On behalf of Jordan aka "Rogertheupsguy":
I have a question about using the REPEAT LOAD command and foundation supports.
In the past whenever I have used a foundation support I have always used LOAD COMB. However, reading through this thread I have discovered the REPEAT LOAD command. Referencing the Technical Manual I discover that the REPEAT LOAD should be used if the model involves non-linear situations.
Fearing that I was performing incorrect analysis, I wanted to compare. Now I took the model that was posted in this thread and added the LOAD COMB to the compare the results and I noticed no change in the plate forces or bearing between the LOAD COMB and the REPEAT LOAD.
I am curious why the results are the same for this situation and what situation will the result differ for a foundation support?
Jordan
REPEAT LOAD is an instruction to the program to analyze a set of loading conditions as though they act concurrently. It "always" provides a technically correct analysis of a structure, because it considers all of the specificed forces acting together, as they do in the real structure. For this reason, it will provide technically correct results in situations that incorporate any non-linearities such as tension-only or compression-only members, compression-only foundations, etc. It is also ESSENTIAL to combine load effects with the REPEAT LOAD specification if a PDELTA analysis is being performed, or else the analysis won't "see" the P at the same time that it predicts the Delta. So it will be a meaningless analysis.
By comparison, LOAD COMB does not instruct the program to analyze the model as if the loads acted concurrently. Instead, it is just an instruction to combine (in a variety of possible ways) the results that were obtained from the independent analysis of the constituent load cases. It is important to note that there are instances where LOAD COMB is a perfectly legitimate method of obtaining combined results. If the model has no non-linearities, and if an elastic (stiffness) analysis is being performed, and if superposition is valid, then LOAD COMB provides an analytically efficient way to obtain results without performing any unnecessary analysis.
Having said all of that, what is the downside to REPEAT LOADS and why not use them all the time. Well, the downside is that they require some additional analysis efforts and compared to LOAD COMB, so run times may be somewhat longer. But depending on the complexity of your typical models, that may be a negligible effect. Personally, I promote the idea of using REPEAT LOADS as the general rule, and considering LOAD COMB as the exeption to the rule...to be taken advantage of only when technically permissible.
Cheers,
Chris
This model will be a good demonstration of a LOAD COMB gone wrong:
STAAD PLANE START JOB INFORMATION ENGINEER DATE 23-Nov-09 END JOB INFORMATION INPUT WIDTH 79 UNIT FEET KIP JOINT COORDINATES 1 0 0 0; 2 0 10 0; 3 10 10 0; 4 10 0 0; MEMBER INCIDENCES 1 1 2; 2 2 3; 3 3 4; DEFINE MATERIAL START ISOTROPIC STEEL E 4.176e+006 POISSON 0.3 DENSITY 0.489024 ALPHA 6e-006 DAMP 0.03 END DEFINE MATERIAL MEMBER PROPERTY AMERICAN 1 TO 3 TABLE ST W8X28 CONSTANTS MATERIAL STEEL ALL SUPPORTS 1 4 FIXED BUT MX MY MZ KFY 10000 SPRING COMPRESSION 1 4 KFY LOAD 1 LOADTYPE None TITLE GRAVITY LOAD JOINT LOAD 2 3 FY -20 LOAD 2 LOADTYPE None TITLE LATERAL LOAD JOINT LOAD 2 FX 10 LOAD 3 LOADTYPE None TITLE GRAV + LAT REPEAT LOAD 1 0.67 2 1.0 LOAD COMB 4 GRAVITY + LATERAL 1 0.67 2 1.0 PERFORM ANALYSIS LOAD LIST 3 4 PRINT SUPPORT REACTION LIST 1 4 FINISH
This model is a single portal frame with pinned bases and a compression-only spec on the supports. It contains a downward-acting gravity load, a laterally-acting load that tends to produce uplift on the left support, a REPEAT LOAD spec that correctly reports the reactions at both supports, and a LOAD COMB spec that gets the left support reaction wrong.
Run the model and view the support reactions in the output file to see the difference between the REPEAT LOAD and the LOAD COMB.
Don't forget, too, that your foundation support reactions will also be incorrect if you use LOAD COMB in conjunction with a PDELTA ANALYSIS.
Regards,