[kinhsham]

Hi,

I am a student and I am a doing a CFD simulation of a trimaran using STARCCM+ 12.04.011 with the EHP plugin.

So, basically, I am trying to do a resistance estimation of the trimaran and predict the flow pattern around the hull, more specifically under the hull. The trimaran is small, around 7m long. But the thing is the middle hull is only 6 m long and the side hulls is set-backed, i.e. the sterns of the side hulls are 1m behind the stern of the middle hull (so the LOA = 7m).

My 1st question is, in the EHP setup, where should I put my [0,0,0] coordinate?
Should I set it at the stern of the middle hull or the side hulls? Theoretically, the aft perpendicular is at the stern of the middle hull?

Anyway, I tried both, and if I set the 0 coordinates at the stern of the middle hull, I got a lower resistance, about 10% less. Any Idea what cause that?

Okay, the main question. I was using the original setting of EHP, I did not change anything. But the Residuals were high.
Tdr & Tke were around 0.001,
x, y,z momentum were around 0.1, and
continuity & water were around 1.

They separated into three groups, what does it mean?

I tried to play with the base size and it didn't do much.
Originally, I was running with the default base size = 0.15, then I tried others like 1.0, 0.5, 0.3, 0.2, 0.18, 0.12, and 0.10. (all with speed = 2.47 knots or Fn =0.15)
I can see the resistance were converging, with the base size of 1.0, I got 50N, then converged to 36 N...
First of all, is the result reasonable? It seems a bit low to me...
Secondly, whatever the base size was the residuals fell into the range as above, more interestingly, some with coarser meshes(base size = 0.2, 0.18) had less osculation than some finer meshes. Why is that?
So, is the results acceptable? I think it is a little bit unrealistic and the error would be still too high (due to various reasons, we cannot do a real model test, so no experimental data to compare).
What can I do to reduce the residuals? How could I get an trustworthy result?

I saw some posts saying about the time step, I was running with 0.025s. I tried to lower it, it didn't work. The residuals were still high.

One last thing is I tried to run in batch mode with the BatchScript.java, but it didn't work. The simulation ran properly, but at the end it gave an error "ERROR: nulljava.lang.NullPointerException" and won't generate the scenes and plots, and the powerpoint report. Any idea why?

I am sorry that I asked a lot of questions at one time. Any hints and advice would be greatly appreciated.
Thank you so much in advance!

P.S. I was using school computer, but they are kind of old. I can only run 1M cells at max, otherwise it crushes...