Re: [Gfs-users] GfsRiver and friction

[<Wy...@ba...>]

Hi George,
 
Does the problem change if you prescribe the waterlevel downstream and not its gradient ?
 
cheers Jens

----------------------------------------------
Dr.-Ing. Jens Wyrwa
Bundesanstalt für Gewässerkunde (BfG)
Referat Ökologische Wirkungszusammenhänge (U2)
Am Mainzer Tor 1
56068 Koblenz
Tel:      +49(0)261 1306-5254
http://www.bafg.de <http://www.bafg.de/> 
---------------------------------------------- 

 


________________________________

	Von: George Gerber [mailto:geo...@gm...] 
	Gesendet: Samstag, 3. Dezember 2011 11:31
	An: Gfs...@li...
	Betreff: [Gfs-users] GfsRiver and friction
	
	
	Dear all,
	
	Would somebody be so kind as to explain to me why the simple model below does not produce the correct water levels?
	The model below describes a rectangular channel with a 0.1% slope, 1000m length, 3.90625m width and Manning n of 0.01. An inflow of 1m3/s is specified upstream and the waterlevel gradient downstream is set equal to the channel slope.  The flow depth should be 0.231m.
	
	Any help will be greatly appreciated.
	George
	~~~~~~~~~~~~~~~~~~~~~
	Define LEVEL 8
	Define l 1000.0
	Define b (l/pow(2.,LEVEL))
	Define S0 0.001
	Define n 0.01
	Define Q 1.0
	Define Pinit 1e-2
	
	Define A (P*b)
	Define PWet (b+2*P)
	Define R (A/PWet)
	Define grav 9.81
	
	1 0 GfsRiver GfsBox GfsGEdge {x=0.5} {
	    PhysicalParams { L = l g = grav }
	    Init {} { Zb=S0*(l-x) P=Pinit }
	
	    SourceCoriolis 0 {
	       if (P < 1e-6 ) return 1e10;                  /* Avoid division by zero */
	       else return grav*n*n*Velocity/pow(R,4./3);   /* Manning friction model */
	    }
	
	    AdvectionParams { gradient = gfs_center_minmod_gradient }
	
	    Refine LEVEL
	    InitMask (y>l/2-b? 0:1)
	
	    Time { end=50000 }
	    OutputProgress { istep = 10 } stderr
	    OutputLocation { istep = 100 } GaugeMid 500 499.9 0
	}
	GfsBox {
	    left  = Boundary {BcDirichlet U Q/b}
	    right = Boundary {BcNeumann P -S0}
	}