Time History Frequently Asked Questions

Reference [1]: CAEPIPE User's Manual, v5.1J, 2003.

Q-1. What are time functions and how do I use them?

A-1. Typically, a transient fluid flow analysis program computes forces as a function of time at all changes in directions (bends/tees) resulting from a transient event (such as a fluid hammer). These separate force-time histories are then applied separately as Time Varying Loads in CAEPIPE at the corresponding nodes in the piping model.

Time functions are non-dimensional (i.e., a series of time versus value tables). which describe the variation of the forcing function with respect to time. The actual force (or moment) at any time is found by linear interpolation between time points and multiplying by the input scale factor.

In CAEPIPE, the time function you define can have any interval between two time values. You can make that table as fine as you want it to be. Example: 0.0 0.0 (you must have a zero entry for the Value, otherwise CAEPIPE will interpolate linearly for a smaller time, say t=0.1, if this time is smaller than that available in the table (say you had started from 0.2 12000).


Time history analysis begins at time t=0.0. An Illustrative time function (format: time value) follows.


Timefunction name
0.2 0.0
0.3 1000
0.4 4000
0.5 10000
0.6 35000
0.7 73000
0.8 60000
0.9 xxxx (some value)
1.0 xxxx
2.0 xxxx
3.0 xxxx
4.0 xxxx

The time function above is illustrative and not a guideline. The above format can be followed to create indvidual text files that contain one time function, and can be read into CAEPIPE.


Time functions can be renamed inside CAEPIPE. You can delete a time function (use Ctrl+X). It is possible to fill in up to 1000 rows of time-value pairs.

Q-2. Explanation of terms in the Time History Analysis control dialog (found in the Layout window > Loads menu > Time History)


1. Time step: is the smallest time gap for which you want to compute system response. (typically no more than 10% of the period of the highest natural frequency, i.e., higher the frequency, smaller the time step, e.g.. for a max. 33Hz., the time step would be 0.03).


2. Number of time steps: is how long you want to study system response. The (time step times number of time steps) may exceed the range of available data in the time function. The time function is only the forcing function. But, CAEPIPE can compute system response to that forcing function well after the forcing function ceases. For example, the effect of a really heavy steam hammer could linger on for a minute, and your forcing function data could span only 8 sec..


3. Output interval: Is the periodic time step you want CAEPIPE to report results. For example: if you wanted to see results for every 5th time step (0.5s, 1s, 1.5s, 2s and so on), enter 5. You need to check the "Save Results" checkbox (or press F2 in the v4.1 and earlier). The file name that contains all these results is model.RTH.


4. Damping: Is the damping factor expressed as a percentage (not as a fraction). Enter 5 not 0.05 for 5%.

Time step =0.1s,
# of time steps: 200,
Output interval: 5,
Damping: 5,
Save Results checkbox Checked.


CAEPIPE calculates the solutions (modal displ. only) at every 0.1 sec from 0.1 to 20 seconds. CAEPIPE scans this file (.RTH) and finds the maximums and reports these results for forces/moments/displacements etc. under Support loads, Pipe element forces, Displacements for Time history load case. But if you choose Time history under Results menu (Output), then the whole RTH file is displayed (of course, as Displ, Support loads).


At a particular node, one can input the FX, FY, FZ as separate Time Varying loads and CAEPIPE will calculate the resultant.

Q-3. Are time functions momentum forces or pressures or equivalent forces of deflections?

A-3. The time functions that CAEPIPE requires can be any time-varying quantity. These values may be forces/moments in the case of a fluid hammer problem. CAEPIPE takes these values at each timestep (linearly interpolates if necessary) and multiplies it by the user-input scale factor and applies it in the user specified direction (FX, MY etc.).


The values in the case of a fluid hammer are usually got from a transient flow analysis software which outputs the results of a fluid hammer in terms of forces on a system as a function of time. In the transient analysis program, one may be able to get different forces (as a function of time) for different piping components. By noting the direction of these forces, a CAEPIPE user may apply those forces with a scale factor if necessary, at the corresponding piping components in the noted directions. CAEPIPE then will apply those forces as explained in the previous paragraph and will compute the response of the piping system by doing a time-history analysis.


Any phenomenon that gives rise to forces/moments that vary with time can be input into CAEPIPE for time-history analysis. You may hand calculate or use a simulation program to get the force/moment vs. time history values.

Q-4. Regarding the time-history of loads, they would be applied in the plane of changes in direction (at elbows) at a 45deg angle to the longitudinal axes of the pipe runs (assuming 90deg elbows). Technically they would be different at each elbow although a generalized time history might be possible to develop. Can CAEPIPE handle this analysis?

A-4. For time history loads (tabulated as separate X, Y and Z component values versus time), one can apply each component to one global axis, as a separate time function. Example: X component (of Force) vs. time at an elbow in the global X axis, Y component (of Force) vs. time at an elbow in the global Y axis, and so forth.

Q-5. In a time history analysis, why don't the pipe forces for each element add up to the load shown for a support?

A-5. The idea that the pipe forces for each element should add up to give you, say, a support load at a limit stop would be correct in a time independent analysis (W, P, T1, etc.).


But, in a time history analysis, the values you see for any entity (displacements, support loads, pipe forces, moments etc.) are the enveloped values (i.e., the absolute maximum value of [for that node] the entity when compared with values at all time steps). So, there is no telling whether the numbers are going to add up.

Q-6. Is missing mass correction feature included in Time History and Harmonic Analyses?

A-6. No. Time history and Harmonic analyses do not include the missing mass correction presently. It is included in the Response spectrum analysis, however.

Q-7. The release notes for version 4.0F mentions an interface to fluid flow program called Pipenet. Where is it located in the interface within CAEPIPE? Or is it a separate program?

A-7. Starting v4.0F, CAEPIPE does interface with Pipenet Transient module. Two way translators exist.

1. Pipenet Force-Time History data file to CAEPIPE (built into CAEPIPE v4.0F-v4.1))
2. CAEPIPE to PIPENET Geometry translator (available as a separate program)-dated 2/9/98


However, v5.0 does not support the translation as given above for item 1. If there are enough customers using Pipenet, then we may bring it back.