections, you associate a node on the primary structure (or a group of nodes if you avail of the generate option) with a set of nodes on the secondary structure. You may have noticed that there is currently no facility for defining node sets in Flexcom, although the use of element sets is widespread throughout the program. So you actually define a relevant element set instead, and an appropriate node set is created internally for you. Based on the proximity of the various nodes at the beginning of the analysis, the program creates an initial set of (effectively standard) connections between the primary and secondary structures. The modelling procedure is then very similar to that of the standard connections – resistance to relative motion of the pipes in the lateral plane is achieved via an appropriate augmentation of the global stiffness matrix. The main difference lies in the fact that the connections are interchangeable. The program continually monitors the relative axial locations of the connected nodes over the course of the analysis, and the set of connections is updated as and when required.
Another advantage of using sliding connections is that it is also possible to model a degree of resistance to relative motion of the pipes in the axial direction. This may be may characterised by a linear axial stiffness value or a non-linear force-deflection relationship.
Connections are constructed based on the instantaneous locations of the primary and secondary nodes. The distance between a primary and a secondary node is computed in a local axis system which is formed based on the primary element. The modelling procedure is similar to that discussed in Contact Modelling, and interested readers are referred to that section for further details as required.
Once the initial set of connections has been established the analysis proceeds as normal. After every solution time step, the set of connections is revised and updated to reflect the latest nodal positions. A margin of 5% is used to avoid connections alternating back and forth at the changeover point, when there are two secondary nodes which are almost equidistant axially from the primary node. So once a connection has been established between a primary node and a secondary node, a competing secondary node must be at least 5% closer to the primary node before the interchange happens.
By definition, sliding connections are based on primary nodes. It is also important to note that each primary node can only be connected to one secondary node at any given time. So for example, if the mesh density on the secondary section is more refined than the primary section, the number of active connections will be equal to the number of primary nodes. If the primary section has a higher mesh density, then it is possible that some secondary nodes can be connected to more than one primary node. Advanced users who wish to examine the active connections at any point during the simulation may use the *PRINT keyword.
A situation can sometimes occur where a secondary section is being inserted into a primary section (such as a J-Tube Pull-In), or being removed from it. In such circumstances, some or all of the secondary section is completely detached from primary section, and the connections are deemed to be inactive. For each primary node, Flexcom computes the axial distance between it and the nearest secondary node. If this distance is greater than a threshold value, then the connection is deemed to be inactive. The threshold value is equal to the length of the longest element in the secondary element set. Assuming for simplicity that all elements in the primary and secondary sections have equal length, connections start to become inactive when the nearest secondary node is more than one element length away from the primary node.
While a connection is inactive, it makes no contribution to the global stiffness matrix. Refer to Contact Modelling for further details. Note also that even if a connection becomes inactive at a particular time, it may become active again subsequently.
Sliding connections are appropriate for modelling scenarios where there is significant relative axial motion between the primary and secondary structures. For example, J-Tube Pull-In scenarios are well suited to sliding connections.
In certain circumstances, it may be desirable to allow the software to initially determine appropriate connections between a primary and secondary section, and to subsequently treat these connections as standard connections. For example, when setting up a landing string model, the initial set-down of the drill string within the marine riser can make it very difficult to manually identify the optimal set of pipe-in-pipe connections in advance of the initial static analysis. Designating the connections as sliding allows the program to automatically determine the optimal connections, minimising effort on the part of the user. While there may be significant relative axial motion between the initial model definition and the converged static solution, there is comparatively little axial motion during the actual simulation itself (e.g. when the model is subjected to wave loading). Invoking the *NO PIP SLIDING keyword in subsequent restart analyses thereby ensures computational efficiency (any overhead associated with the monitoring of nodal locations is eliminated), and can also provide enhanced numerical stability (connectivity of the finite element model remains consistent throughout the simulation). For example, Completion/Workover Risers would be suitable for modelling using Sliding Connections, in conjunction with the *NO PIP SLIDING keyword.
Regarding sliding connections, there may be a slight computational overhead associated with the continual monitoring of relative axial locations over the course of the analysis. For this reason, it is recommended that you utilise standard connections in situations where significant relative axial motion is unlikely to occur – for example, if there are bulkheads present. For example, a Top-Tensioned Production Riser would be suitable for modelling using Standard Connections.
Flexcom, like most comparable finite element codes, has a bandwidth optimisation facility that internally renumbers the node and elements of the finite element mesh to minimise the required array storage. By default this minimisation occurs once only (in the initial static analysis) because mesh connectivity does not change thereafter in the overwhelming majority of cases. One exception is when sliding pipe-in-pipe connections change nodes, as discussed above. Again by default, Flexcom does not repeat bandwidth optimisation every time this happens, because the potential saving in array storage is likely to be small, and the computing overhead associated with the process could significantly affect analysis run-time. But you can override this default and instruct Flexcom to repeatedly perform bandwidth optimisation, via the BANDWIDTH=UPDATED option under the *PIP CONNECTION keyword. Doing this only make senses if (i) sliding pipe-in-pipe connections change between nodes only intermittently, but (ii) the magnitude of the change when it occurs is significant enough to make the optimisation worth the computing effort. As these conditions will be very rarely met, the option will be very rarely invoked.
•*PIP CONNECTION is used to define pipe-in-pipe connections between nodes of the finite element model.
•*PIP STIFFNESS is used to define force-deflection curves for non-linear pipe-in-pipe connection stiffnesses.
•*NO PIP SLIDING is used to disable the interchangeable nature of sliding pipe-in-pipe connections.
•*PRINT is used to request additional printed output to the main output file. Specifically, the OUTPUT=PIP CONNECTIONS option facilitates a detailed inspection of the connected nodes at any point during the simulation, and provides greater transparency regarding the internal workings of the software.
If you would like to see an example of how these keywords are used in practice, refer to H02 - J-Tube Pull-In.