Whenever one computes joint or element code checks or fatigue, or beam loads the last values computed will be stored for use with pictures. Storing only a single value is in keeping with the philosophy of checking all cases and using the maximum to size the structure.

With animations, however, we have a different problem. Here we may want to look at the deflections of the structure or at code checks as a function of time. Before you can do this, you must have load cases that are snapshots at times during a time domain simulation. If so, then you must tell MOSES to store the results so that you can use them for pictures. One does this with the option

     -FOR_ANIMATION

on the Post-Processing command: commands:

     BEAM_POST CODE_CHECK
     JOINT_POST CODE_CHECK
     JOINT_POST DISPLACEMENTS

If you use this option, then

• no output of the results will be generated.
• Results will be generated only for the current process. If you have more than one process, then you need to generate the data for each process; i.e. use &DESCRIBE PROCESS to change the current process.
• Once the data has been generated, it will be used in place of the "system wide" data for all events greater than 0.
• If you define a combination case, then it will inherit the time association of the first member of the combination.
• If you have a combination case generated with the -TIME option, then it will have the event behavior of the time specified. I will not, however, be useful unless you use the FR_2TIME to associate a process with these events.

Perhaps it is good to look at some examples. Suppose that you have a time domain simulation and that you generated load cases with:

     &loop t 1 1000
     lcase -process %t &string(o_number T0000)
     &endloop

Now, suppose that you wanted to look at the deflections as a function of time. You can do this with

     strpost
     joint deflection -load t@
     end
     &picture iso -events 0 1000 -deflect 200 only -movie

Of course, once the deflections are associated with the events the making of the picture is the same as with any animation.

Now, suppose you have a combination case (suppose you are doing an LRFD code check, etc.) Then

     &loop t 1 1000
     &set num = &string(o_number 0000)
     cases -combine %(c)%num  %(t) 1 some -1
     &endloop

will have a time association while will

     &loop t 1 1000
     &set num = &string(o_number 0000)
     cases -combine %(c)%num  some -1 %(t) 1
     &endloop

will not.