Tension-only members are truss/cable members that are capable of carrying tensile forces only. Thus, they are automatically inactivated for load cases that create compression in them.

Linear Tension/ Compression Analysis  

Compression-only members are truss members that are capable of carrying compressive forces only. Thus, they are automatically inactivated for load cases that create tension in them. Member Releases are not allowed on members with this attribute.

The procedure for analysis of Tension-only or Compression-only members requires iterations for every load case and therefore may be quite involved. you may also consider using the INACTIVE specification (instead of Tension/Compression) if the solution time becomes unacceptably high.

If a CHANGE command is used (because of a change in the list of tension members, cable tension, supports, etc.), then the SET NL command must be used to convey to STAAD that multiple analyses and multiple structural conditions are involved.

1. Loads that have been defined on members declared as MEMBER TENSION or MEMBER COMPRESSION will be active even when the member becomes INACTIVE during the process of analysis. This applies to SELFWEIGHT, MEMBER LOADS, PRESTRESS & POSTSTRESS LOADS, TEMPERATURE LOAD, etc.

2. A member declared as a TENSION only member or a COMPRESSION only member will carry axial forces only. It will not carry moments or shear forces. In other words, it is a truss member.  

3. Do not use Load Combination to combine these cases. Tension/Compression cases are Nonlinear and should not be linearly combined as in Load Combination. Use a primary load case with the Repeat Load command.  

Member Tension 0

This command switches off all tension/compression only specifications for load cases which are specified subsequent to this command, usually entered after a CHANGE command. There is no list associated with this command. Hence, for any further primary load cases, the tension/compression only attributed is disabled for all members.

Notes

1. See TR.5 Set Command Specification for explanation of the SET NL command. The number that follows this command is an upper bound on the total number of primary load cases in the file.

2. STAAD performs up to 10 iterations automatically, stopping if converged. If not converged, a warning message will be in the output. Enter a SET ITERLIM i command (i > 10) before the first load case to increase the default number of iterations. Since convergence may not be possible using this procedure, do not set the limit too high.   

3. The principle used in the analysis is the following.

   • The program reads the list of members declared as MEMBER TENSION and/or COMPRESSION.

   • The analysis is performed for the entire structure and the member forces are computed.

   • For the members declared as MEMBER TENSION / COMPRESSION, the program checks the axial force to determine whether it is tensile or compressive. If the member cannot take that load, the member is "switched off" from the structure. 

   • The analysis is performed again without the switched off members.

   • Up to 10 iterations of the above steps are made for each load case, unless a higher value is set using the command ITERLIM.

   • This method does not always converge and may become unstable. Check the output for instability messages. Do not use results if the last iteration was unstable.

4. A revised MEMBER TENSION / COMPRESSION command and its accompanying list of members may be provided after a CHANGE command. If entered, the new MEMBER TENSION/COMPRESSION commands replace all prior such commands. If these commands are not entered after a CHANGE, then the previous commands will still be applicable.  

5. The MEMBER TENSION command should not be used if the following load cases are present : Response Spectrum load case, Time History Load case, or Moving Load case. If used, the MEMBER TENSION /COMPRESSION will be ignored in all load cases. 

6. If UBC Load cases are included, then follow each UBC load case with an Analysis command, then a Change command.

7. If the model requires a PDelta analysis, then only the first method as described in TR.37.2 P-Delta Analysis Options (i.e., small or large delta) may be used.