Posted Fri, 28 Jun 2013 00:15:54 GMT by Payam Aghsaee (Now at IIHR)
I have just started using MIKE3 and I am trying to model a simple open channel flow. The channel section is trapeziodal with maximum depth of 1m and a constant slope of 0.01. The channel length is 20 m (0.2 m elevation difference between to open boundaries at each end).
I took the datum at the bank of upstream section so all bed elevations are negative. I would like to model a uniform flow with given depths at both ends. For initial surface eleveations I choose a constant depth of 0.4 meters so all serface levation values are negative.
I set the boundaries as constant surface elevation of -.6 for upstream and -.8 for the downstream section. After running the model blows up and says: water level -0.6!
Any suggestions?
Appreciate your help
Posted Thu, 04 Jul 2013 12:57:20 GMT by jal@dhigroup.com
A tip is to have a look in the model run data file which tracks the modelling performance, it usually gives a good idea about where the problem lies. It automatically saves a text file in the same location as your MIKE3 setup file.

But, as for  your problem; The initial WL is above the downstream boundary condition I believe? Just try to set the initial value as a constant surface elevation instead, which correspoonds to your downstream BC. Also, if you use a water depth, you get a water column up on your channel banks as well when you use a trapezoidal channel, which might cause instabillity. I would try that first (as your model is very small it will fill up right away anyhow).

Best regards,
Jonas

You must be signed in to post in this forum.