Gentlemen
I am modeling an pipe system for an underground gold mine. The system consists of a chilled water pipe (5 celsius) and a warm water return (35 celsius) running parallel to each other. These pipes are u-bolted every 3m to a structural beam, which is in turn "hung" from the "roof" of the mine by chains on either side.
Some sizes to gain perspective:
Pipe diameter: 500mm NB
Total pipe length: 3709m
Chain length: 800mm
Number of elements: 4900 (approx)
Rod Increment (degrees): tried all
Rod Tolerance (degrees): tried all
Caesar version: latest
The system is shown in the schematics below. The model is attached, and the config file is uploaded as "caesar.cfg.txt".
My questions are as follows:
Q1. Why will the model not converge? (+Y supports have been investigated, but too many assumptions need to be made...)
Q2. The first schematic shows the anchors as red dots with points in the model labelled 1 to 4. If I split the system into 3 parts (ie: 1-2, 2-3 and 3-4), then part 1-2 converges, but not the other two parts. Part 1-2 is more complex than the other two parts - so why won't they converge?
Q3. This model takes approximately 30 minutes to run (Intel Core2 Duo @ 2.2GHz, 4GB ram, Win7-64). Why is it taking so long?
Q4. Caesar allows 512MB of ram to be allocated to it. Whilst this model is running, I have some 2.3GB of ram free. Can more ram not be allocated to speeding up the run?
Q5. The C-Noded Lim, Gui, Y restraints disappear from the input and output display. Is this memory related (tested on both ATI Mobility Radeon HD 2400 XT and Nvidia Ge-Force 9600M GT, in all Caesar II graphic driver modes)?
Many thanks for your attention.
(zoom in to schematics if not clear)