Aust Workshops - TUFLOW Modelling Bends, Structures and ...

Australian 2011/2012 TUFLOW Workshops 

1D and 2D Modelling 

Bends, Structures and Obstructions 

Bill Syme 

1D vs 2D 

Understanding the Difference 

Form Losses 

• Energy dissipated as heat due to changes in velocity 

magnitude and direction 

• Pronounced at 

• Bends 

• Flow constrictions (structures) 

• Form loss coefficient 

• Proportion of dynamic head (V 2 /2g) lost 

• V = 1m/s; Dynamic Head = 0.05m 

• V = 4m/s; Dynamic Head = 0.82m 

3 

Bill Syme, BMT WBM, support@tuflow.com 

1


Right-Angled Bend 

1D vs 2D 

4 

 

3 

3 

 

1 

 

 

 

1 2 2 

A 

0.5 

Water Surface Profiles (V = 2m/s) 

0.4 

Coarse 2D Model 

1D Model 

1D Model 

Water Level (m) 

0.3 

0.2 

0.1 

A 

0 

-0.1 

0 50 100 150 200 250 300 350 400 450 

Distance (m) 

4 

River 

Bends 

23.2 

23.4 

23.2 23.2 23.2 23.2 23.2 23.2 23.2 23.2 23.2 

23.4 23.4 23.4 23.4 23.4 23.4 23.4 23.4 23.4 

23.6 23.6 23.6 23.6 23.6 23.6 23.6 23.6 23.6 

23.4 23.4 23.4 23.4 23.4 23.4 23.4 23.4 

23.2 23.2 23.2 23.2 23.2 23.2 23.2 23.2 23.2 

3 

• 4 m/s 

• 20 m deep 

23.6 23.6 

23.6 

• 0.4m 

superelevation 

23.8 23.8 23.8 23.8 23.8 23.8 23.8 23.8 23.8 

24 24 24 24 24 24 24 24 24 

• 1D: 

• Need additional 

losses 

(eg. higher n) 

• No superelevation 

24.4 24.4 24.4 24.4 24.4 24.4 24.4 24.4 24.4 

24.2 

23.8 23.8 23.8 23.8 23.8 23.8 23.8 23.8 23.8 

24 

23.8 

23.6 23.6 23.6 23.6 23.6 23.6 23.6 23.6 23.6 

23.4 23.4 23.4 23.4 23.4 23.4 23.4 23.4 23.4 

23 23 23 23 23 23 23 23 23 

23.2 23.2 23.2 23.2 23.2 23.2 23.2 23.2 23.2 

22.8 22.8 22.8 22.8 22.8 22.8 22.8 22.8 22.8 

22.6 22.6 22.6 22.6 22.6 22.6 22.6 22.6 22.6 

22 

4 

5 

2D vs 3D? 

2D 

3D 

23 cm 30 cm 

6 

6 


2


Bends - Conclusions 

1D and 2D Approaches 

• 1D 

• 2D 

• Apply extra losses by 

• Form loss coefficient, or 

• Increasing Manning’s n 

• Do not model superelevation 

• Form losses inherent / Models superelevation 

• However 

• Are model elements too coarse to simulate all losses? 

• Are there losses in the vertical plane? (Helicoidal circulations) 

• Additional form losses may be required 

7 

Hydraulic Structures 

• Hydraulic Structures 

• Bridges and Embankments 

• Large Culverts 

• Hydraulics is Complex (3D) 

• 1D: Traditional Approach 

• 2D: Looks impressive, but is it accurate? 

• 1D/2D: Best of both? 

8 

2D: Looks impressive, but is it accurate? 

? 

9 


3


1D: Traditional Approach 

Uses Contraction/Expansion Losses 

10 

1D Culvert Entrance and 

Exit Loss Coefficients 

• Coefficients adjusted according to approach and departure velocities in a 1D 

network (n/a yet when connected to 2D) 

• Can fix losses (ie. no adjustment) if desired 

• Default unadjusted values typically 0.5 and 1.0 

• Energy loss is C*V s2 /2g 

V 

Centrance 

_ adjusted = Centrance 

 

1 

V 

V 

_ 1 

departure 

Cexit 

adjusted = Cexit 

 

Vstructure 

 

approach 

structure 

2 

 

 

 

11 

2D: No Contraction/Expansion Losses? 

12 


4


2D Scheme Modifications 

(2d_fc or 2d_fcsh layers) 

Cell Obvert 

Partially block cell sides 

Deck FLC 

Form Loss 

Coefficient 

13 

2D Layered Adjustments 

(2d_lfcsh layers) 

Blockage = 0% 

FLC = 0 

Blockage = 50% 

FLC = 0.5 


FLC = 0.8 

Blockage = 5% 

Form Loss Coeff = 0.1 

14 

2.88 m 1 

1 

2 2 3 

3 

2.30 m 

4 

0.7 m/s 

2.86 m 

2.42 m 

2.9 m/s 

0.8 m/s 

3 

Water Surface Profiles - Outlet Controlled 

2.9 

2.8 

1D Model 


2.7 

2.6 

2.5 

Upstream 

of Culvert 

Downstream 

of Culvert 

2.4 

2.3 

2.2 

0 50 100 150 200 250 300 

Distance (m) 

15 


5


 

Modify 2D Cells 

to represent 

culvert 

Water Surface Profiles - Outlet Controlled 

3 

2.9 

1D Model 

2.8 

2D Model (Culvert as 2D Cells) 


2.7 

2.6 

2.5 

Upstream 

of Culvert 

Downstream 

of Culvert 

2.4 

2.3 

2.2 

0 50 100 150 200 250 300 

Distance (m) 

16 

So 2D isn’t perfect! 

What are our options? 

• Don’t use 2D! 

• Adapt 2D Solution 

• Insert 1D Solution 

17 

“Calibrating” 

2D Structures 

• For example, 

if we apply a 

0.2 FLC, 

ie. add 

0.2*V 2 /2g 

energy loss 

 


3 

2.9 

2.8 

2.7 

2.6 

2.5 

2.4 

2.3 

2.2 

Water Surface Water Profiles Surface - Outlet Profiles Controlled - Outlet - Controlled Adjusted Form Losses 

Upstream 

of Culvert 

Downstream 

of Culvert 

0 50 100 150 200 250 300 

Distance (m) 

1D 1D Model 

2D 2D Model (Culvert as as 2D 2D Cells) 


6


1D/2D 

Link 

Options 

2.88 m 1 

1 

0.7 m/s 

2.86 m 

2.42 Modify m 

Block 2D Cells 

2 2 3 to and represent insert 

3 

2.9 m/s 

0.8 m/s 

culvert 1D element 

2.30 m 

4 

• SX Link 

• HX Link 

(Preserves 

momentum) 


(m) 

3 

2.9 

2.9 

2.8 

2.8 

2.7 

2.7 

2.6 

2.6 

2.5 

2.5 

2.4 

2.4 

2.3 

2.3 

Water Surface Profiles - - Outlet Controlled 

1D 1D 

Model 

Model 

2D 2D Model Model (Culvert 

1D Model as as 2D 2D Cells) 

2D Model (Culvert as 1D Element) 

2D 2D Model Model (Culvert (Culvert as 1D as Element) 2D Cells) 

2D (1D Culvert & Momentum) 

Upstream 

Downstream 

Upstream Downstream 

of Culvert 

of of Culvert 

of Culvert of Culvert 

2.2 

2.2 

0 50 100 150 200 250 300 

0 50 100 150 200 250 300 

Distance (m) 

Distance (m) 

19 

“Calibrating” 

1D Culvert 

linked to 2D 

 

 

 

• Culvert as 1D 

Element 

• Reduce Outlet 

Loss 

Coefficient by 

0.2 


Water Surface Profiles - Outlet Controlled - Adjusted Form Losses 

3 

1D Model 

2.9 

1D Model 

2D Model (Culvert 1D Model as 2D Cells) 

2.8 

2D Model (Culvert as 2D Cells) 

2D Model (Culvert as 1D Element) 

2.7 

2.6 

Upstream 

Downstream 

of Culvert 

of Culvert 

2.5 

2.4 

2.3 

2.2 

0 50 100 150 200 250 300 

Distance (m) 

20 

Modelling Culverts - Conclusions 

• Culvert as 2D Cell(s) 

• 2D solution models 70 to 80% of losses 

• Need 20 to 30% additional form losses 

• Culvert as 1D Element 

• Over predicts losses by 0 to 70% 

• Small – 0% over prediction 

• Large – up to 70% over prediction 

• Reduce inlet / outlet losses of 1D element(s) 

21 21 


7


Embankments / Levees 

(Weir Flow) 

• Approach 

• Test submergence 

across cell side 

• BC Weir equation if 

unsubmerged 

• No adjustment if 

submerged 

 

2.4 

2.2 

2 

Cell sides set to 1m high 

for 2D Model (Thin Weir) 

Comparison Upstream Water Level Hydrographs 

Unsubmerged 

Submerged 

400 

350 

300 

• Thin Weir Test 


1.8 

1.6 

1.4 

1D 

1D 

Model 

Model 

& Theory 

Theory 

D/S Water Level 

2D Model (Thin Weir) 



Inflow Hydrograph 

Inflow 

Inflow 

Hydrograph 

Hydrograph 

250 

200 

150 

Inflow (m 3 /s) 

1.2 

100 

1 

Weir Crest Level 

50 

0.8 

0 

0:00 0:30 1:00 1:30 2:00 2:30 3:00 3:30 

Time (hh:mm) 

22 

Oblique Weirs 

• Flow oblique to 

grid 

• Weir at 45˚ test Unit Flow Across a Broad-Crested Weir - Upstream Controlled 

3.50 

• Correct using weir 

coefficient 

Unit Flow (m 2 /s) 

3.00 

2.50 

2.00 

1.50 

1.00 

BC Weir Formula 

BC Weir Formula 

2D Model BC (Weir Formula Coef = 1.0) 

2D Model (Weir Coef 1.0) 

2D Model (Weir Coef = 1.1) 

0.50 

0.00 

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 

Upstream Head above Sill (m) 

23 

Conclusions 

• 2D contracts and expands flow lines 

• Inherently models form losses 

• May not model 100% of losses 

• Need ability to add form losses (calibrate) 

• Need momentum and viscosity terms 

• Linking 1D structures into 2D 

• Useful when the structure is small 

• Large structures (relative to 2D cell size) may over predict losses 

• May need to reduce inlet / outlet losses (calibrate) 

• Check and UNDERSTAND your results 

24 24 


8


1D/2D Linking 

1D/2D Interface Linking 

1D Carved Through 2D 

• Define 1D/2D Interface 

(digitise along the top or crest of the bank) 

• Connect 1D nodes to 

1D/2D interface 

(digitise perpendicular to flow) 

10.0 

10.2 

10.1 

26 26 



• If levee, digitise HX 

line along 

levee crest 

HX 

Thick Z Line needed 

HX 

• Use a Thick Z Line to 

ensure HX cell is set 

to levee crest height 

and overtopping 

occurs at correct 

height 

27 27 


9




• Cross-sections must extend from 

HX line to HX line 

(ToB to ToB) 

• Use a Thick Z Line along 

HX lines to set ToB 

levels at HX cells 

28 28 

1D/2D Interface Troubleshooting 

• Ensure Cell elevations are representative of spill levels 

–use a Thick Z Line 

• Most common cause by far is bumpy HX cell elevations 

• Poor 1D resolution 

• Missing connections 

• Add additional FLC (energy loss) works well 

(2010 version can be added directly to HX line using “a” attribute) 

29 

29 

1D/2D Structure Linking 

Culvert Through an Embankment 

• 2D water levels define 

water level gradient 

across culvert 

• Flow through culvert 

applied as a sink/source 

to 2D cells 

• Make sure width of 2D SX 

cells EXCEEDS width of 

1D flow (at all elevations) 

30 30 


10


1D/2D Pipe Network Linking 

Underground Pipe Network 

• Use pits to connect 

pipe network with 

overland 2D 

• 2D water level at 

cell drives 1D pipe 

hydraulics 

(unless pit is not full) 

• Net pipe flow in/out 

of pit applied as 

sink/source to 2D 

cell 

31 31 

Modelling Urban Areas 

Culvert Flow 

Inlet Control Regimes 

HW 

HW 

TW 

TW 

A: Unsubmerged Entrance, 

Supercritical Slope 

B: Submerged Entrance, 

Supercritical Slope 

HW 

TW 

HW 

TW 

K: Unsubmerged Entrance, 

Submerged Exit 

Critical at Entrance 

L: Submerged Entrance, 


Orifice Flow at Entrance 

33 33 


11


Culvert Flow 

Outlet Control Regimes 

HW 

TW 

HW 

TW 

No Flow 

HW 

No Flow 

TW 

Gate Closed 

C: Unsubmerged Entrance, 

Critical Exit 

D: Unsubmerged Entrance, 

Subcritical Exit 

G: No Flow 

Dry or Flap-Gate Closed 

HW 

TW 

HW 

TW 

HW 

TW 

HW 

TW 

E: Submerged Entrance, 

Unsubmerged Exit 

F: Submerged Entrance, 


H: Adverse Slope, 

Submerged Entrance 

J: Adverse Slope, 

Unsubmerged Entrance 

(Critical or Subcritical at Exit) 

34 34 

Pits 

(Drains / Gully Traps) 

• Convey the water between above ground 

and below ground 

• Recommendation is to use Q pits and y-Q curves 

• Apply appropriate curves! 

35 

35 

Pit Database 

• See Section 4.5.1.2 of 2008 Manual 

36 

36 


12


Manholes 

• Represent pipe junctions 

• Simulate energy losses at junctions 

• Must have at least one pipe in and one pipe out 

37 

37 

Junction Energy Losses 

Node or “NO” Manhole 

• “Structure Losses == ADJUST” (the default) or Channel “A” Flag 

• Inlet/Outlet Losses of pipes/manholes are adjusted based on 

approach/departure velocities 

(see Section 4.7.4.1 2008 Manual) 

• Adjusted down to zero if velocity unchanged through node 

• “Structure Losses == FIX” or Channel “F” Flag 

• Full pipe inlet/outlet losses are applied 

• Can significantly overestimate losses 

38 

38 


“FX” and “EN” Manholes 

• For “FX” and “EN” Manholes 

• Exit loss coefficients of all inlet pipes ignored 

• Entrance loss coefficients of all outlet pipes ignored 

• Manhole approach applied instead 

• Any pipe Form_Loss (bend loss) values are applied 

39 

39 


13



“FX” Manhole 

• K_Fixed attribute sets total losses for manhole 

(default = 0.0, ie. no losses) 

• Proportion/Multiplier of outlet pipe velocity head 

• Can exceed 1 

• User specified based on literature guidelines 

40 

40 


“EN” Manhole 

• Based on following loss coefficients 

• K in – expansion from water flowing into manhole 

• K Ɵ – losses due to approach-departure angles of pipes 

• K drop – drop losses due to change in pipe inverts 

• K out – contraction losses into outlet pipe(s) 

• K f – Any user specified additional fixed losses 

• Loss coefficients recalculated every timestep 

• Equations in 2010 manual 

41 

41 

Pipe Network Tips 

• Converting GIS to 1d_nwk 

• Keep backward traceability 

• Append some/all GIS attributes to 1d_nwk attributes 

• Data Integrity 

(Snapping!) 

• Use 1d_nwk_N_check 

(Nodes colour coded based on snapped channels) 

42 42 


14


Pipe Network Stability Tips 

• 1D timestep for pipe models usually in range from 0.1s to 1.0s 

• Beware of very short/steep pipes 

• Sometimes additional storage added – sensitivity test! 

43 

43 

Modifying Networks 

• Can upgrade or modify existing pipe(s) by simply overriding with 

repeat pipe(s) in separate 1d_nwk layer 

• Select and save pipe(s) to be modified 

• Save as new 1d_nwk layer 

• Modify pipe(s) 

• Add in new “Read MI Network” line 

• Cross-check using 1d_nwk_check layer and/or viewing .eof file 

44 

44 

Detailed Urban Models 

45 

45 


15


Detailed Urban Models 

• 1,600 pipes / culverts 

• 900 pits (drains) 

• 600 manholes 

• 1.8 million wet cells at peak 

• 0% Mass Error 

46 46 

Modelling Buildings? 

47 47 

Modelling Buildings 

Block Cells Out 

100% 

0.137m 

0m/s 

1.9m/s 

48 48 


16


Building Walls Blocked 

Open Upstream 

100% 

0.145m 

0m/s 

1.93m/s 

49 49 

Roughened Up Scenario 

(n = 0.3) 

79% 

0.122m 

0.38m/s 

1.41m/s 

50 50 

Porous (Blockage = 90%) 

Energy Loss (0.1*V 2 /2g) 

93% 

0.107m 

1.28m/s 

1.78m/s 

51 51 


17


Modelling Fences! 

52 52 

Urban Areas – Buildings and Fences 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

53 

Modelling Fences! 

• Able to raise element sides 

• Element sides wet and dry 

• Layered parameters 

• eg. vary blockage 

and losses with height 

• Collapse element sides 

• Switch between u/s and d/s controlled 

weir flow 

54 


18


Collapsible Fences Animation 

55 

Modelling Blockages!? 

56 

2D Layered Adjustments 

Blockage = 0% 

FLC = 0 


FLC = 0.5 


FLC = 0.8 

Blockage = 5% 

Form Loss Coeff = 0.1 

57 


19


58 58 


20

Aust Workshops - TUFLOW Modelling Bends, Structures and ...

Create successful ePaper yourself

Delete template?

Save as template?