equilibrium calculation appears dependent on order of constituents

[GoogleCodeExporter]

What steps will reproduce the problem?
1. run attached python script with attached files

What is the expected output? What do you see instead?
The following should print to the screen:
"DEBUG:root:Works worked!
ERROR:root:Fails worked!"
Instead I see:
"DEBUG:root:Works worked!
ERROR:root:Fails failed!"

What version of the product are you using? On what operating system?
v2.1.0 on Windows 7, 32-bit

Please provide any additional information below.
The only difference between these two phase descriptions is the order in which 
the constituents are declared.

Original issue reported on code.google.com by kevin.ho...@gmail.com on 12 Mar 2014 at 10:22

Attachments:

• eqErrorSample.zip

[GoogleCodeExporter]

Confirmed in trunk r2800 as well. 

If the 'gibbs' or 'vcs' solvers are explicitly specified, the 'fails' example 
can be solved successfully, so the basic issue is with the 'element_potential' 
solver. However, using the automatic solver selection also fails, so there 
seems to be a problem with the mechanism for falling back to the Gibbs solver 
as well.

Original comment by yarmond on 13 Mar 2014 at 2:58

• Changed state: Accepted
• Added labels: component-Equilibrium