saplanti
Mechanical
- Nov 27, 2007
- 780
Dear All,
It is not my nature to put a discussion on that I already have pretty good opinion. However, I must say it is an obligation for me now to get your opinion on the issue:
We are working on a piping system connected on to a 4-stage reciprocating compressor and the flow is pulsating. Pulsation bottles have already been designed for inlet/outlet of each stage. We have CaesarII only as a tool for analysis.
We are asked to design the pipe for 4th harmonic (24.8 Hz) of rotating frequency of the compressor shaft in addition to the static analysis (please do not ask why we do not use 20% margin in accordance with the code guidance; this decision was made in a technical meeting.) in addition to use a clamp with disc springs (with a spring stiffness in one direction only on the pipe) to apply a preload for keeping the pipe held down under the pulsating pipe forces during the thermal expansion/contraction at support locations. The clamp consists of two sideway guide structural members welded onto the supporting structure.
The claim is this: The clamp with spring is going to provide right supporting level that can be considered that pipe is supported in three directions by the preload (the fourth direction is supported by the structural member). Therefore, the structural stiffness instead of spring stiffness shall be used in the preload direction in the natural frequency analysis. Using structural stiffness instead of using clamp spring stiffness is adequate approach to calculate the natural frequency (although the spring stiffness is 15000-20000 N/cm and supporting structural stiffness is 80000-150000 N/cm).
At the moment we do not know the pulsation forces at support locations (we were told not to use the guidance of API 618 or API 688 either because the pulsating forces and stiffness required will be a lot less than the guides give). Therefore we do not have right information to design the clamp with springs properly; the pulsation study will be handled by others after we finalize the static and natural frequency analysis. They will review our design and nominate new support locations if needed.
I could not get an agreement with others on the implementation of clamp with springs although I can simulate the spring action on the pipe to be used in the static and natural frequency analyses within a single model.
They claim that it is required to separate the natural frequency analysis and static (thermal & seismic) analysis; first run the natural frequency analysis using structural stiffness in the 4 support directions (not considering the stiffness of springs on the clamps) by going through sustain load static analysis only and iterate the analysis until reaching the target natural frequency of the piping. (This approach does not answer how to simulate the friction effect in the axial direction in the natural frequency analysis).
Then, go back and use these support locations in the static analysis, allow thermal displacements in the clamp locations in the axial direction of the pipe and lateral directions if necessary by adding large gaps. The clamp spring shall never be used against the thermal or seismic forces. This may require rotation of clamp support. Use the spring force in the applied direction on the pipe to give the pipe friction effect on the supports and nozzles attached. This is sufficient.
However, this approach does not answer what to do in the natural frequency analysis in case we can not locate clamps next to the elbows to eliminate axial (elbow–to-elbow) shaking frequency, to be able to increase the natural frequency to the target frequency without changing the pipe route.
I would like to get your expert opinion on the method above to do the static and natural frequency analyses and implementation of the clamp with spring in the natural frequency analysis.
Thanks in advance.
Kind regards,
Ibrahim Demir
It is not my nature to put a discussion on that I already have pretty good opinion. However, I must say it is an obligation for me now to get your opinion on the issue:
We are working on a piping system connected on to a 4-stage reciprocating compressor and the flow is pulsating. Pulsation bottles have already been designed for inlet/outlet of each stage. We have CaesarII only as a tool for analysis.
We are asked to design the pipe for 4th harmonic (24.8 Hz) of rotating frequency of the compressor shaft in addition to the static analysis (please do not ask why we do not use 20% margin in accordance with the code guidance; this decision was made in a technical meeting.) in addition to use a clamp with disc springs (with a spring stiffness in one direction only on the pipe) to apply a preload for keeping the pipe held down under the pulsating pipe forces during the thermal expansion/contraction at support locations. The clamp consists of two sideway guide structural members welded onto the supporting structure.
The claim is this: The clamp with spring is going to provide right supporting level that can be considered that pipe is supported in three directions by the preload (the fourth direction is supported by the structural member). Therefore, the structural stiffness instead of spring stiffness shall be used in the preload direction in the natural frequency analysis. Using structural stiffness instead of using clamp spring stiffness is adequate approach to calculate the natural frequency (although the spring stiffness is 15000-20000 N/cm and supporting structural stiffness is 80000-150000 N/cm).
At the moment we do not know the pulsation forces at support locations (we were told not to use the guidance of API 618 or API 688 either because the pulsating forces and stiffness required will be a lot less than the guides give). Therefore we do not have right information to design the clamp with springs properly; the pulsation study will be handled by others after we finalize the static and natural frequency analysis. They will review our design and nominate new support locations if needed.
I could not get an agreement with others on the implementation of clamp with springs although I can simulate the spring action on the pipe to be used in the static and natural frequency analyses within a single model.
They claim that it is required to separate the natural frequency analysis and static (thermal & seismic) analysis; first run the natural frequency analysis using structural stiffness in the 4 support directions (not considering the stiffness of springs on the clamps) by going through sustain load static analysis only and iterate the analysis until reaching the target natural frequency of the piping. (This approach does not answer how to simulate the friction effect in the axial direction in the natural frequency analysis).
Then, go back and use these support locations in the static analysis, allow thermal displacements in the clamp locations in the axial direction of the pipe and lateral directions if necessary by adding large gaps. The clamp spring shall never be used against the thermal or seismic forces. This may require rotation of clamp support. Use the spring force in the applied direction on the pipe to give the pipe friction effect on the supports and nozzles attached. This is sufficient.
However, this approach does not answer what to do in the natural frequency analysis in case we can not locate clamps next to the elbows to eliminate axial (elbow–to-elbow) shaking frequency, to be able to increase the natural frequency to the target frequency without changing the pipe route.
I would like to get your expert opinion on the method above to do the static and natural frequency analyses and implementation of the clamp with spring in the natural frequency analysis.
Thanks in advance.
Kind regards,
Ibrahim Demir
