Please login with a confirmed email address before reporting spam
     
    
 
                                                Posted:
                            
                                5 years ago                            
                            
                                3 juin 2020, 14:53 UTC−4                            
                        
                        
                                                    Hi Ming,
The variable Hxy should indeed give the derivative of Hx with respect to y. Thus, the current density in the z direction should be Jz=Hyx-Hxy as you suggested. However, I am not sure why you are getting a non-zero value for the expression mentioned above. What kind of value are you getting for the expression? Also, are you using a moving mesh?
Cheers,
Alex
                                                 
                                                
                            Hi Ming,
The variable Hxy should indeed give the derivative of Hx with respect to y. Thus, the current density in the z direction should be Jz=Hyx-Hxy as you suggested. However, I am not sure why you are getting a non-zero value for the expression mentioned above. What kind of value are you getting for the expression? Also, are you using a moving mesh? 
Cheers,
Alex                        
                                                
                                                                                                            
                                             
                                            
                                                
    
        Please login with a confirmed email address before reporting spam
     
    
 
                                                Posted:
                            
                                5 years ago                            
                            
                                3 juin 2020, 21:18 UTC−4                            
                        
                        
                                                    Dear Alex,
Thank you for your suggestion.
Firetly, I just use a triangular mesh, it's not a moving mesh.
And, I use the mfh part. The expression: mfh.curlHz-mfh.Jz is zero, but the value of mfh.Jz-(Hyx-Hxy) is from  -2.56e-13 to 2.34e-13.
Yours,
Ming
                                                 
                                                
                            Dear Alex,
Thank you for your suggestion. 
Firetly, I just use a triangular mesh, it's not a moving mesh.
And, I use the mfh part. The expression: mfh.curlHz-mfh.Jz is zero, but the value of mfh.Jz-(Hyx-Hxy) is from  -2.56e-13 to 2.34e-13.  
Yours,
Ming                        
                                                
                                                                                                            
                                             
                                            
                                                
    
        Please login with a confirmed email address before reporting spam
     
    
 
                                                Posted:
                            
                                5 years ago                            
                            
                                4 juin 2020, 09:25 UTC−4                            
                        
                        
                                                    Hi Ming,
This is probably due to numerical error rather than the definition of Hyx. Taking the derivative of the H field will lead to numerical error, and since the H field is usually on the order of 1e6 A/m, the error of 1e-13 seems reasonable. The reason why mfh.curlHz-mfh.Jz gives zero is that mfh.Jz is defined as mfh.curlHz, so Comsol recognizes that they are exactly equal.
Cheers,
Alex
                                                 
                                                
                            Hi Ming,
This is probably due to numerical error rather than the definition of Hyx. Taking the derivative of the H field will lead to numerical error, and since the H field is usually on the order of 1e6 A/m, the error of 1e-13 seems reasonable. The reason why mfh.curlHz-mfh.Jz gives zero is that mfh.Jz is defined as mfh.curlHz, so Comsol recognizes that they are exactly equal.
Cheers,
Alex                        
                                                
                                                                                                            
                                             
                        
                        
                                                
    
        Please login with a confirmed email address before reporting spam
     
    
 
                                                Posted:
                            
                                5 years ago                            
                            
                                5 juin 2020, 01:03 UTC−4                            
                        
                        
                                                    Dear Alex
Thanks for your suggestion. This helps me a lot. I was confused for days, now I can have a good sleep. Haha
Yours,
Ming
                                                 
                                                
                            Dear Alex
Thanks for your suggestion. This helps me a lot. I was confused for days, now I can have a good sleep. Haha
Yours, 
Ming