There is no issue with the solver just that it is diverging. Just set you reference pressure on the higher side to your reality conditions(>20 MPa) and set opening relative pressures to zero.

Also have you checked the direction of rotation of your gears????? Why dont you try starting your problem under steady state conditions and then move on to your transient conditions.

others
 

It‘s the issue of mesh,for mesh quality of the domain is below 0.3.


Thanks all !

But this simulation is not basically consistant with the experiment.
In the first picture,this groove doman is connected with fluid domain with the method of "Domain Interface".

Hello，Still a question about my simulation? I suppose the unreal pressure picture above is due to leakage. So is there anyone who has familiar experience with gear pump? Since the massFlow at outlet is not consistant with the actual situation, what's the reason ?And what can be done?
Many thanks.