Ivar KJELBERG
                                                                                                                                                    COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)
                                                         
                            
                         
                                                
    
        Please login with a confirmed email address before reporting spam
     
    
 
                                                Posted:
                            
                                1 decade ago                            
                            
                                2015/04/23 2:53 GMT-4                            
                        
                        
                                                    Hi
my experience with COMSOL "violating" conservation laws are always due to poor model resolution, or approximation in defining the model. Items like mesh resolution, integration step ... nonphysical BC's or numerical singularities can easily introduce large errors. That is why model V&V is so important, and COMSOL, as you are doing it allows to check many variables by different means.
Now for us out here to be able to answer and consider your case in a better way, it would be much easier if you explained your model, or made a simplified version that you shear with us.
For me, when I see that my results from my COMSOL model is "violating" a law, I suspect first of all my model set-up and not the software, as COMSOL has been checked by many ways and by a large community. 
But it happens, even for the best, that some typos enter the formulas under the hood, or that we users set up a case that was not well checked and that COMSOL need some corrections, so it's always worth to chase discrepancies to try to understand them, but that means studying in detail, not just broad statements.
--
Good luck
Ivar                                                
                                                
                            Hi
my experience with COMSOL "violating" conservation laws are always due to poor model resolution, or approximation in defining the model. Items like mesh resolution, integration step ... nonphysical BC's or numerical singularities can easily introduce large errors. That is why model V&V is so important, and COMSOL, as you are doing it allows to check many variables by different means.
Now for us out here to be able to answer and consider your case in a better way, it would be much easier if you explained your model, or made a simplified version that you shear with us.
For me, when I see that my results from my COMSOL model is "violating" a law, I suspect first of all my model set-up and not the software, as COMSOL has been checked by many ways and by a large community. 
But it happens, even for the best, that some typos enter the formulas under the hood, or that we users set up a case that was not well checked and that COMSOL need some corrections, so it's always worth to chase discrepancies to try to understand them, but that means studying in detail, not just broad statements.
--
Good luck
Ivar                        
                                                
                                                                                                            
                                             
                        
                        
                                                
    
        Please login with a confirmed email address before reporting spam
     
    
 
                                                Posted:
                            
                                1 decade ago                            
                            
                                2015/04/24 18:21 GMT-4                            
                        
                        
                                                    Thanks Ivar,
You were right - upon reflection I noticed my boundary conditions were off.   
James                                                
                                                
                            Thanks Ivar,
You were right - upon reflection I noticed my boundary conditions were off.   
James