Posted by: Kenny Robertson
limits of a system - 10/02/03 04:41 AM
I have a system that is being installed on site and requires allowable nozzle loads to be met at it's limits. It is a orifice metering system with 3 16" streams that run horizontally 1m above the ground and then rise up into a common raised 24" header some 4m above the ground and perpendicular to the meter runs. The header is not directly supported but rather there are supports below on the 'tails' leading up to it.
The problem is that when we analyse the system on it's own with moments and forces acting on the 'nozzles' to simulate the required allowables, the lack of constraint imposed by having no connecting pipework causes the system to twist freely and behave un-realistically under operating conditions.
By the same token adding some imaginary pipework (running across a minimum of 3 supports to an anchor) helps constrain the actions of the connecting nozzles, but may not fully develop the required forces and moments that the allowables are suggesting. Adding extra forces and moments at the nozzle seems an option, but I am concerned that a portion of these will be absorbed by the imaginary piping we have modelled in and thus may not be testing the ability of our nozzles to absorb the allowables required.
Does anybody have any suggestions on how to get a realistic fix on this problem? Apart from the basic sustained/displacement checks we also need to ensure that the loads do not cause flange overload at other points in the system and is also safe seismically. Or is this an indeterminate system that can only be analysed as part of the bigger system it is connected to.
All help and comments gratefully recieved!!!
The problem is that when we analyse the system on it's own with moments and forces acting on the 'nozzles' to simulate the required allowables, the lack of constraint imposed by having no connecting pipework causes the system to twist freely and behave un-realistically under operating conditions.
By the same token adding some imaginary pipework (running across a minimum of 3 supports to an anchor) helps constrain the actions of the connecting nozzles, but may not fully develop the required forces and moments that the allowables are suggesting. Adding extra forces and moments at the nozzle seems an option, but I am concerned that a portion of these will be absorbed by the imaginary piping we have modelled in and thus may not be testing the ability of our nozzles to absorb the allowables required.
Does anybody have any suggestions on how to get a realistic fix on this problem? Apart from the basic sustained/displacement checks we also need to ensure that the loads do not cause flange overload at other points in the system and is also safe seismically. Or is this an indeterminate system that can only be analysed as part of the bigger system it is connected to.
All help and comments gratefully recieved!!!